18th ISoP Annual Meeting “Pharmacovigilance without borders” Geneva, Switzerland, 11–14 November, 2018

ABSTRACTS

18th ISoP Annual Meeting

“Pharmacovigilance without borders”

Geneva, Switzerland

11–14 November, 2018

1 International Society of Pharmacovigilance

The International Society of Pharmacovigilance (ISoP) is devoted to developing its activities on a worldwide basis towards supporting safer use of medicines in clinical practice.

ISoP aims to promote the use of all types of information and methodologies in providing optimal drug treatment for patients. The Society is not only for clinical pharmacologists, pharmaceutical industry representatives, epidemiologists and regulators, but also for practising clinicians, other healthcare professionals and anyone else who is interested in learning about better ways for patients to receive and use medicines safely.

Countries where there are ISoP members:

From Argentina to Vietnam, from countries in South-America to North-America, from Europe to Asia and Australia via Africa, we have members in all five continents.

‘‘By becoming a member of ISoP, you will have the opportunity to share your knowledge and ideas and to contribute to improving pharmacovigilance activities worldwide.’’

Sten Olsson, President of the International Society of Pharmacovigilance.

ISoP Membership incentives include:

• Biannual newsletters (ISoP Star)

• Training workshops

• Reduced fees for Annual Meeting and training courses

• Discounted online subscription to the Drug Safety journal

• Other offers/discounts on books

• Involvement in ISoP’s Chapters and Special Interest Groups

For more information you can visit www.isoponline.org, the Society’s official website.

International Society of Pharmacovigilance

ISoP Secretariat Ltd

140 Emmanuel Road, London SW12 0HS, UK

Tel and Fax: +44 (0)20 3256 0027

administration@isoponline.org

ISoP 2018 Local Organising Committee

Chair: Victoria Rollason, University Hospitals of Geneva (Switzerland)

Mario Bertazzoli, Helsinn Healthcare Lugano (Switzerland)

Marie Besson, University Hospitals of Geneva (Switzerland)

Jules Desmeules, University Hospitals of Geneva (Switzerland)

François Girardin, University Hospitals of Geneva (Switzerland)

Roseline Ing Lorenzini, University Hospitals of Geneva (Switzerland)

Christian Lovis, University Hospitals of Geneva (Switzerland)

David Niedrig, University Children’s Hospital Zurich and drugsafety.ch (Switzerland)

Frédérique Rodieux, University Hospitals of Geneva (Switzerland)

Stefan Russmann, Klinik Hirslanden Zürich and drugsafety.ch (Switzerland)

Caroline Samer, University Hospitals of Geneva (Switzerland)

ISoP 2018 Scientific Committee

Chairs: Ian C K Wong, University of Hong Kong and UCL School of Pharmacy (UK)

Mira Harrison-Woolrych (New Zealand)

Victoria Rollason, University Hospitals of Geneva (Switzerland)

Pia Caduff, Uppsala Monitoring Centre (Sweden)

Bruce Donzanti, Genentech, Inc (USA)

Noha Iessa, WHO (Switzerland)

Ambrose Isah, University of Benin City (Nigeria)

Richard Hill, Therapeutic Goods Administration (Australia)

Gurumurthy Parthasarathi, JSS University Mysore (India)

Mónica Tarapués, Central University of Ecuador (Ecuador)

Lynn Zhou, Sanofi (China)

ISoP 2018 Poster Prize Committee

Chair: Pia Caduff, Uppsala Monitoring Centre (Sweden)

Caroline Samer and Marco Tuccori

ISoP Executive Committee and Advisory Board 2016–2019

Sten Olsson, President (Sweden)

Ian C K Wong, Vice-President (UK)

Mira Harrison-Woolrych, Secretary General (New Zealand)

Jean-Christophe Delumeau, Treasurer (Singapore)

Board Members

Hilda Ampadu, Africa/Education and Training Programme (Ghana)

Brian Edwards, Coordinator Chapters (UK)

Deirdre McCarthy, Coordinator Special Interest Groups (USA)

Jan Petracek, America/Education and Training Programme (Czech Republic)

Phil Tregunno, Asia/Education and Training Programme (UK)

Marco Tuccori, Europe/Education and Training Programme (Italy)

Hervé Le Louët, Past-President (France)

Disclaimer

ISoP requests that a high standard of science is followed concerning publications and presentations at all its Annual Meetings and training courses. However, ISoP as a whole or its Advisory Board and Executive Committee (EC) or appointed Scientific Committees, or its members, do not take any responsibility for the completeness or correctness of data or references given by authors in publications and presentations at official scientific meetings.

It is not within the remit of ISoP, the Advisory Board, the EC or Scientific Committees in particular, to seek clarification or detailed information from authors about data in submitted abstracts. Moreover, it is not within the scope of ISoP and its committees to monitor compliance with any legal obligations, e.g., reporting requirements or regulatory actions.

2 O-001 Measuring the Impact of Medication Errors on Public Health: What are the Right Parameters and Criteria?

2.1 R. Soulaymani^*1, G. Benabdallah¹

2.1.1 ¹CAPM WHO CC, Rabat, Morocco

Background/Introduction: A medication Error (ME) is an unintended failure in the treatment process that leads to, or has the potential to lead to, harm to the patient^1,2 and a preventable adverse drug reaction is an injury that is the result of an error at any stage of the treatment process³. These 2 definitions seem to highlight the close relation between ME and adverse drug reactions.

Health injuries due to preventable ADRs are often responsible for hospitalizations, additional length of stay and healthcare costs which are estimated between US$ 6 billion to US$ 29 billion per year^4,5. The burden of ME has been highlighted with the results of the IOM⁶ report which revealed that ME were responsible for 7000 death per year in USA. In LMICs, the probability of patients being harmed is higher than it is in developed countries.

The particularity of ME with serious harm is that they could have legal consequences leading to law impact for healthcare professionals and health institutions.

Objective/Aim: The aim of our presentation is to determine the right parameters and criteria needed to measure the impact of ME on public health through a review of available data at the international level and through the long and efficient experience of the Centre Antipoison et de Pharmacovigilance du Maroc in the field of pharmacovigilance and the integration of medication errors as part of adverse drug reactions (ADRs).

Methods: Literature review of available data related to the assessment of the impact of medication errors on public health.

Results: It is difficult to define the limits between ADRs and MEs because in many cases MEs are preventable ADRs and then lead to overlapping, yet it can be confusing and underappreciated concept. The detection and analysis of medication errors are not well integrated in all PV systems. Furthermore, assessing the impact of medication errors on public health remains a challenge (using of specific tools as pharmaco-epidemiology studies and pharmaco-economy) even for systems leaders in medication errors management.

Conclusion: To be able to assess the impact of ME on public health, Pharmacovigilance Centres need to have an overall view of medication errors data and to develop a close collaboration with patient safety organizations, Poison Control Centres and hospitals that use specific methods to detect any failure in the systems. How are these methods efficient and sufficient to?

References:

• 1. Ferner RE, Aronson JK. Clarification of terminology in medication errors: definitions and classification. Drug Saf 2006;29(11):1011–22

• 2. Goedecke T, Ord K et al., Medication errors: New EU good practice guide on risk minimisation and error prevention. Drug Saf 2016;39(6):491–500

• 3. Reporting and learning systems for medication errors: the role of pharmacovigilance centres: terminology and definitions: consensus during the Delphi survey: http://apps.who.int/medicinedocs/fr/m/abstract/Js21625en/

• 4. Hug BL, Keohane C, Seger DL, Yoon C, Bates DW. The costs of adverse drug events in community hospitals. Jt Comm J Qual Patient Saf 2012; 38(3):120–6

• 5. Bates, David W.; Spell, Nathan; Cullen, David J., et al. The costs of adverse drug events in hospitalized patients. JAMA 1977; 277:307–11

• 6. Kohn LT, Corrigan JM, Donaldson MS. To err is human: building a safer health system. Institute of Medicine (US) Committee on Quality of Health Care in America; Washington (DC): National Academies Press (US); 2000

Disclosure of Interest: None declared.

3 O-002 Opportunities and Pitfalls When Measuring Harm Reduction Through Pharmacovigilance Activities

3.1 A. Kant^*1, L. Peters², H. Gardarsdottir³, F. van Hunsel¹

3.1.1 ¹Pharmacovigilance Centre Lareb, Den Bosch, the Netherlands; ²Master Student, Graduate School of Life Sciences, Utrecht University, the Netherlands; ³Utrecht Institute for Pharmaceutical Sciences, Utrecht University, Utrecht, The Netherlands

Background/Introduction: Pharmacovigilance (PV) activities aim to reduce harm by better use of medicines and promote public health. Assessing the impact of PV actions on public health at population level is an area currently under-investigated [1, 2]. Determining and measuring the right outcomes can be challenging and complicated due to lack of publically available data. Measuring impact should not only include an assessment of the intended but also take into account unintended effects and other simultaneous events such as changes in clinical practice, or secular trends in health outcomes.

Objective/Aim: To investigate the feasibility of measuring impact of pharmacovigilance activities based on publically available data in the Netherlands.

Methods: Five different identified safety risks related to use of medicines were assessed.

• 1. thrombotic risk due to (off-label) use of Diane-35 (cyproterone/ethinyloestradiol)

• 2. pergolide and bromocriptine and the risk of cardiac valvulopathy

• 3. proton pump inhibitors and the risk of hypomagnesaemia

• 4. rosiglitazone withdrawal due to cardiovascular effects

• 5. valproate and the risk of congenital abnormalities and developmental disorders

Assessment was done using the following approach:

• a. Based on the information on the safety signal and following pharmacovigilance actions a definition of impact on health was formulated for each signal;

• b. based on this pre-defined definition of impact of  the specific PV activities, data on change of health were searched in publically available data sources;

• c. If not available, data were searched needed to estimate the impact on health: data on change in use of the medicine or other safety measurements concerning the use of the medicine and the AR or RR of the safety issue.

For b. and c. also other factors that could have influenced the health outcomes were taken into account.

The feasibility of pre-defining a definition of impact, and the public availability of the needed data were evaluated.

Results: Determining the definition of the impact was feasible. However, measuring or estimating the impact was hampered by a lack of publically available data both on outcomes and drug use. For some safety signals (2, 4) it was challenging to determine a time frame. For only one signal (5) data on change of health were publically available, although only in a small cohort. For signal 3 no data on drug use were available, and for signal 4 data on the risk were doubtful, so an estimation of the impact was not possible. For two signals (1, 2) a crude estimation of the impact could be made with a varying degree of assumptions.

Conclusion: Lack of data hampers assessment of the impact of PV activities, but with assumptions a crude estimation of the impact on health can be possible.

References

• 1. Pharmacovigilance Activities (Rev 1) (EMA/165407/2017) [Internet]. 2017 [cited 2018 Apr 10]. Available from: http://www.ema.europa.eu/docs/en_GB/document_library/Other/2016/01/WC500199756.pdf

• 2. Goedecke T, Morales DR, Pacurariu A, Kurz X. Measuring the impact of medicines regulatory interventions—systematic review and methodological considerations: methods for measuring impact of medicines regulatory interventions. Br J Clin Pharmacol 2018;84(3):419–33

Disclosure of Interest: None declared.

4 O-003 Serious Consequences from Medication Errors Identified in Vigibase

4.1 A. Zekarias^*1, R. Chandler¹, H. Taavola¹, K. Star¹, P. Caduff-Janosa¹

4.1.1 ¹Uppsala Monitoring Centre, Uppsala, Sweden

Background/Introduction: Medication errors (MEs) are “failures in the drug treatment process that lead to, or have the potential to lead to, harm to the patient”[1]. National pharmacovigilance centres have reported MEs to VigiBase for more than two decades (first report:1994) and as per April 2018 VigiBase contains 700 000 reports on MEs. Up until today, very few ME-signals have been identified within the global database of the WHO Programme.

Objective/Aim: To perform a pilot study to investigate the feasibility of using VigiBase to identify medication errors resulting in adverse drug reactions (ADRs) for which there is an opportunity for further risk minimization measures.

Methods: Data was retrieved from VigiBase in December 2017. A drug-event combinations list was generated where ‘events’ were preferred terms from the MedDRA High-Level Group Term “Medication errors and other product use errors and issues”. The combinations were prioritized according to vigiRank [2]. The drug-event combinations represented the following criteria; at least one report received after 2014, ≤ 30 reports per combination and reports from ≥ 2 countries.

During the initial manual assessment of the drug-event combinations, the assessors reviewed a list of pre-defined questions (e.g. regarding patterns of the reported events). The drug-event combinations suggesting a need for further risk minimization measures were assessed in-depth to determine if they should be communicated within the WHO Programme.

Results: Eighty drug-event combinations were initially assessed; six (7.5%) were communicated as signals. Three signals included serious ADRs and one of these reported a fatal outcome. The signals consisted of five types of errors; dosage not adapted to renal clearance as instructed in the product information causing ADRs from overdose; incorrect drug administration rate due to unclear dosing instructions leading to reactions such as mouth ulceration and thrombocytopenia; generic substitution leading to accidental overdose resulting in bleeding; inappropriate dose administration causing dizziness and confusion affecting patients’ quality of life; and co-administration of drugs resulting in death despite warnings of drug–drug interaction in the label. The signals were communicated within the WHO Programme in April 2018.

Conclusion: Medication errors with serious consequences can be identified in a global database of spontaneous reports. While a detailed root-cause analysis can seldom be performed on such a dataset, we identified inadequate labelling, similarities in trade names as well as unclear communication by the health care professionals as actionable causes for the errors reported. Continued use of VigiBase for identifying MEs is encouraged to prevent unnecessary patient harm.

References

• 1. Aronson J.K. Medication errors: what they are, how they happen, and how to avoid them. Q J Med 2009; 102:513–52.

• 2. Caster O., Juhlin K., Watson S., Norén N. Improved statistical signal detection in pharmacovigilance by combining multiple strength-of-evidence aspects in vigiRank. Drug Safety 2014; 37(8):617–28.

Disclosure of Interest: None declared.

5 O-004 The Proforma Project Presentation

5.1 E. Aklillu^*1

5.1.1 ¹Department of Laboratory Medicine, Karolinska Institutet, Stockholm, Sweden

Background/Introduction: Supply of Medicines in Africa is increasing but inadequate pharmacovigilance regulatory capacity to process new drug registration and monitoring drug safety profiles locally remains a challenge. The increasing number of clinical trials and the large-scale mass drug administration and immunizations programs being deployed through public health programs in sub Sharan Africa underscores the urgent need to develop/strengthen the pharmacovigilance system for patient safety.

The PROFORMA project (http://proforma.ki.se/) is a joint venture between academia, regulatory authorities, and Reginal and Global pharmacovigilance stakeholders. PROFORMA is funded by European Union’s Framework Program for Research and Innovation Horizon 2020 via the European and Developing Countries Clinical Trials Partnership Association (EDCTP2).

Objective/Aim: Our main objective is to strengthen the national pharmacovigilance infrastructure and post-marketing surveillance system involving mass drug administration and immunizations being deployed under public health programs in Ethiopia, Kenya, Tanzania, and Rwanda.

Methods: The project intends to develop the national pharmacovigilance regulatory training programs by forging partnerships with local academic institutions (training-of-the-trainers for sustainable training programs) and regulatory authorities (practical training to change policy into practice) by using existing structures. Using the Pharmacovigilance Indicators tools, PROFORMA will first assess the current pharmacovigilance system and practice in Ethiopia, Tanzania, Rwanda and Kenya, to identify the missing pharmacovigilance structural elements systems, strength, deficiencies, and gaps. Based on the identified gaps, comprehensive interventional measures aligned with the local needs and priorities will be introduced in each country. PROFORMA will map all ongoing pharmacovigilance activities and tools to foster collaborations and to lower duplication. Harmonized pharmacovigilance tools and indicators will be utilized. The main regulatory functions that need capacity building will be identified and prioritized. We aim to generate a cohort of pharmacovigilance trained human resources from all stockholders including patients, healthcare providers, regulatory staffs that are engaged in pharmacovigilance data collection, analysis, interpretation and data sharing. Emphasis will be given to implement pharmacovigilance in clinical trials regulation and post-marketing surveillance in public health programmes involving mass drug administration and immunization programs. A total of 12 postgraduates will be trained to serves as part of the future PV expert regional task force. The ultimate goal is to establish/strengthen sustainable pharmacovigilance system in East Africa that is aligned with the large-scale African medicine regulatory harmonization and WHO’s Pharmacovigilance program.

Results: A total of 12 postgraduates will be trained to serves as part of the future PV expert regional task force. The ultimate goal is to establish/strengthen sustainable pharmacovigilance system in East Africa that is aligned with the large-scale African medicine regulatory harmonization and WHO’s Pharmacovigilance program.

Conclusion: Overview of the project aims and brief description of the different work packages and the associated tasks will be presented.

Disclosure of Interest: None declared.

6 O-005 Pavia: Strengthening Pharmacovigilance in Africa

6.1 F. Cobelens^*1,2, A. Isah³

6.1.1 ¹Global Health, Academic Medical Center; ²Amsterdam Institute for Global Health and Development, Amsterdam, The Netherlands; ³University of Benin, Benin City, Nigeria

Background/Introduction: PAVIA, a project funded by the European-Developing Countries Clinical Trials Platform (EDCTP), aims to improve the readiness of sub-Saharan Africa (SSA) health systems to effectively deliver new medical products for poverty-related diseases and monitor their post-market safety. It will be conducted in Nigeria, Tanzania, Ethiopia, Eswatini (formerly Swaziland) by 13 collaborating partners.

Objective/Aim: PAVIA’s objectives are: (I) To strengthen governance of pharmacovigilance (PV) systems, by strengthening regulatory and organisational structures and defining clear roles and responsibilities for all stakeholders; (II) To improve efficiency and effectiveness of national surveillance systems, by strengthening active (sentinel) surveillance of adverse drug reactions and implementation of tools and technologies for their detection, reporting, analysis and dissemination; (III) To build capacity and skills to sufficiently conduct drug safety monitoring activities throughout the country; and (IV) To improve readiness of health systems within SSA, by improving performance assessment of PV systems allowing identification of enablers and barriers for implementation.

Methods: PAVIA aims to strengthen national PV systems through a highly collaborative effort involving public health programmes (PHPs), National Medical Regulatory Authorities (NMRAs) and medical research institutions (MRIs). PAVIA will initially build up medicines safety surveillance activities in the context of the introduction of new drugs and treatment regimens for multidrug-resistant tuberculosis (MDR-TB) currently being introduced through the national tuberculosis programmes (NTPs). In each of the four project countries the national PV Centre/NMRA will collaborate with the NTP as well as a local MRI in a “triangular” cooperation in which the MRI provides clinical expertise. This “triangle” will provide a channel for reporting and interpreting safety signals in MDR-TB treatment, serve as a training ground for country PV staff and clinicians, and provide a demonstration project for similar linkages with other disease control programs.

The project will start with country-specific baseline assessments and situational analyses, which through a process of country stakeholder engagement will lead to country-specific roadmaps for PV strengthening, including recommendations and policy briefs to strengthen policies, laws and financial sustainability. Activities will include hands-on training of PV Centre/NMRA staff in database entry, data analysis, causality assessment and signal detection. PV systems (e.g. database connectivity) will be improved, and clinicians and other health care providers will be trained in PV by a blended learning approach where PAVIA will develop the materials and train “master trainers”. At the end of the project, best practices, enablers and barriers will be compiled in a blueprint that can guide other SSA countries in strengthening their PV systems. PAVIA’s alignment with relevant global and regional initiatives is ensured through their representation its Advisory Board.

Results: In view.

Conclusion: Through a highly collaborative approach between PV centers, public health programmes and clinicians, the coverage, accessibility and effectiveness of existing evidence-based interventions will be strengthened not only for MDR-TB but also for other poverty-related diseases.

Disclosure of Interest: None declared.

7 O-006 What Future Healthcare Professionals Need to Know About Pharmacovigilance

7.1 R. Van Eekeren^*1

7.1.1 ¹Netherlands Pharmacovigilance Centre Lareb, ‘s-Hertogenbosch, The Netherlands

Background/Introduction: To prevent patients from harm of adverse drug reactions (ADRs), healthcare professionals (HCPs) have to be aware of potential risks and occurrences of ADRs. To build knowledge and raise awareness about ADRs among healthcare professionals more education in the field of ADRs and pharmacovigilance (PV) is needed [1]. Currently, there is no standard for teaching PV at universities for medical, pharmacy, dentistry and nursing students and for this reason a core curriculum needs to be developed [2].

Objective/Aim: The development of a pharmacovigilance core curriculum for university teaching.

Methods: Two stakeholders meetings in 2016 and 2017 addressed and agreed the content, learning outcomes and availability for a PV core curriculum (PVCC) for university education [2]. These meetings were attended by PV professionals and university staff who are actively involved in PV education.

Results: Five key aspects summarize PV competencies that every student should develop. The PV key aspects are: (1) Importance of PV in the context of pharmacotherapy, (2) Preventing ADRs, (3) Recognizing ADRs, (4) Managing ADRs and (5) Reporting ADRs. The content of the PVCC is available on the internet, via www.pv-education.org. At this web portal educational materials and background information are available for all university teachers and PV staff that are involved in therapeutics education [2].

Conclusion: Since most medical, pharmacy and nursing students will enter clinical practice after graduation, the PVCC for university teaching focusses on the clinical aspects of ADRs. Since current educational programmes have limited time, the PVCC is developed as separate modules that can be integrated into existing courses such as pharmacology and therapeutics [3]. However, the content can also be used as a stand-alone course. A Teach-the-Teacher programme is developed which supports university staff to integrate PV key aspects into current courses.

With the five PV key aspects of the PV core curriculum for university education, future healthcare professionals will learn how to use drugs safely in clinical practice and how to share responsibility in ensuring the safety of drugs as well.

References:

• 1. Brinkman DJ, Tichelaar J, Mokkink LB, Christiaens T, Likic R, et al. Key learning outcomes for clinical pharmacology and therapeutics education in Europe: a modified Delphi study. Clin Pharmacol Ther 2017 https://doi.org/10.1002/cpt962 [Epub ahead of print]

• 2. Eekeren van R, Rolfes L, Koster AS, Magro L, Parthasarathi G, et al. What future healthcare professionals need to know about pharmacovigilance: introduction of the WHO PV core curriculum for university teaching with focus on clinical aspects. Drug Saf (accepted); https://doi.org/10.1007/s40264-018-0681-z

• 3. Hartman J, Harmark L, Puijenbroek van EP. A global view of undergraduate education in pharmacovigilance. Eur J Clin Pharmacol 2017;7: 891–9

Disclosure of Interest: None declared.

8 O-007 Anywhere, Anytime: Distance Learning in Pharmacovigilance

8.1 A. Hegerius^*1, P. Caduff¹, R. Savage^1,2, J. Ellenius¹

8.1.1 ¹Uppsala Monitoring Centre, Uppsala, Sweden; ²Department of General Practice, University of Otago, Christchurch, New Zealand

Background/Introduction: There is a growing demand for globally accessible training in pharmacovigilance. The Uppsala Monitoring Centre (UMC) has a long history of supporting the member countries of the WHO Programme for International Drug Monitoring, e.g., through the annual pharmacovigilance course in Uppsala. However, many people are unable to attend this course and the UMC has therefore launched a new distance learning platform with a first course in signal detection and causality assessment [1].

Objective/Aim: To develop and evaluate a distance learning course in pharmacovigilance.

Methods: The course was developed using the microlearning approach. The content, created using Articulate 360 Rise [2], was divided into four modules subdivided into 27 lessons. Each lesson is a short video lecture with a transcript and a clear learning objective. There is a quiz at the end of each module.

Initially, a small pilot study was performed to assess the preferred format of the course. Four months after the launch, an anonymous user survey was performed to assess usability and usefulness. The evaluation model was inspired by a refined version of the unified theory of acceptance and use of technology (UTAUT) [3, 4]. The questionnaire contained 25 questions and was sent to 2067 course participants from 137 countries, out of which 841 had completed the course and 1226 had at least started.

Results: Five hundred and fifty course participants responded to the survey (26.6%). The majority came from the WHO region SEARO (43.3%), followed by EURO (24.2%) and PAHO (19.3%). Twenty-two percent of the course participants were pharmacovigilance centre staff (our primary target audience). However, the course was also very popular among other pharmacovigilance professionals. A selection of questions and answers from the survey are presented in the table.

Almost all course participants (98.5%) responded that they would be able to apply what they had learnt to their practice. When asked to rate the course, 49.9% selected ‘Excellent’, 47.3% ‘Good’ and 2.8% ‘Average’. No one selected ‘Poor’ or ‘Inadequate’. A high proportion, 99.2%, would recommend the course to others and 99.8% intend to take future UMC courses relevant to them. Suggestions for improvements mainly concerned requests for more practical examples and quiz questions, improved display of transcripts and a forum for interaction.

Conclusion: Extending the UMC training portfolio with distance learning was well-received with high overall course satisfaction. Microlearning with video lectures, transcripts and quizzes was an appropriate educational strategy.

References:

• 1. Uppsala Monitoring Centre. Online courses [Internet]. Uppsala: Uppsala Monitoring Centre; 2017. [cited 2018 May 25] Available from: https://www.who-umc.org/education-training/online-courses

• 2. Articulate. Articulate 360 Rise, [authoring app]. New York: Articulate Global, Inc; 2017. [cited 2018 May 25]. Available from: https://articulate.com/360/rise

• 3. Chen, Jian-Liang. The effects of education compatibility and technological expectancy on e-learning acceptance. Comput Educ 2011;57:1501–11

• 4. Venkatesh, V, Morris, MG, Davis, FD. User acceptance of information technology: toward a unified view. MIS Q 2003;27:425–78

Disclosure of Interest: None declared.

9 O-008 Learning by Doing: Implementation of Pharmacovigilance in the Clinical Setting in a National Referral Hospital in Kenya

9.1 F. A. Okalebo^*1, M. N. Oluka¹, A. N. Guantai¹

9.1.1 ¹School of Pharmacy, Nairobi, Kenya

Background/Introduction: Training in detection of adverse drug reactions and reporting can be theoretical and learners may fail to see how pharmacovigilance can be incorporated into routine patient care. In an effort to move away from didactic teaching, the Department of Pharmacology and Pharmacognosy, University of Nairobi made efforts to promote bedside training in adverse event detection and reporting.

Objective/Aim: The objective is to present a case study of implementation of training in pharmacovigilance in Clinical rotations in a National Referral Hospital in Kenya. Opportunities and challenges will be highlighted.

Methods: A qualitative case narrative approach will be used to describe experiences in implementing practical training in pharmacovigilance.

Results: The trainers’ came up with 3 clinical rotation sites where the disease conditions present a large opportunity for detection of adverse drug reactions. These clinical sites were outpatient hematological clinic, in-patient psychiatric ward and out-patient HIV/AIDS clinic. After discussion with the pharmacist in-charge of the HIV/AIDS clinic, the in-patient multi-drug resistant clinic and in-charge of and a female medical ward were added to the rotation sites. The lecturers came up with a list of competencies expected in each of these sites. Based on these, a log book and tools for use by learners were developed. Learners were required to families themselves medication related expected in patients on anti-retroviral, anti-tuberculosis and psychiatric drugs. After 1 month of implementation of the clinical rotations, some of the challenges identified were an overwhelming occurrence of adverse drug events that necessitated a change in the rotations schedule; inappropriate tools; understaffing; and inappropriate tools. These problems were resolved through discussion with learners. Learners expressed satisfaction with the clinical rotations.

Conclusion: Incorporation of pharmacovigilance activities in clinical rotations provide a rich opportunity for training in pharmacovigilance and integration of medication management in routine clinical practice.

Disclosure of Interest: None declared.

10 O-009 Long Term and Clinical Effects of an Pharmacovigilance Educational Intervention in Specialist Oncology Nurses

10.1 M. Reumerman^*1, J. Tichelaar¹, R. van Eekeren², M. Richir¹, M. van Agtmael¹

10.1.1 ¹Section Pharmacotherapy, VUmc, Amsterdam, The Netherlands; ²Netherlands Pharmacovigilance Centre Lareb, ‘s hertogenbosch, The Netherlands

Background/Introduction: The level of underreporting of adverse drug reaction (ADR), especially under nurses, is high (0.9% reporting rate/year) and hinders optimal ADR monitoring. Since (specialist oncology) nurses administer most drugs they could play a major role in monitoring and reporting ADRs. Only several interventions in healthcare professionals have proven effective in increasing knowledge, however failed to produce clinical and durable effects.

Objective/Aim: This study aims to investigate the clinical and long-term effects of a pharmacovigilance educational intervention in specialist oncology nurses (SONs).

Methods: A case–control intervention study in the three postgraduate school who offer the course “prescribing qualification” for SONs was initiated. An innovative educational intervention in pharmacovigilance was added in one postgraduate school. The clinical value was assess by analyzing the number of reported ADR-reports to the Netherlands pharmacovigilance center. SONs competences on ADR-reporting were evaluated using e-questionnaires at T0: direct after graduation and T1: 1 year after graduation.

Results: Thus far, sixty SONs in the intervention group (74% of total) and ten SONs (59% of total) in the control group were included. Seventy ADRs were reported during and after the intervention while the control group did not report any. Six SONs (10%) completed a report within 1 year after the course, which showed an 11 time increase compared to baseline nurse ADR reporting rates. SONs in the intervention group showed an sustainable long-term knowledge-score (T₀: 82%/T₁: 83%) while the control group had a continuous lower level of knowledge (T₀: 65%/T₁: 69%). SONs in the intervention group found themselves more relevant in reporting ADRs than the control group (3.66 ± 0.36 vs 1.75 ± 0.59, 1–5 min/max).

Conclusion: This is the first study that shows a significant and relevant increase in the quantity of reported ADRs after a single educational intervention. The intervention had a sustainable effect on pharmacovigilance knowledge and increases SONs responsibility for their role in ADR-reporting.

Disclosure of Interest: None declared.

11 O-010 Communication: The Key to Implementation of Risk Management and Patient Safety

11.1 P. Bahri^*1,2, K. Ilic ³, U. Hagemann⁴

11.1.1 ¹ISoP CommSIG Coordinator,; ²European Medicines Agency (in personal capacity), London, United Kingdom; ³ISoP Comm SIG Member, Washington, DC, United States; ⁴ISoP CommSIG Member, Berlin, Germany

Background/Introduction: Over the last 10 years, aspirations regarding communication in pharmacovigilance have evolved from linear information flows with educative and instructive intentions to facilitating a dialogue with and within healthcare, involving patients for shared therapeutic decision-making and understanding of how to use medicines safely. We experience however that important information does still not reach those who need it and that the dialogue between patients and healthcare professionals does often not happen at all. Moreover the adoption of safe use behaviours by healthcare professionals as required by risk management programmes is not easily successful, with different challenges in countries around the globe.

Objective/Aim: This session intends to introduce a new way of thinking of risk minimisation measures as interventions within complex systems, and making communication the key to their implementation.

Methods: There will be a range of flash talks and time for discussion of conference participants for sharing their experiences and ideas. This session is sponsored by the ISoP CommSIG on medicinal product risk communication in collaboration with the ISoP special interest groups on women health and medication errors.

Results: A range of flash talks will offer with perspectives from patients, medicines safety specialists from four continents and a guest expert from complex intervention practices, underpinned by real life examples of interventions taking into account human factors to foster safe use behaviours and avoid medication errors.

Conclusion: The session will conclude with a wrap up by the co-chairs on key messages and suggestions for the presenters and participants.

Disclosure of Interest: P. Bahri: None Declared, K. Ilic Shareholder of: Consultant for pharmaceutical industry, U. Hagemann: None declared.

12 O-011 Human Factor Science for Improving the Impact of Risk Communication

12.1 B. Edwards^*1

12.1.1 ¹UK Pharmaceutical Ergonomics & Human Factors Group, Ashtead, Surrey, United Kingdom

Background/Introduction: Ultimately safe and proper use of healthcare products depends on human performance within an effective and supportive system. Unfortunately, this has not been previously thought through by applying organisational science. That is why in November 2015, we established a Pharmaceutical Human Factors & Ergonomics Special Interest Group (PHF&E SIG) within the Chartered Institute Ergonomics & Human Factors with the aim of promoting debate and develop programmes of work that include the capacity and capability of human factors engineering (HFE) and identify what support is needed to enable pharmaceutical organisations of all forms and sizes to embed HFE principles and practices into their culture, systems and processes.

Objective/Aim: This unique group is open to all who work in the pharmaceutical sector at any level and in any discipline who are committed to enhancing human performance towards greater safety and efficiency.

Methods: Since the inception of the PHF&E SIG, a series of sub-groups have been set up with membership from different parts of the system looking at various aspects of improving human performance. The four active groups concern: (1) modelling the system for medicines in the UK, (2) introducing human factors into existing training courses for drug safety, clinical research and pharmacists developing cross-system human factors competencies, (3) evaluating how best to perform human factors studies with healthcare products, (4) how to apply technology in the system and (5) implementation of principles of human-centred organisations (ISO 27500).

Results: From the experiences of these sub-groups we have identified how barriers in adopting safe use behaviours continue to exist despite intensive efforts of risk management programmes. Part of this failure is not ever having analysed the complexity of healthcare systems and designing processes that are evidenced-based, adaptable and flexible for those who work in them. Thus poor communication is inevitable. The lack of training about working and coping in complex systems is notable. We are not aware of sufficient evidence demonstrating the dynamics of how individual human factors (intrinsic) interact with social and organisational issues (extrinsic human factors) impact on the quality and outcomes of risk communication for healthcare products.

Conclusion: Barriers in adopting safe use behaviours continue to exist despite intensive efforts of risk management programmes. Part of this failure is not analysing the complexity of healthcare systems and designing processes that are evidenced-based, adaptable and flexible for those who work in them. The lack of training about working and coping in complex systems is notable. We are not aware of sufficient evidence showing how individual on one hand and social and organisational issues (human factors) impact on the quality and outcomes of risk communication for medicines. By increasing research for such evidence and blending human factors science with good communication practice, outcomes should improve.

References:.y increasing practical real-life research about how to apply such organisational evidence and blending human factors science with good communication practice, outcomes should improve.

Disclosure of Interest: None declared.

13 O-012 Health Care Communication Empowerment: A Successful Programme in Colombia

13.1 A. Caro-Rojas^*1

13.1.1 ¹President, Asociación Colombiana De Farmacovigilancia, Bogotá, Colombia

Background/Introduction: The communication is a process involve in all moment in our life, we are always in constant expressing ideas verbally or not verbally. The communication influence the work environment and in a hospital and affect the system working and the results with the patient.

Nowadays the human factors are a relevant topic in patient safety, It is possible to implement strategies to improve the communication (teamwork, empathy, collaborative leadership,) and in this way reduce the patient safety events included medication errors. The good communication include writing and reading properly, ask and discuss, this tools are necessary in the medicine dispensing and administration. Understand the information of the labelling medication helps to control medication errors.

Objective/Aim: To describe the experience with pharmacy and medicine students, using roleplay and theatre practices.

Methods: The communication skills can be teached to university students, it show a experience in Pharmacy and medicine students who understand the importance of recognize the other professional through roleplay and theatre practices. Parallelly the use of similar strategies was implemented in a institution with good results in decrease of medication events. The strategies was called Health Care Communication empowerment defined as: use of non-conventional tools for the development of communication skills, which the empowerment of the actors in the provision of health services, including the patient, their family and caregivers.

Results: A total of 400 aprox. has been involved in this practice. All the students recognize de importance of the medication safe use and consider that strategy using roleplay are linked with emotions and sensitize.

In other scenarios, nurses and doctors have been sensitized in the safe use of medications, with this strategy the respective decrease in adverse events was obtained.

Conclusion: Innovative strategies can be implemented to sensitize the persons in safe use of medicines and minimize medication errors.

Disclosure of Interest: None declared.

14 O-013 Central to Risk Communication in Healthcare: The National Pharmacovigilance Centre in Tunisia

14.1 M. Bouhlel¹, R. Daghfous ^*1, S. El Aidli¹

14.1.1 ¹Tunisian National Centre of Pharmacovigilance, Tunis, Tunisia

Background/Introduction: The Tunisian pharmacovigilance system essentially includes direct communication with patients during a dedicated consultation. In addition, there is communication with the healthcare professionals with regard to their patients. Thereby, the pharmacovigilance centre establishes communication also between the patient and their healthcare professionals.

Objective/Aim: The work aims to present of the Tunisian pharmacovigilance system and to demonstrate its efficiency in risk communication between patients and healthcare stakeholders.

Methods: A retrospective study was made during the last decade in the National Centre of Pharmacovigilance based on statistical data collected from consultation with patients, patients’ family members, doctors and other stakeholders.

The study was also focusing on the different kinds of issues raised.

Results: The majority of consultation was made with patients, then with the patient’s family members, and later with the doctors and the other stakeholders.

The results showed that detected issues were related to the various side effects due directly to medicinal products, to the misusage of medicines and not to the suspected used drug itself.

The reports of the pharmacovigilance investigation were submitted directly to each doctor treating patients and/or to each patient, after a consultation explaining the content of the report.

Conclusion: The Tunisian pharmacovigilance centre supports a dialogue between all different parties when investigating medication errors and identifying the root cause through consultation made with all stakeholders.

Disclosure of Interest: None declared.

15 O-014 National Campaign in an Israeli Chain of Community-Based Pharmacies to Decrease the Risks of Anticoagulants

15.1 R. I. Fermont^*1,2, A. Amitai^3,4, A. Livneh^1,5, V. Kalamaro^6,7, R. Litman^8,9, P. Pitts¹⁰, N. Sevdalis¹¹, P. Bahri^12,13, B. Edwards^14,15, A. Elad¹⁶

15.1.1 ¹ISOP ISRAEL Chapter, Israel; ²IFC Strategic Safety Consulting, Jerusalem, Israel; ³The Pharmaceutical Society of Israel, Tel Aviv; ⁴Clinical Pharmacology, Meir Medical Centre, Kfar Saba, Israel; ⁵Ministry of Health, Jerusalem, Israel; ⁶Clinical Pharmacists, The Pharmaceutical Society of Israel, Tel Aviv, Israel; ⁷Clinical Pharmacy, Cystic Fibrosis Israeli Foundation, Petah Tikva, Israel; ⁸Anesthesiology and Critical Care Medicine, The Children’s Hospital of Philadelphia, Philadelphia, United States; ⁹Medical Direction, Institute for Safe Medication Practices, Horsham, United States; ¹⁰Center for Medicine in the Public Interest, New York, United States; ¹¹Centre for Implementation Science, Kings College London, United Kingdom; ¹²SIGCOMM, ISOP, ¹³Pharmacovigilance, European Medicine Agency, ¹⁴Board of Directors, ISOP, London, United Kingdom; ¹⁵Pharmacovigilance, NDA, Leatherhead, United Kingdom; ¹⁶The Healthcare System Healing Campaign, Zichon Yaakov, Israel

Background/Introduction: The community pharmacists are often the last rampart before drug administration to the patients. Therefore, they should play a significant role in risk prevention, medication errors or interactions detection, and safe use advice to patients.

An Israeli chain of community-based pharmacies is launching a national campaign on the risks of anticoagulants. The project is led by the Pharmaceutical Society of Israel (PSI), in partnership with ISOP ISRAEL.

Objective/Aim: 1. To decrease the risk of anticoagulants via improving risk awareness.

2. To set up a safety methodology applicable to other high-risk medications.

3. To improve the perceived value of the community pharmacist as a Public Health significant actor.

Methods: 1. Training of all the chain pharmacists on risk communication for anticoagulants with key messages and key questions to the patients.

2. Tools identification: ISMP (Institute for Safe Medication Practices) “Self-assessment tools”^1,2,3 are the basis for the key messages. Other tools will be identified by ISOP ISRAEL with the support of ISOP risk communication group and ISMP.

3. On-site advising: For each patient prescribed with anticoagulants, several parameters will be checked: first symptoms awareness for bleeding and thromboembolism leading to immediate medical consultation, dosage, indication, drug and food interactions, compliance and detection of potential adverse reactions. Patients will receive also information leaflets and a safety card.

4. Key performance indicators (KPI) will be integrated the distribution software as tick box to check the number patients advised, medication error or adverse reaction detection.

5. Data collection: In each of the five regions, once a month, a coordinating pharmacist will select and share case(s) where pharmacist’s advice had the most significant impact. 

Results: Timeline.

• Project start: June 2018

• June–October 2018: preparation of training and tools identification

• October 2018: community pharmacists training

• November 2018: starting the national campaign

• June 2019: Interim results presentation in the ISOP ISRAEL symposium

• October 2019: Final results presentation in ISOP Bogota 2019

Conclusion: For this original initiative, we will have to adjust tools and methods during the project to prove its effectiveness and feasibility. We can reasonably expect this campaign to trigger/increase risk awareness on anticoagulants in Healthcare Professionals and patients. The partnership between PSI, ISOP ISRAEL, ISOP and ISMP is innovative and brings all stakeholders together—which, if successful, provides a model to use for further campaigns on other high-risk medications.

References:

• 1. ISMP Medication Safety Self Assessment^® for Community/Ambulatory Pharmacy, January 23, 2017, https://www.ismp.org/sites/default/files/attachments/2018-01/ISMP117C-Pharma%20SA-FINAL%20020317.pdf

• 2. ISMP Medication Safety Self Assessment^® for High-Alert Medications January 25, 2018, https://www.ismp.org/sites/default/files/attachments/2018-01/EntireAssessmentWorkbook.pdf

• 3. Medication Safety Self Assessment^® for Antithrombotic Therapy, March 1, 2017, https://www.ismp.org/sites/default/files/attachments/2017-11/ISMP128-Antithrombotic%20SA-021017.pdf

Disclosure of Interest: None declared.

16 O-015 Medical Devices: Definition, Classifications, and Nomenclatures

16.1 J. Aronson^*1

16.1.1 ¹Centre for Evidence Based Medicine, University of Oxford, Oxford, United Kingdom

Background/Introduction: Medical devices are being increasingly used.

Objective/Aim: Defining a medical device: The WHO’s definition [1] is:

Any instrument, apparatus, implement, machine, appliance, implant, reagent for in vitro use, software, material or other similar or related article, intended by the manufacturer to be used, alone or in combination, for human beings, for one or more of the specific medical purpose(s) of:

• diagnosis, prevention, monitoring, treatment or alleviation of disease,

• diagnosis, monitoring, treatment, alleviation of or compensation for an injury,

• investigation, replacement, modification, or support of the anatomy or of a physiological process,

• supporting or sustaining life,

• control of conception,

• disinfection of medical devices,

• providing information by means of in vitro examination of specimens derived from the human body;

and does not achieve its primary intended action by pharmacological, immunological or metabolic means, in or on the human body, but which may be assisted in its intended function by such means.

Methods: Analysing the definition: A device is “something devised or framed by art or inventive power; an invention, contrivance; especially a mechanical contrivance for some particular purpose” (Oxford English Dictionary). The WHO’s definition attempts comprehensive coverage of all imaginable types of device. However, the items included overlap in meaning, and as new devices appear the definition will become outdated.

Results: A proposed definition: A simpler definition of medical devices would encompass their important features. The following proposed definition builds on that published by the Bundesinstitut für Arzneimittel und Medizinprodukte [2]:

Medical device n: A product intended to benefit human healthcare, whose effects are primarily achieved by physical, chemical, or computerized means, rather than pharmacologically, immunologically, metabolically, or cellularly.

Classifications: The FDA has classified about 1700 types of devices in 16 areas into three different classes, I, II, and III, with different regulatory requirements, depending on how much control is required for a favourable benefit to harm balance [3].

In contrast, the EU classifies devices into four categories, I, IIa, IIb, and III, based on purpose, duration of use, invasiveness, use of a source of energy, and whether something is measured [4].

Nomenclatures: The Universal Medical Device Nomenclature System™ (UMDNS) is a standard international nomenclature and computer coding system for medical devices [5]. The Global Medical Device Nomenclature (GMDN) is a list of generic names used to identify all medical device products [6].

Conclusion: Harmonization and simplification of classifications and nomenclatures is desirable.

References:

• 1. http://www.who.int/medical_devices/full_deffinition/en

• 2. https://www.bfarm.de/EN/MedicalDevices/_node.html;jsessionid=9133BE8F49877DB89997E9920F72C89D.2_cid329

• 3. https://www.fda.gov/MedicalDevices/DeviceRegulationandGuidance/Overview/ClassifyYourDevice

• 4. https://www.gov.uk/topic/medicines-medical-devices-blood/medical-devices-regulation-safety

• 5. https://www.ecri.org/components/UMDNS/Pages/default.aspx

• 6. https://www.gmdnagency.org

Disclosure of Interest: None declared.

17 O-016 A Multidisciplinary Management of Oncologic Patients Treated with Immune Checkpoint Inhibitors (Icpi)

17.1 C. Le Beller¹, P. Mignen¹, E. Fabre², L. Weiss³, J. Pouchot⁴, G. Malamut⁵, M.-A. Dragon-Durey³, E. De Guillebon², A. Lillo-Le Louet^*1, S. Oudard²

17.1.1 ¹Pharmacovigilance Hopital Européen Georges Pompidou, Paris, France; ²Oncology, ³Immunology; ⁴Internal Medicine; ⁵Gastroenterology, Hôpital Européen Georges Pompidou, Paris, France

Background/Introduction: ICPI approvals have changed the management and prognosis of advanced cancer. However, ICPI have a specific toxicity profile which modifies the oncologists’ practices. Immune-related adverse effects (irAEs) can affect all organs, far from usual cytotoxic AEs observed with classical anticancer treatment [1, 2]. Considering their action on the immune system, patients with underlying disease (autoimmune, graft recipients, chronic infections) were not evaluated in clinical trials. Since January 2017 in our hospital, a multidisciplinary team with monthly meetings (MTM) discusses ICPI indications in patients with an underlying disease, manages irAEs or decides ICPI continuation. The MTM gathers oncologists and subspecialists in immunology, internal medicine, gastroenterology, endocrinology, nephrology, radiology and pharmacovigilance.

Objective/Aim: To present an overview of patient’s cases assessed by MTM in 2017.

Methods: We performed a retrospective study of the patients’ cases presented in the MTM during 2017. We used MTM reports and patients’ clinical data from the hospital computerized medical record.

Results: Among the 172 patients who received ICPI, 41 circumstances for 37 patients (18 women and 19 men, aged from 44 to 86 years) were analyzed and discussed in 2017.

16 patients were evaluated prospectively due to an underlying condition:

• an autoimmune disease: rheumatoid arthritis (1), multiple sclerosis (1), Sjögren’s syndrome (1), positivity of ANA (3)

• a chronic viral disease: hepatitis C (2), HBV (1), HIV (1)

• heart transplantation (1)

• other disorders: lymphocytosis (2), allo-immunization HLA (1), idiopathic pulmonary fibrosis (1), hyperthyroidism (1).

In all these 16 patients but one (heart transplantation), the ICPI benefit/risk ratio was considered as favorable.

For the remaining 21 patients, the MTM analyzed 25 AEs: tumor progression or pseudoprogression (5), hepatic injury (5), hyperthyroidism (4), cutaneous reaction (3), cytopenia (2), colitis (2), pulmonary fibrosis, oligoarthritis, weight gain and a suspected drug interaction with phytotherapy (one of each case). For 2 hematological AEs, the relationship with ICPI was ruled out. After discussion, ICPI was continued in 15 patients associated with corticotherapy or curative treatment (11). Additional investigations (biological survey, TDM, liver biopsy, colonoscopy…) were recommended for 8 patients. ICPI were withdrawn in 6 patients.

Conclusion: ICPI is a new class of drugs available for cancer treatment with growing indications and a specific and unusual safety profile. It is a challenge in current oncology practice as several questions are still pending about its benefit in some population and/or in association with other drugs. A new organization with MTM helps clinicians to share experiences with a direct benefit for their patients.

References

• 1. Michot JM, Bigenwald C, Champiat S et al. Immune-related adverse events with immune checkpoint blockade: a comprehensive review. Eur J Cancer 2016;54:139–48

• 2. Sosa A, Lopez Cadena E, Simon Olive C et al. Clinical assessment of immune-related adverse events. Ther Adv Med Oncol 2018; 10: 1–11

Disclosure of Interest: None declared.

18 O-017 Management of Adverse Experiences from the Use of Herbal Medicines: The Consumers’ Perspectives

18.1 C. Kongkaew^*1, M. Sreerattanachotchai¹, S. Sudnongbua²

18.1.1 ¹Faculty of Pharmaceutical Sciences, Naresuan University, Thailand; ²Department of Community Health, Faculty of Public Health, Naresuan University, Phitsanulok, Thailand

Background/Introduction: Herbal medicines have been used worldwide. There are several the herbal medicine reporting mechanisms, especially spontaneous reporting system, but the number of reports were low and may be underreported. To improve reporting, it is crucial to understand how consumers perceive adverse events from herbal medicines and act upon them.

Objective/Aim: To describe consumer perceptions of how they manage adverse experiences following use of herbal medicines.

Methods: A multiple case study was undertaken in sub-district, in the south-east of Thailand during March/April 2017. In-depth interviews were employed as the method of data collection. Nine participants who had had adverse experiences were selected by criterion sampling.

Results: Key informants perceived the adverse experiences during a course of herbal medicine treatment. They self-diagnosed the adverse experiences, they observed abnormalities of the body using their own experiences. Some of the herbalists did not know that the experience was an adverse reaction to herbal medicine that it may also be called “Sai” symptom. Identification of suspected herbal medicines was deduced by (i) ruling-out other potential causes (ii) a recovery after cessation of treatment and (iii) the return of the symptoms after re-administration of the herbal. The perceived adverse reactions were managed by: (a) taking no action if symptoms were mild and that herbal medicines were believed to be safe (b) self-care including dose moderation or discontinuation, sleeping, drinking a lot to increase excretion, or detoxification by taking laural clock vine (Thunbergia laurifolia), (c) seeking advice/care at sub-district health promoting hospital or drugstore and (d) perform both self-care and care-seeking behaviors.

Conclusion: Consumers of herbal medicine were at risk of adverse events. Therefore, surveillance systems for herbal medicines should be established by promoting consumer and/or healthcare professional reporting in primary care thereby improving the safe use of herbal medicines.

Disclosure of Interest: None declared.

19 O-018 Efforts Made by the University in Costa Rica to Promote Pharmacovigilance

19.1 V. Hall Ramirez^*1

19.1.1 ¹Pharmaceutical Care and Clinical Pharmacy Department, Pharmacy Faculty, University of Costa Rica, San Jose, Costa Rica

Background/Introduction: The Pharmacy Faculty of the University of Costa Rica, compromise, since 1897 with the Costa Rica`s population health, through the training of pharmaceutical professionals, the provision of services and the development of activities that directly affect the public health of the country, had made several efforts to strengthen the role of the pharmacist in ensuring the safety of medicines used by patients, not only through a new curriculum but also with the participation of the professors in several activities regarding Pharmacovigilance.

Objective/Aim: Describe the efforts made the Pharmacy Faculty of the University of Costa Rica and the challenges the academia have in promoting Pharmacovigilance.

Methods: Study Type: Descriptive. Method: Compilation of the activities carried out by the Pharmacy Faculty around Pharmacovigilance integrating the three fundamental pillars of the University of Costa Rica: teaching, research and social action.

Results: In Costa Rica, the National System of Pharmacovigilance, allows the reporting of suspected adverse drug reactions by health professionals and also directly by the pharmaceutical industry, that’s the reason why the training and the continuing education on this topic is essential for the correct functioning of the System.

The Pharmacy Faculty developed a new curriculum in 2016, in which it can be identify several courses related to Pharmacovigilance, defined as the science and activities relating to the detection, assessment, understanding and prevention of adverse effects or any other drug-related problem [1]. The topic is included in different courses with complementary approaches, as for example Pharmaceutical Care, Public Health Elements, Pharmacoepidemiology Topics, Pharmaceutical Legislation and Deontology, Industrial Pharmacy and Introduction to Pharmacy.

Considering that the counterfeit medicines are also part of Pharmacovigilance, the theme is part of the topics included in Pharmaceutical Analysis and Quality Control courses. These activities are complemented by the Pharmaceutical Research Institute and it’s dependencies through their everyday activities and several projects.

On the other hand, the Faculty has developed, joining efforts with different stakeholders, including pharmaceutical associations, industry and health authorities, several continuing education courses and workshops for pharmacist, so they can be more proactive in the filling of the Individual Case Safety Report (ICRS) and send them to the Costa Rican National Pharmacovigilance Center.

Conclusion: The Pharmacy Faculty of the University of Costa Rica has made several efforts to promote the development and strengthening of Pharmacovigilance, working together with different stakeholders, but all the activities done by now, can be improved so it can be guaranteed to the population that the medicines they use are safe and that the professionals are making their biggest effort to ensure that.

References

• 1. World Health Organization. Pharmacovigilance. [Internet] [Accessed on June 4th, 2018]. Available in http://www.who.int/medicines/areas/quality_safety/safety_efficacy/pharmvigi/en/

Disclosure of Interest: None declared.

20 O-019 How the Colombian Pharmacovigilance Association Join the Stakeholders in Pharmacovigilance, and Promote the Best Practices in Latam?

20.1 A. Caro-Rojas^*1

20.1.1 ¹President, Asociación Colombiana De Pharmacovigilancia, Bogotá, Colombia

Background/Introduction: The Colombian Pharmacovigilance Association (ACFV) was founded 2 years ago with a propose “To be a national and international leader, who integrates all the actors who are interested in patient safety and particularly in the safe and efficient use of the medicines.” In this propose the Association has been working with the academy, government, pharmaceutical industry, scientific associations, hospitals, and other stakeholders, earing, learning and understanding about the needs of each one and implementing strategies that answer that problems. The association has worked in regulatory and academic topics, and specially in networking with other Associations and nets in Latam in the project “Bogotá Declaration”.

Objective/Aim: To describe how the ACFV, has been integrated the different stakeholders to share information and develop strategies for medication use in Colombia and Latinamerica.

Methods: The ACFV has been working with the academy, government, pharmaceutical industry, scientific associations, hospitals, and other stakeholders, earing, learning and understanding about the needs of each one and implementing strategies that answer that problems.

The ACFV has doing a net of nets with the participation of people from Mexico, Costa Rica, Ecuador, Peru, Bolivia, Chile, Argentina and of course Colombia. In this net and by social media networks the ACFV share information for use safe of medicines and identify needs of information.

In this way the ACFV has developed virtual conferences to help to share pharmacovigilance information.

Other strategy has been workshops to build solutions like communication between stakeholders (Industry, academy and governement).

Results: In this moment the Virtual net is about 2700 persons in the basic structure, and about 10,000 persons for the other networks interacting for all the stakeholders.

Conclusion: Is important to have strategies that joint stakeholders about pharmacovigilance topics.

Disclosure of Interest: None declared.

21 O-020 Systems Factors Associated with the Use of Intravenous Insulin Infusions: Current Evidence and Future Directions

21.1 B. Edwards^*1

21.1.1 ¹UK Pharmaceutical Ergonomics & Human Factors Group, Ashtead, Surrey, United Kingdom

Background/Introduction: We know that human and other systematic causal factors underlie most medication errors. One of the challenges is how best to persuasively demonstrate this. UK National Health Service Never Events may be a good place to start. Overdose of Insulin due to abbreviations or an incorrect device is an agreed NHS Never Event [1]. The European Medicines Agency has produced specific guidance on prevention of medication errors with high strength insulins [2]. Although much focus has been placed on technical aspects of insulin use, less emphasis has been placed on systemic causal factors and barriers to successful intravenous insulin infusions. Such a case study might then be extrapolated to other high risk medicines and therapeutic areas to illustrate the importance of human factors and systems theory.

Objective/Aim: A workshop was organised at the School of Pharmacy, University of Reading, UK on April 20th 2018 which focussed on human factors at the level of the system, the process and the individual to identify systems-based solutions to optimise use of intravenous insulin. This involved representatives from healthcare, the industry, patients and human factors specialists.

Methods: At a systems level, participants were asked to reflect on hierarchy and leadership, decision-making, team working and collaboration, communication and evidence dissemination, incentives and motivators, principles of quality management leading to controls and feedback, standards and competencies. Discussion of the process concerned the people (role definition), training, working environments (how they differ), error management, metrics and measurement.

Results: In the context of systems and processes there was discussion about quality of evidence on which decisions are made and design and suitability of products available.

Conclusion: A report will be created from this workshop which will make recommendations about how best to disseminate current research findings about i.v. insulin, engage key stakeholders in shaping future directions for research/practical applications and identify barriers and gaps hindering optimal use of intravenous insulin and achievement of the NHS Never Event target. There will be initial proposals for system redesign including the need for a larger conference or meeting to move this topic forward.

References:

• 1. Never Events list 2018 NHS Improvement January 2018 https://improvement.nhs.uk/documents/2266/Never_Events_list_2018_FINAL_v5.pdf

• 2. Guidance on prevention of medication errors with high-strength insulins European Medicines Agency 27 November 2015 EMA/134145/2015 http://www.ema.europa.eu/docs/en_GB/document_library/Recommendation_on_medication_errors/2015/11/WC500197133.pdf

Disclosure of Interest: None declared.

22 O-021 Strengthening HIV Pharmacovigilance in China

22.1 J. Bao^*1

22.1.1 ¹Frontier Biotechnologies, Inc, Nanjing, China

Background/Introduction: AIDS is a harmful infectious disease caused by the HIV virus that attacks the body’s immune system. HIV has caused more than 35 million deaths worldwide and 1 million people worldwide died of HIV-related illnesses in 2016. However, it is estimated that only 70% of people living with HIV are aware of their infection status. Since the first case reported in China in 1885, the incidence of AIDS has grown rapidly. In 2016 (January 1st, 2016 to 22:00, Dec. 31st), the nationwide (excluding Hong Kong, Macau, Taiwan, the same below) reported 6,944,240 communicable diseases and 18,237 deaths. Among them, the number of AIDS cases was 54,360, an increase of 8.0% over the previous year. The incidence of AIDS was 3.9656/100,000 and the number of AIDS deaths was 14,091, an increase of 10.5% from 2015. The mortality rate of AIDS was 1.028 per 100,000.

Objective/Aim: HIV treatment is complex and requires a life-long combination medication. Therefore, a well-designed pharmacovigilance system for HIV drugs presents a unique challenge and opportunity for drug safety as well as public health system.

Methods: This presentation will present the current ADRs associated with the Anti-HIV drug use, as well as the ADRs reporting system and data received for ADRs.

Results: The current reporting system as well as changing regulatory landscape for PV in China after the country joint the ICH will also be discussed. The presentation will also be highlighted with a case study to build the system for the new HIV drug’s AE collecting and reporting.

Conclusion: Little is known about the toxicity profile of ARV drugs in China. The relevant comorbid conditions such as tuberculosis and hepatitis also have complicated the safety profile of ARVs. A strengthened PV system and well-designed risk management plan is necessary for ARVs especially for the newly approved ones.

Disclosure of Interest: None declared.

23 O-022 Clinical Application and Regulation on Oral Diabetic Drugs in China: Current Status and Further Perspective

23.1 X. H. Yang¹, J. K. Huang^*1, L. Zhang¹

23.1.1 ¹Dongfang Hospital Affiliated to Beijing University of Chinese Medicine, Beijing, China

Background/Introduction: Chinese medical and healthcare service is characterized by the coexistence of traditional Chinese medicine (TCM)and Western Medicine (WM). Integrated therapy of Chinese and Western Medicine (ITCWM) can not only improve the symptoms of diabetes, prevent and treat the complications of diabetes effectively, but also enhance the quality of life, moreover it reduces medical expenses. Recent clinical and experimental studies have shown that has obvious advantages in the treatment of diabetes.

Objective/Aim: To summarize the clinical application and risk control modes of oral diabetic drugs (ODDs) in China.

Methods: This authors summarize the status of ODDs registration and management in China, reviewing the participatory ways of physicians, scientific community, enterprises and regulators in the whole life cycle risk management based on the risk characteristics of various Chinese ODDs.

Results: The classification of Chinese authorized ODDs are divided into chemical drugs (CDs), Pure TCM drugs (PTCMDs) and Chinese and Western Compound drugs (CWCDs). CDs have advantage in clearing hypoglycemia, while PTCMDs with fewer side effects in improving complications, symptoms and reducing the side effects of CDs. The safety of CDs and CWCDs have been of great concern lately. The roles concerning risk management have been involved in the risk management by formulating the guidelines on the diabetic therapy and rational application of ODDs, strengthening the publicity and training of safe medication, bringing safe medication course into medical college curriculum, controlling drug marketing approval strictly related to potential drug risks, expanding channels of drug safety information collection and communication between regulators, enterprises, healthcare professionals and the public. Moreover, exploring research modes concerning pharmacovigilance, strengthening marketed sampling and quality supervision and cracking down on counterfeit drugs and inferior drugs are also important [1–5].

Conclusion: ITCWM has complementary advantages in diabetic treatment. It is effective ways promoting rational and safe use of ODDs by establishing convenient patient ADR spontaneous reporting system, carrying out basic research on drug interaction, developing active surveillance large-scale evidence-based clinical studies actively.

References:

• 1. Center for Drug Evaluation, CFDA. Key issues for clinical trials of cardiovascular risk assessment of Type 2 diabetic new drugs. http://www.cde.org.cn/dzkw.do?method=largePage&id=311954 accessed 31 May 2018

• 2. Wang HY, Zhang J, Gao CY, et al. Risk control guidance for clinical studies of cardiovascular end point events in type 2 diabetic patients in China. http://www.cde.org.cn/dzkw.do?method=largePage&id=312453. Accessed 31 May 2018

• 3. Zhang L, Wong L, He Y, et al. Pharmacovigilance in China: current situation, successes and challenges. Drug Saf 2014;37:765–70

• 4. Zhang L, Yang XH. Evaluation and rational use of Chinese patent medicines for diabetes. China J Tradi Chinese Med Pharm 2009;24:1270–3

• 5. Zhang L, Yang XH. Adverse drug reactions and rational use of Xiaoke Wan. Chinese J Pharmacovigilance 2006;3:76–80

Disclosure of Interest: None declared.

24 O-023 Universal Health Coverage is Possible with Effective Pharmacovigilance at Current Expenditure Level in Nigeria

24.1 E. Okoro^*1, B. Oyejola², A. Giwa³

24.1.1 ¹Department of Medicine, ²Department of Statistics, University of Ilorin, Ilorin, Nigeria; ³Department of Jurisprudence & International Law, Delta State University, Abraka, Nigeria

Background/Introduction: Nigeria with > 190 million people will be the third most populous nation by 2050. And citizens each spend on average US$82 annually towards a total health expenditure of US$115 per capita when accessing services. This can rise to US $ 316–4000 for 20–40% of the adult population exhibiting hypertension ± type 2 diabetes. This pattern of personal health expenditure exceeds the financial threshold of 34–86 US$ per capita experts recommends for financing universal health coverage (UHC) including basic treatment of type 2 diabetes/hypertension. Even so, untimely deaths linked to both conditions remain alarming while this is plunging globally from effective interventions.

Objective/Aim: This symposium presentation attempts to highlight possible contributory factors why treatment quality is suboptimal and UHC remains inadequate in Nigeria.

Methods: Mainly academic reports of Nigerian origin after 1999 when National health insurance policy became operational locally were reviewed.

Results: Up to 60–70% of individuals in some cohorts of hypertension/type 2 diabetes each spend US $ 316–4000 annually [1–3] for  treatment whose quality is barely effective in < 20% to optimally diminish  the risk of adverse cardiovascular events. The practice as national treatment standard seem inconsistent with the best possible value-for-money based on dominant disease behavior in Nigeria population [1–5]. In addition, treatment costs can be inflated sometimes beyond internationally quoted prices. Instructively, one Netherlands based organization, operating community health insurance business in Nigeria, recently demonstrated effective antihypertensive therapy, often, the only cardiovascular intervention majority (> 90%) with type 2 diabetes/hypertension require to optimally diminish the risk of sudden death, stroke, etc. with just US$ 28 per person annually. Given what individuals currently pay without reimbursements, these observations reinforce EU recent declaration of Nigeria having enough internal revenue to finance its own development. Moreover, Nigeria’s UHC is 4–5% against > 40% achieved in Ghana with a comparative level of total health expenditure per capita and national health insurance scheme 5 years younger. And today, over 40% of youths in Northern Nigeria may have become addicted to opiates/other substances 13 years after dismantling measures that initially contained their over-prescription while liberalizing access [1, 6]. Incidentally, the epidemic of abuse is fed by a network of criminal gangs/cultists, benefiting indigenous pharmaceutical companies and some health care professionals.

Conclusion: Effective pharmacovigilance can mobilize/optimize Nigeria’s abundant internal revenue for UHC while also minimizing interventions becoming instruments of mass exploitation/destruction.

References:

• 1. Okoro EO. “Towards better prevention/control of hypertension and diabetes” 143rd Inaugural lecture, 27 February 2014, University of Ilorin, Nigeria see www.unilorin.edu.ng

• 2. Okonkwo IL, Ekpemiro JN, Okwor EU, Okpala PU, Adeyemo FO. Economic burden and catastrophic cost among people living with type2diabetes attending a tertiary health institution in South–East zone, Nigeria, BMC Res Notes 2015;8:527

• 3. Okoro C. Assessing the quality of care received by diabetes patients under Nigeria National Health insurance Scheme: does enrollment in health insurance matter? Boston University openBU; http://open.bu.edu

• 4. Chinenye S, Uloko AE, Anthonia O, OgberaE, OAFasanmade OA, et al. Profile of Nigerians with diabetes mellitus—Diabcare Nigeria study group (2008): Results of a multicenter study, India. Journal of Endocrinology 2012, 16, 558–564

• 5. Okoro EO, Oyejola BA. Aspirin and Diabetes Care in Nigeria : Treatment or Expliotation? J Clinic Res Bioeth 2015, 6:2. https://doi.org/10.4172/2155-9627.1000227

• 6. Kazeem Y. Coughing Up: The major ingredient in Nigeria’s codeine abuse crisis is corruption at major drug makers http://qz.com/1266823/block/bbc-investigates-nigeria-codeine-opiod-crisis-from-cough-syrups (visited last on 22 May 2018)

Disclosure of Interest: None declared.

25 O-024 Opportunities of Prospective Observational Entis Studies

25.1 C. Weber-Schoendorfer^*1

25.1.1 ¹Pharmakovigilanzzentrum Embryonaltoxikologie, Charité-Universitätsmedizin Berlin, Corporate Member of Freie Universität Berlin, Humboldt-Universität zu Berlin, and Berlin Institute of Health, Berlin, Germany

Background/Introduction: The main task of the European Network of Teratology Information Services (ENTIS) is to contribute to the prevention of birth defects and developmental disorders. Callers to the services—either health care professionals and/or pregnant patients—search for risk assessment regarding drug use during pregnancy.

Objective/Aim: This offers the unique possibility to combine counselling with the follow-up of critically exposed and unexposed pregnancies.

Methods: The majority of calls comes in during early pregnancy, thus neither pregnancy outcome nor the results of prenatal diagnostic tests are known. Using a standardized set of questions and variables across services, these prospectively followed up pregnancy courses are the base for observational multicenter cohort studies. By this design, all potential pregnancy outcomes are registered. The range of adverse outcomes can be analyzed, which is important for both, physicians and patients.

Results: TIS studies are cost-saving and are able to provide an equally recruited comparison cohort. Results can be controlled for a variety of confounders. Drug intake is documented in real-time. Through motivated and often thankful callers, the non-responder rate is low. By using a multicenter approach, it is possible to effectively accumulate a meaningful number of exposed pregnancies. Since the founding of ENTIS in 1990, a variety of multicenter studies has been published in peer-reviewed journals. Data quality and data harmonization across sites have been improved over the years as well as the overall study quality by following the STROBE guidelines [1]. However, the sample size of ENTIS studies is usually moderate; the follow-up period is typically limited to weeks after birth and TIS populations might not be representative for the pregnant populations.

Conclusion: Despite these disadvantages, the follow up of exposed pregnancy courses after consultation offers a unique opportunity of drug surveillance in order to contribute to the prevention of birth defects or to enhance data on drug safety.

References:

• 1. von Elm E, Altman DG, Egger M, Pocock SJ, Gøtzsche PC, Vandenbroucke JP. STROBE Initiative. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Lancet. 2007;370:1453–7.

Disclosure of Interest: None declared.

26 O-025 A Population-Based Cohort, Pregnant: Day-to-Day Experience with Medication use During Pregnancy and Breastfeeding

26.1 S. Vorstenbosch^*1

26.1.1 ¹Monitoring, Netherlands Pharmacovigilance Centre Lareb, ‘s-Hertogenbosch, The Netherlands

Background/Introduction: In general, safety issues related to medication use during pregnancy and breastfeeding are not addressed in pre-marketing studies. Despite the lack of information, medication use among pregnant women is frequent, with estimates ranging from 40 to 96% [1–5]. This widespread experience does not translate into knowledge that could be used in the pharmacotherapeutic care of pregnant and postpartum women.

In the Netherlands, the Dutch Pregnancy Drug Register pREGnant systemically documents relevant and high-quality information at a central place. Ultimately to obtain insight into medication use among pregnant and breastfeeding women and safety issues concerning maternal and fetal/infant health. In that way, it provides a step towards translation of daily practice with medication use in pregnancy and breastfeeding into actionable knowledge.

Objective/Aim: To discuss the design and strengths and limitations of the Dutch Pregnancy Drug Register pREGnant and its value in signal detection, epidemiologic studies and counseling of health care providers and patients.

Methods: The Dutch Pregnancy Drug Register pREGnant is an initiative of the Teratology Information Service of the Netherlands Pharmacovigilance Centre Lareb. The register has a prospective cohort design. The target population of pREGnant are pregnant women throughout the Netherlands (approximately 170,000 births/year). They are recruited in several ways, e.g., through a digital invitation of their health care professional around the first prenatal care visit, information leaflets or promotion on relevant websites. Data are primarily collected through web-based questionnaires completed by participating women (three times during pregnancy and three times during the infant’s first year of life). Topics in these questionnaires are current pregnancy, obstetric history, maternal lifestyle, health and medication use, delivery, and infant health. If permission is given, the self-reported data can be complemented with information retrieved from Perined, the perinatal registry of the Netherlands, obstetric and medical records, and/or pharmacy records.

Results: The unique aspects of the Dutch Pregnancy Drug Register pREGnant will be presented and discussed in light of its value in increasing knowledge on the safety aspects related to medication use during pregnancy and breastfeeding.

Conclusion: The Dutch Pregnancy Drug Register pREGnant systematically documents high-quality data on exposure, outcome and co-variables. This is a prerequisite to obtain more insight in medication use among pregnant and breastfeeding women and the potential effects on maternal and fetal/infant health.

References:

• 1. Bakker MK, Jentink J, Vroom F, van den Berg PB, de Walle HEK, de Jong-van den Berg. Drug prescription patterns before, during and after pregnancy for chronic, occasional and pregnancy-related drugs in the Netherlands. Bjog 2006;113:559–68

• 2. Cleary BJ, Butt H, Strawbridge JD, Gallagher PJ, Fahey T, Murphy DJ. Medication use in early pregnancy-prevalence and determinants of use in a prospective cohort of women. Pharmacoepidemiol Drug Saf 2010;19:408–17

• 3. Crespin S, Bourrel R, Hurault-Delarue C, Lapeyre-Mestre M, Montastruc JL, mase-Michel C. Drug prescribing before and during pregnancy in South West France: a retrospective study. Drug Saf 2011;34:595–604

• 4. Irvine L, Flynn RWV, Libby G, Crombie IK, Evans JMM. Drugs dispensed in primary care during pregnancy: a record-linkage analysis in tayside, Scotland. Drug Saf 2010;33:593–604

• 5. Mitchell AA, Gilboa SM, Werler MM, Kelley KE, Louik C, Hernandez-Diaz S. Medication use during pregnancy, with particular focus on prescription drugs: 1976–2008. Am J Obstet Gynecol 2011;205:51.e1–51.e8.

Disclosure of Interest: None declared.

27 O-026 Experience with the Who Central Registry for the Epidemiological Surveillance of Drug Safety in Pregnancy

27.1 C. Halleux^*1

27.1.1 ¹Special Programme for Research and Training in Tropical Diseases, World Health Organization, Geneva, Switzerland

Background/Introduction: Healthcare providers, pregnant women and policy-makers need valid information in order to make informed decisions about the use of drugs during pregnancy. Unfortunately evidence of drug safety is usually very limited and hence there is a clear need for more data. This is particularly valid in context when important questions remain on the safety of drugs during pregnancy or specific periods of pregnancy for drugs used by women of childbearing age and/or intentionally used in pregnancy, including for example anti-retroviral or anti-malarials where treatment cannot be postponed or suspended.

Pregnancy exposure registries and birth outcome surveillances systems are the most common approaches used to collect post marketing drug safety data in pregnancy.

While evidence on safety of drug in pregnancy is limited, options offer by data sharing has received considerable attention in recent years with legal, ethical and practical recommendations on how industry, researchers, regulatory bodies, funders and sponsors should facilitate and practice data sharing. Pooled data provides advantages of a larger database, greater statistical power and more robust evidence upon which decision-makers can make policies and recommendations.

Objective/Aim: Based on the existing need for broader evidence on the drug safety in pregnancy, the Special Programme for Research and Training in Tropical Diseases and other departments at the World Health Organization (WHO) have worked to develop a central pregnancy registry. The central registry provides a platform for pooling data from different pregnancy exposure registries and birth outcome surveillance projects that have collected safety data at country level. It built on existing pregnancy exposure registries and the objective is to provide a single harmonised platform into which contributors (sites/countries/projects) can export anonymized individual case data collected.

Methods: The standardised central registry will hold information on pregnancies, drug exposure, births, maternal outcome and congenital anomalies.

Though not limited to Africa or developing countries, the central database is being developed with a special focus on Africa. This responds to the critical need for active toxicity surveillance in countries where HIV and infectious disease burden is high and better understanding the safety of treatment exposure is substantial.

Results: Since it initiation the central pregnancy registry has grown but the challenges of data sharing still constitutes certain limitation.

Conclusion: Safety data following exposure to drugs during pregnancy are scarce. The WHO central registry for the epidemiological surveillance of drug safety in pregnancy represents an opportunity to optimise use of data.

Further efforts are still needed however to strengthen surveillance of drug safety in pregnant women and encourage data sharing.

Disclosure of Interest: None declared.

28 O-027 Back to the Future: The Case Narrative and Artificial Intelligence

28.1 R. Chandler^*1, N. Norén¹

28.1.1 ¹Research, Uppsala Monitoring Centre, Uppsala, Sweden

Background/Introduction: In his commentary “Why Doctors Need Stories” published in the NY Times, psychiatrist Peter Kramer states “…while we wait for conclusive science, stories will preserve diversity in our theories of mind” [1]. Signal detection in pharmacovigilance is a hypothesis-generating exercise, a clinical science using spontaneous reports of adverse events (AE) and other information sources to assess the probability of a causal relationship between a medicinal product and an event. It is logical therefore that clinical story contained with AE reports in the form of the case narrative is integral to the development of hypotheses of drug safety concerns. And yet, preservation of central role of the case narrative in pharmacovigilance is potentially being challenged with the application of automated case processing to handle increasing levels of spontaneous reporting and the increased requirements to ensure the protection of individual patient data.

Objective/Aim: The aims of this symposium are to underscore the importance of the narrative in the practice of medicine, and specifically within the practice of pharmacovigilance and pharmacoepidemiology, and to become familiarised with how AI (novel machine learning/data science methods) can help make best possible use of the case narratives.

Results: The first speaker, Eugene van Puijenbroek, will discuss the role of clinical reasoning and heuristics in the diagnostic process of clinical medical practice and will highlight their importance in causality assessments central to pharmacovigilance. Within the current infrastructure of the practice of pharmacovigilance, a reliance on disproportionality analysis upon structured data in the form of drug—ADR pairs limits the ability of a safety assessor to best understand the clinical reasoning of the reporter. Review of the case narrative is integral to capturing the nuances of diagnostic considerations and to achieving higher quality assessments of individual case safety reports.

The second speaker, Lucie Gattepaille, will introduce machine learning and its use in pharmacovigilance for the purpose of narrative mining. A number of machine learning methods to capture valuable information from large-scale screening of case narratives will be described. Such methods may enable us to automatically extract information important for clinical assessment and may lead to insights that the sole use of structured fields would leave hidden. These innovations can also be harnessed to ensure patient confidentiality.

Conclusion: Current and future advances in data science methodology can bring us back to basics of pharmacovigilance: “listening” to a story of a patient’s experience and making better clinical decisions regarding medicines safety.

References:

• 1. https://opinionator.blogs.nytimes.com/2014/10/18/why-doctors-need-stories/

Disclosure of Interest: None declared.

29 O-028 Information on Clinical Reasoning and Heuristics, a Missing Link in Pharmacovigilance

29.1 E. van Puijenbroek^*1,2

29.1.1 ¹Pharmacovigilance Centre Lareb, ‘s-Hertogenbosch, the Netherlands; ²Groningen Research Institute of Pharmacy, Pharmacotherapy, -Epidemiology & -Economics, University of Groningen, Groningen, the Netherlands

Background/Introduction: The analysis of Individual Case Safety Reports (ICSRs) mainly relies on the detailed clinical and pharmacological information provided by the reporting healthcare professional or patient. This information is usually aimed at the description of the suspected adverse drug reaction. However, the considerations of the reporter for an adverse drug reaction (ADR) to be present, is rarely mentioned on reporting forms. These considerations can be either based on clinical reasoning, be heuristic in nature, or a combination of both. For the assessors of ICSRs, knowledge of these considerations could provide insight in the clinical decision making process and may help to accept or refute alternative diagnoses. Therefore, in the analysis of ICSRs, insight in this process may enable us to distinguish meaningful from less meaningful observations.

Objective/Aim: In this presentation the role of clinical reasoning and heuristics in the diagnostic process will be discussed, with special attention for current developments in pharmacovigilance.

Methods: Review of literature on clinical reasoning and heuristics in clinical practice.

Results: Clinical reasoning is defined as the sum of thinking and the decision making processes associated with clinical practice. Multiple elements of the clinical situation are used to analyse the occurrence of clinical signs. Clinical reasoning has a strict structure but may vary in the way the case is interpreted depending on the experience of the clinician [1]. Heuristics on the other hand are simple, efficient rules usually focusing on one aspect of a complex problem and ignoring others. They can be based on intuitive judgments but can also be used as deliberate mental strategies when working from limited information [2]. Both processes play an important role in the diagnostic process. During the professional development there is a gradual change from structured formal clinical reasoning towards heuristic reasoning [1]. Whereas in the diagnostic process novices strongly rely on clinical reasoning, clinical experts rely more on previous experience and exploratory problem-solving heuristic techniques.

Conclusion: When assessing ICSRs, knowledge of the considerations to report may help us to substantiate causal reasoning. Although elements of clinical reasoning can be found in the discharge letters, elements describing the heuristic processes usually are not explicitly mentioned and may require active follow up. The increasing number of reports that has to be analysed on a national and global level, as well as the introduction of other sources of reporting like patient reporting, have implications for the way this type of information can be used. In disproportionality analysis, applied when analysing large numbers of ICSRs, additional analysis of case narratives and retrieving information about the reason to report may be needed. In clinical complex cases reported by patients, pharmacovigilance centres may have to track down details of the clinical assessment by the healthcare professionals involved. A better understanding of the reasoning behind reporting links the thought process of the reporting healthcare professional with data on the ICSR and may therefore improve the quality of the case assessment.

References:

• 1. Brush JE. How expert clinicians intuitively recognise a medical diagnosis Am J Med 2017;(130):629–34

• 2. Gigerenzer G, et al. Heuristic decision making Annu Rev Psychol. 2011;62:451–82

Disclosure of Interest: None declared.

30 O-029 Enhancing Pharmacovigilance in Sub-Saharan Africa: Sharing Experiences from a GSK Pilot Initiative in Malawi

30.1 V. Jusot^*1, O. Menang², F. Chimimba³, N. Dzabala³, C. Sambakunsi⁴, C. Chulu⁴, J.-U. Stegmann¹, Y. Guerra Mendoza¹

30.1.1 ¹GSK, Wavre, Belgium; ²PATH, Geneva, Switzerland; ³College of Medicine, University of Malawi; Blantyre, ⁴Pharmacy, Medicines and Poisons Board, Lilongwe, Malawi

Background/Introduction: Historically, the safety profile for most medicines in low- and middle-income countries (LMICs) have been established using spontaneous safety data from high income countries with established pharmacovigilance (PV) systems [1]. However, their safety profile may differ markedly due to a variety of factors including environmental and genetic influences. With more medicines being developed exclusively for use in sub-Saharan Africa (SSA), a region with suboptimal PV systems, a strengthened safety surveillance system is critical for reliable adverse events (AE) reporting and requires effective collaboration between the key players in the field of PV. Currently, there are numerous initiatives aimed at strengthening PV capacity in developing countries. Established collaborative partnerships offer the benefit of pooled technical expertise and resources to ensure synergy and sustainability.

Objective/Aim: Facilitate spontaneous reporting of medicine-related AEs in SSA through capacity building.

Methods: Countries with a basic or minimal PV system were selected. Malawi was the first country to implement the project, in collaboration with the Ministry of Health (MOH), regulatory authority, college of Medicine and PATH, through PV trainings and mentoring of health care professionals (HCPs) in their facilities. Pre-selected key performance indicators were evaluated.

Results: Malawi implementation began in October 2016. PV trainings and regular mentoring has been provided to 34 nationwide focal points and over 300 HCPs sensitized at 37 healthcare facilities. Reported AEs significantly increased from 4 before November 2016 to 204 in May 2018, qualifying Malawi to become a full member of the World Health Organization’s Programme for International Drug Monitoring [2]. The quality and timeliness of AE reports require further improvement.

Conclusion: Intensified PV training and mentoring has shown to be effective in the improvement of passive safety surveillance. Partnerships with country MOH and supranational stakeholders are a key pre-requisite for engagement and country-level ownership. Lessons learned from this experience will assist in seeking further partnerships into other PV-related initiatives with additional LMICs.

Funding source: GlaxoSmithKline Biologicals SA.

References

• 1. Amarasinghe A, Black S, Bonhoeffer J et al. Effective vaccine safety systems in all countries: a challenge for more equitable access to immunization. Vaccine 2013; 31: B108–14

• 2. Join the WHO Programme for International Drug Monitoring. Updated in October 2010. Retrieved May 17, 2018, from https://www.who-umc.org/global-pharmacovigilance/who-programme/join-the-who-programme/

Disclosure of Interest: V. Jusot Shareholder of: GSK, Employee of: GSK, O. Menang: None Declared, F. Chimimba: None Declared, N. Dzabala: None Declared, C. Sambakunsi Grant/Research support from: Reports a grant from GSK in the frame of this work, C. Chulu: None Declared, J.-U. Stegmann Shareholder of: GSK, Employee of: GSK, Y. Guerra Mendoza Shareholder of: GSK, Employee of: GSK.

31 O-030 Strengthening Pharmacovigilance Capacity for Vaccine Manufacturers in Low-And-Middle-Income Countries (Lmics)

31.1 O. Menang^*1, V. Sharma², N. Bhat³, M. Socquet¹, F. Berlanda-Scorza⁴, J. Flores⁴

31.1.1 ¹PATH, Geneva, Switzerland; ²PATH, New Delhi, India; ³PATH, Seattle, United States; ⁴PATH, Washington D.C., United States

Background/Introduction: It is estimated that in the next 3–5 years, about 43 vaccines, 30 drugs and 14 diagnostic products targeted for LMICs will be developed [1]. Many of these new vaccines and drugs target diseases unique to LMICs and will be marketed in these regions soon after licensure. As these products advance from the pre- to post-licensure stage, it is imperative that pharmacovigilance (PV) systems are established to ensure adequate safety surveillance. Manufacturers cannot adequately monitor the safety of their products without functional PV systems. Though developed countries mandate that manufacturers must establish PV systems [2], this is not yet the case in majority of LMICs. However, LMIC manufacturers, often with limited PV system functionality, are required to demonstrate capacity for passive and active safety surveillance when seeking WHO prequalification of their products. PATH provides technical and logistical assistance to vaccine manufacturers in LMICs; PV support comprises establishing new or strengthening existing systems based on global and local regulatory requirements.

Objective/Aim: Establish or strengthen pharmacovigilance quality systems for vaccine manufacturers to ensure adequate safety surveillance of authorized products and appropriate benefit/risk evaluation.

Methods: Four-step procedure consisting of: (1) Questionnaire-based assessment and gap analysis of existing PV system using a PV system assessment tool (2) Develop business-friendly PV strengthening plan considering regulatory requirements and industry best practices (3) Implement identified interventions including PV training, standard operating procedures, IT database (4) Continuous monitoring, improvement and communication with NRAs for alignment.

Results: Sustainable functional PV quality systems established for vaccine manufacturers, with structures, processes and outcomes, for continuous risk–benefit evaluation in compliance with local and international requirements.

Conclusion: Going forward, novel vaccines, particularly those specifically designed for use in LMICs, may not be able to rely on safety and risk–benefit assessments conducted in developed countries. Therefore, adequate safety surveillance and evaluation by both manufacturers and NRAs in LMICs is critical at product launch. Such entities may require support in order to achieve the capacity to meet the requirements for effective and cost-efficient pharmacovigilance.

References:

• 1. Lalvani P, Milstein J. Access to new health products in low income countries and the challenge of pharmacovigilance. Retrieved May 28, 2018, from: http://pdpaccess.org/downloads/projects/full-papers/Pharmacovigilance%20Discussion%20Paper.pdf

• 2. European Medicines Agency. Guideline on good pharmacovigilance practices (GVP) Module I—Pharmacovigilance systems and their quality systems: EMA/541760/2011; 2012

Disclosure of Interest: O. Menang Employee of: PATH, V. Sharma Employee of: PATH, N. Bhat Employee of: PATH, M. Socquet Employee of: PATH, F. Berlanda-Scorza Employee of: PATH, J. Flores Employee of: PATH.

32 O-031 Development of Pharmacovigilance System in a Resource-Limited Country, the Experience of the Democratic Republic of Congo

32.1 D. Nzolo Bomene^*1, Y. Lula¹, N. Ntamabyaliro¹, A. Engo¹, G. Tona¹

32.1.1 ¹Unit of Clinical Pharmacology and Pharmacovigilance, University of Kinshasa, Kinshasa, Democratic Republic of the Congo

Background/Introduction: Implementation of a pharmacovigilance system in resource-limited countries has always been a challenge. Despite all the effort, at global and regional level, Africa is still contributing with less than 1% to the World Health Organization’s (WHO) global Individual Case Safety Report (ICSR) database (VigiBase). The Democratic Republic of Congo (DRC) has one of the most active pharmacovigilance centres in sub-Saharan Africa and is currently used as a reference for the development of pharmacovigilance system in the Central of Africa.

Objective/Aim: To describe the experience of DRC to set up a pharmacovigilance system in a resource-limited country in sub-Saharan Africa.

Methods: An analysis of the DRC pharmacovigilance system since launch were performed. Best practices, bottlenecks and main challenges has been identified. Additionaly, VigiLyze (Full dataset, data lock point 03/06/2018) has been consulted to document the contribution of DRC to the WHO global Individual ICSR database.

Results: As best practices, the national pharmacovigilance centre in DRC has been placed at the Unit of Clinical Pharmacology and Pharmacovigilance, at the University of Kinshasa. It employs academic staffs performing research in Pharmacovigilance. It has included pharmacovigilance in the student curriculum. It interacts with the national health system and provides training and supervision at central, intermediate and operational level. It interacts with public health programs like the national immunization program and those using drugs against malaria, tuberculosis, neglected diseases, etc. It uses different strategies to capture ICSRs and consisting in spontaneous reporting, stimulated reported and active pharmacovigilance, including community-based and hospital-based studies. It works in close collaboration with the national regulatory authority about drug safety. The main challenges included the paucity of financial support, the big size of the country with limited accessibility and the sustainability of pharmacovigilance activities in different area at the end of funding projects.

The different strategies used by the DRC national pharmacovigilance centre since 2010 allowed DRC to contribute with 14,540 ICSRS in VigiBase. Half of these ICSRs come from community-based pharmacovigilance during mass vaccination campaigns against yellow fever (30.7%) and poliomyelitis (13.8%).

Conclusion: The experience of the Democratic Republic of Congo shows that it is possible to develop a pharmacovigilance system in a resource-limited country. While lack of funding is one of the challenges, use of universities, collaboration with public health programs and community-based pharmacovigilance constituted different solution to boost pharmacovigilance system.

Disclosure of Interest: None declared.

33 O-032 Drug Utilization and Drug–Drug Interaction Studies in the Field: Examples of South–North Collaboration Projects

33.1 A. Kuemmerle^*1,2, S. Akakpo³, J. Liwono^3,4, S. Duparc³, C. Burri^1,2, A. Stergachis⁵

33.1.1 ¹Swiss Tropical and Public Health Institute, Basel, Switzerland; ²University of Basel, Basel, Switzerland; ³Medicines for Malaria Venture, Geneva, Switzerland; ⁴Clinical Pharmacology and Pharmacovigilance Unit, Kinshasa, The Democratic Republic of the Congo, ⁵School of Pharmacy and School of Public Health, University of Washington, Seattle, United States

Background/Introduction: In low-resource countries, increased urbanization is contributing to a dual burden of infectious diseases and non-communicable diseases. Multiple comorbidities and the corresponding concomitant use of multiple medicines increase the risk of drug–drug interactions (DDI). For example, co-infection with malaria and HIV results in concomitant administration of antimalarial and antiretroviral medicines. Medicines within these two classes of drugs have high DDI risks when used simultaneously^(1,2). Additionally, in spite of their common use, it is not broadly recognized that herbal medicines can alter the efficacy and safety of co-administered prescription drugs⁽³⁾. In sub-Saharan Africa, traditional medicine is widely used in rural and urban areas⁽⁴⁾. As a consequence, the overlap between conventional medicines and herbal medicines is real among countries in health system transition and may lead to drug-herbal interactions (DHI)⁽⁵⁾.

Objective/Aim: The first objective is to present Swiss Tropical and Public Health Institute’s (SwissTPH) recent and current research in this field conducted with partners from the South and North. Medicine for Malaria Venture’s (MMV) and SwissTPH will also share their point of view from a Product Development Partnership and Research Institution perspective. As a second objective the speakers will present the protocol of their recently planned study. Its aim will be to assess the potential DDIs and DHIs occurring in patients with uncomplicated malaria receiving antimalarial treatment.

Methods: We plan to perform a cross-sectional study in urban Kinshasa, Democratic Republic Congo. The study will be carried out at sentinel health centers and within selected community settings. Each participant and prescriber, after having signed an informed consent form, will be requested to complete a questionnaire designed to collect information on antimalarial treatments, comorbidities, other medications, herbals and socio-economic factors. Drug pairs and drug–herbals pairs will be assessed by at least two different pharmacologists using drug interaction detection software or manually if necessary for drug/herbal interactions. Drug-pairs will be subsequently categorized as deleterious, of potential clinical relevance, of weak clinical significance or not interacting.

Results: Results will be reported on the prevalence of potential DDIs and DHIs among patients as well as factors associated with them. Findings from recent studies, and the expected distribution of new antimalarial drugs in Sub-Saharan Africa, were of MMV and SwissTPH interest since it could generate evidence on real-world drug utilization and the risk of deleterious DDIs.

Conclusion: The long-term goal of our projects is to inform the development of original and locally tailored interventions to create awareness about and mitigate the risk of polymedication in patients receiving antimalarial treatment.

References

• 1. University of Liverpool. HIV Drug Interactions Tables 2018 [cited 2018 01 March]. Available from: https://www.hiv-druginteractions.org/

• 2. Seden K, Gibbons S, Marzolini C, Schapiro JM, Burger DM, Back DJ, et al. Development of an evidence evaluation and synthesis system for drug–drug interactions, and its application to a systematic review of HIV and malaria co-infection. PLoS One. 2017;12:e0173509

• 3. Wheaton AG, Blanck HM, Gizlice Z, Reyes M. Medicinal herb use in a population-based survey of adults: prevalence and frequency of use, reasons for use, and use among their children. Ann Epidemiol. 2005;15:678–85

• 4. Boukandou Mounanga M, Mewono L, Aboughe Angone S. Toxicity studies of medicinal plants used in sub-Saharan Africa. J Ethnopharmacol. 2015;174:618–27

• 5. Thomford NE, Dzobo K, Chopera D, Wonkam A, Skelton M, Blackhurst D, et al. Pharmacogenomics implications of using herbal medicinal plants on African populations in health transition. Pharmaceuticals (Basel) 2015;8:637–63.

Disclosure of Interest: None declared.

34 O-033 Distilling the Value of Narratives with Machine Learning

34.1 L. Gattepaille^*1

34.1.1 ¹Uppsala Monitoring Centre, Uppsala, Sweden

Background/Introduction: With the steady increase in computing resources and the unparalleled amounts of data available, machine learning is a promising technology that has already offered significant improvement in areas such as computer vision and speech recognition. An interesting application of machine learning for pharmacovigilance is narrative mining. Case narratives often contain useful, detailed and nuanced information that might complement what can be found in the structured fields of case reports. Machine learning methods can help us automatically extract and harness this information at scale and may lead to insights that the sole use of structured fields would leave hidden.

Objective/Aim: The Uppsala Monitoring Centre is a firm believer in the unique value of free-text narrative accounts of suspected adverse drug reactions and seeks to develop machine learning methods to distil valuable information from large-scale screening of case narratives.

Methods: For example, we have sought to automatically identify the severity of an adverse event from the narrative, an aspect that is not currently present in the structured description of case reports. Furthermore, we have developed named entity recognition tools that can be used recognize ADRs and their associated drugs, as well as distinguish them from indication mentions. In order to enable signal detection and assessment at the global scale, we should seek to make possible the sharing of case narratives across organisations without compromising patient confidentiality.

Results: To this end, we are developing machine learning techniques to automatically de-identify case narratives allow patient privacy to be protected while keeping the crucial clinical information intact for review. 

Conclusion: Finally, I will briefly present potential areas of interest in narrative mining, illustrating that narratives are a rich resource that provides great potential for a better characterization of case reports, and that efforts in the field should push the science of pharmacovigilance forward.

Disclosure of Interest: None declared.

35 O-034 Similar Sounding Names in Medications Safety: Risky or Not?

35.1 Y. Jani^*1,2, D. Gerrett³

35.1.1 ¹Centre for Medicines Optimisation Research & Education, UCLH NHS Foundation Trust, London, United Kingdom; ²Research Department of Practice & Policy, UCL School of Pharmacy, London, United Kingdom; ³NHS Improvement, London, United Kingdom

Background/Introduction: Medication errors and related harm are an established global patient safety concern. Look-alike sound-alike medicine names and products are implicated as common caustive and contributory factors. However there is little research on the patient safety benefits of International Non-proprietary Name (INN) common stems: a part of the name that expresses the pharmacologically-related group to which the substance belongs.

Objective/Aim: The aim was to determine if there was an association between medication suffixes and pharmacological function. The objectives were to ascertain any differences in association amongst doctors, nurses and pharmacists; and to identify if the association was linked to experience of the professional.

Methods: A combination of actual medicines and fictitious names with suffix extensions based on common drug names where there was a known link between extension and pharmacological function were presented as sentences. Confounder sentences were designed such that there was no known suffix extension and no relationship with a pharmacological function. 88 sentences were developed of which 44 were presented to the participants at random through the survey software.

A convenience sample of doctors, nurses and pharmacists was recruited and invited to complete an online survey to indicate whether the name (actual or fictitious) matched the indication in the sentence presented. 

Results: 10 doctors, 36 nurses (of which 6 = students), 14 pharmacists (of which 5 = pre-registration) and 12 pharmacy staff completed the survey.

Nurses were the least likely and pharmacists and pharmacy staff most likely to correctly match suffixes to the pharmacological function, as shown by an ascending polynomial relationship from the nursing profession through medicine, to pharmacy. There was no difference in correcly matching medicine name to function by experience. ‘cillin’ and ‘mycin’ antimicrobials were the most identifiable drug classes across all professions.

Conclusion: This study supports an association between medicine name suffixes and pharmacological function (presented as suggested indications). The use of International Non-proprietary Names may have a positive safety impact in recognition of critical medicines such as antimicrobials.

Disclosure of Interest: None declared.

36 O-035 Patient Involvement in Adverse Drug Reaction Detection and Reporting

36.1 R. Karwa^*1,2, M. Maina³, B. Jakait^3,4, A. Gardner^4,5, C. Ngetich^3,6

36.1.1 ¹Pharmacy Practice, Purdue University, West Lafayette, United States; ²Pharmacology, Moi University College of Health Sciences, Eldoret, Kenya; ³Pharmacy, Moi Teaching and Referral Hospital, Eldoret, Kenya; ⁴Executive, Academic Model Providing Access to Healthcare, Eldoret, Kenya; ⁵Infectious Diseases, Indiana University, Indianapolis, United States; ⁶Pharmacy, Academic Model Providing Access to Healthcare, Eldoret, Kenya

Background/Introduction: Approximately two-thirds of people living with HIV globally reside in Sub-Saharan Africa. In 2015, 12 million patients in SSA were on antiretroviral therapy (ART) [1]. ART-related drug toxicity is a growing concern in low and middle-income countries, where under-developed pharmacovigilance (PV) infrastructure is common, as it represents a major contributor to treatment interruptions and poor adherence [2, 3]. HIV peer health workers, trained HIV-positive patients have previously been shown to improve outpatient clinic follow up and to increase adherence to ART [4−6]. To address the gap in PV, we set up a peer health worker program.

Objective/Aim: To describe the implementation and impact of an HIV peer health worker program targeting adherence, ADR reporting and medication safety issues in western Kenya.

Methods: In 2012, we conducted a study evaluating ADR reporting approaches, including an arm where HIV peer health workers administered surveys, in an HIV-positive population in western Kenya. The study site, the Academic Model Providing Access to Healthcare (AMPATH), is a partnership with Moi University, Moi Teaching and Referral Hospital (MTRH), the Ministry of Health and a number of North American universities, that provides comprehensive services in the public health sector for a rural, underserved catchment population of over 4 million people. Patient interactions with peers revealed not only ADRs experienced but also complex concerns, adherence issues and an eagerness among patients to converse with someone who held a shared-disease experience. To capitalize on this finding, we held stakeholder meetings with local providers to understand how to transition from a study to a clinical program. Through additional grant funding, we retained 4 peers to set up the program.

Results: The transition from a study to an HIV peer health worker program was successful in creating both inpatient and outpatient programs. Two HIV peer health workers were deployed to the inpatient setting at MTRH, where they provide adherence and disclosure counseling, linkage to outpatient care, medication refills, ADR documentation and post-discharge follow up for approximately 900 patients per year. In an outpatient pharmacy setting, 2 peers provide medication counseling and ADR assessments through group presentations and individual sessions.  Additionally, a novel HIV resistance clinic, which provides care to patients who are failing 2nd line protease inhibitor ART was established with peers playing a pivotal role. Peers interact with the patients first, generating a relationship and an accurate medical history, including ADR assessments prior to provider–patient interactions. Over 200 patients have been enrolled into the resistance clinic with a majority of the patients successfully being retained on second line ART. Peers are currently being employed in the programmatic roll-out of dolutegravir to document ADRs experienced in a western Kenyan population.

Conclusion: The experiences at AMPATH indicate that trained patients can play crucial roles in promoting adherence, enhancing medication safety and improving ADR detection and reporting.

References:

• 1. WHO. HIV/AIDs. 2017  [cited 2018 24/05/]; Available from: http://www.afro.who.int/health-topics/hivaids

• 2. Messou, E., et al., Antiretroviral treatment changes in adults from Cote d’Ivoire: the roles of tuberculosis and pregnancy. Aids 2010; 24: p. 93–9

• 3. Braitstein, P., et al., Sustainability of first-line antiretroviral regimens: findings from a large HIV treatment program in western Kenya. J Acquir Immune Defic Syndr 2010; 53: p. 254–9

• 4. Enriquez, M., et al., A Peer-Led HIV Medication Adherence Intervention Targeting Adults Linked to Medical Care but without a Suppressed Viral Load. J Int Assoc Provid AIDS Care 2015; 14: p. 441–8

• 5. Hatcher, A.M., et al., Predictors of linkage to care following community-based HIV counseling and testing in rural Kenya. AIDS Behav 2012; 16: p. 1295–307

• 6. Richter, L., et al., Pregnant women living with HIV (WLH) supported at clinics by peer WLH: a cluster randomized controlled trial. AIDS Behav 2014; 18: p. 706–15

Disclosure of Interest: None declared.

37 O-036 A Delphi Study on the Development of a risk Proportionality Framework for the Selection of Risk Minimisation Interventions in Asia

37.1 Y. Moride^*1,2, S. Frise^3,4, H. Le Louët⁵, G. Castillon², J.-C. Delumeau⁶

37.1.1 ¹Université de Montreal, Montreal, Canada; ²YOLARX Consultants, Montreal, Canada; ³AstraZeneca, Mississauga, ⁴Dalla Lana School of Public Health, University of Toronto, Toronto, Canada; ⁵Department of Pharmacology, Université Paris-Est, Créteil, France; ⁶Bayer, South East Asia, Singapore

Background/Introduction: Existing risk minimisation interventions (RMIs), frequently used in the European Union and the United States, may not be applicable to Asian countries.

Objective/Aim: This study aims at developing explicit criteria for the evaluation of risk proportionality applicable to RMIs in Asia.

Methods: Project consists of two phases: A Consultation Phase (pre-Delphi) and A Modified Delphi Study. Consultation phase involved a purposeful sample of 8 experts from regions of the world with the highest experience in therapeutic risk management (EU, US, Australia, Canada, Singapore, Taiwan, Japan). In-depth individual interviews were conducted in order to identify key factors (criteria) and circumstances used in the assessment of risk proportionality and their impact on RMIs (low, medium, high stringency interventions). Using those factors as a starting point, a Delphi panel is used to reach consensus on a structured framework (e.g., a set of explicit criteria and associated weights (scoring system) to assess risk proportionality). Pharmacovigilance professionals representing all target groups (regulatory, academia, industry), distributed across countries belonging to the Association of South-East Asian Nations (ASEAN), as well as East-Asian and South-Asian countries, were contacted.

Results: In the Consultation Phase, the following factors were determinants of whether a RMI is needed: Frequency and seriousness of the AE, benefit–risk ratio, severity of the indication, suddenness and unpredictability of the AE, inability of routine medical practice to minimize the risk, teratogenicity, specific AEs (hepatotoxicity, agranulocytosis, SJS, etc.), genetic profile, extent of drug use in the population, risk of off-label use, drug confusion, novelty of the actions that must be taken to treat or manage the risk. The main country-specific factor influencing the selection of the RMI is the local treatment practice (prescription and follow-up by specialists versus general practitioners), drug coverage, as well as access to specialists or procedures for monitoring (e.g., ECG or biomarkers). These factors were used to develop the questionnaire for the Delphi panel. So far, there are 16 participants out of a total of 20–25. Each respondent has to rank the importance of each factor in determining whether a RMI is justified, from 1 (not-important) to 10 (extremely important).

Conclusion: Themes and determinants of risk proportionality were identified in the Pre-Delphi consultation. Panel responses after two Delphi rounds will be synthesized and presented.

Disclosure of Interest: None declared.

38 O-037 Addressing the Challenges Affecting the Implementation of Routine Risk Minimisation and Risk Minimisation Interventions in Diverse Countries

38.1 J.-C. Delumeau^*1

38.1.1 ¹Pharmacovigilance, BAYER, Singapore, Singapore

Background/Introduction: This presentation is part of the Symposium: Risk Minimisation Interventions for diverse health care systems: Following-up on CIOMS IX. As described in CIOMS IX, the decision to implement a risk minimisation intervention and the selection of the method should include a careful analysis of the specificity and capability of the target health care system. A particular attention should be paid to the burden which may result from implementing the intended risk minimisation intervention and from the method used for measuring its effectiveness. It should be ensured that the intended measure is likely capable of providing the expected benefit rather than detrimentally restrict the access to the medicine.

Objective/Aim: The purpose of this presentation on behalf of the ISoP SiG Group on Risk Minimisation Methods for Asian Countries, is to compile and analyse the experience accumulated by the different stakeholders involved in risk minimisation activities in order to determine the lessons learned that can help selecting the methods and optimise their implementation in the contexts of diverse health care systems.

Methods: Challenges affecting the implementation of risk minimisation will be considered and analysed. Both routine risk minimisation methods and risk minimisation interventions will be addressed. The challenges related to the capability of the health care system or other country-specific aspects will be considered, especially to what extent specific risk minimisation interventions can effectively address the sub-optimal impact of routine risk minimisation in some countries to ensure the safe use of targeted products.

Results: From the analysis of the experience accumulated in the past 4 years, the lessons learned will be determined for conventional educational and access restriction methods as well as for methods powered by Web and App communication technologies. The current landscape will be compared to the trends anticipated by CIOMS IX. Further improvements aimed to address the lessons learned will be proposed, and future perspective will be considered.

Conclusion: The presentation will conclude on how the outcome of the ISoP SIG on Risk Minimisation Methods for Asian countries can help the risk minimisation decision making process in consideration of the above lessons learned.

References:

• 1. Practical Approaches to Risk Minimisation for Medicinal Products. Report on CIOMS Working Group IX. Geneva, 2014.

Disclosure of Interest: J.-C. DELUMEAU Employee of: Bayer.

39 O-038 Listening to Kasparov: Putting Human Experts and Artificial Intelligence Together to Create Advanced Pharmacovigilance

39.1 S. Comfort^*1, M. Munoz²

39.1.1 ¹Genentech, San Francisco, United States; ²US Food and Drug Administration, Silver Spring, United States

Background/Introduction: World champion grandmaster chess master Garry Kasparov was defeated by IBM’s “Deep Blue” chess system in 1997, marking the first time in human history that a world champion was defeated by a computer.   Since then, AI has developed rapidly with AlphaGo beating the world No. 1 ranked human player Ke Jie in a 2017 three-game match of Go.  The growing successes of AI in many fields has led some to feel that it will take over everything and replace humans in many fields, including PV.

Objective/Aim: In this talk, the speakers will present perspectives inspired by Mr. Kasparov’s development of “Advanced Chess” to partner machines and humans to play advanced levels of chess.

Methods: Dr. Comfort will present highlights of Machine Learning efforts in PV as a result of internal and external collaborations (e.g., IBM, PDSS, PVIS, PDSI).

Results: His talk illustrates that beyond the headlines of disruption, partnering with AI will actually allow us to develop an “advanced” level of PV that combines the best aspects of human and machine intelligence. The results can be a level of product safety insights, quality, and risk management solutions for patients that are not achievable today.

Dr. Muñoz will highlight opportunities for AI to support PV practices from a regulator’s perspective. She will share findings from ongoing projects demonstrating the potential for advanced PV, such as those aimed at timely signal identification.

Conclusion:

Disclosure of Interest: None declared.

POSTER PRESENTATIONS

40 ISoP18-1021 Developing a New Methodology for Preparing a Guideline: The Case of Good Pharmacovigilance Outsourcing Practices

40.1 G. Furlan¹, B. Van Leeuwen ², M. Bertazzoli³, B. Edwards^*4

40.1.1 ¹Seqirus, Maidenhead, United Kingdom, ²Astellas, Amsterdam, Netherlands, ³ISOP, Lugano, Switzerland, ⁴ACRES, Leatherhead, United Kingdom

Background/Introduction: Structured workflows, organisation and practices can hinder innovation instead of promoting it [1]. In addition, the legal and bureaucratic considerations that govern formal channels for voicing concerns may inhibit from speaking up [2] and proposing solutions. Legal requirements for innovative work practices take a long time to be implemented. Guidelines often are derived from these requirements at an even later stage. In pharmacovigilance, outsourcing could be an example: it has become more and more common practice attracting the authorities’ attention. The Medicines and Healthcare Products Regulatory Agency (MHRA) has recently published the outsourcing areas identified, during inspections, as problematic for Marketing Authorisation Holder and Service Providers [3]. Articles describing the principles of pharmacovigilance outsourcing have been published [4], but a consensus guideline describing all the steps, principles and points to consider for this topic is missing.

Objective/Aim: The authors, members of a volunteer group, are trying to fill this gap by working through a new way of preparing such a guideline.

Methods: A first draft, also referring to guidelines in other areas of industry [5], has been prepared and is undergoing subsequent cycles of review by increasingly wider number of pharmacovigilance experts, both from the industry and from the service providers. Following each cycle of review, the contributions from all stakeholders are evaluated and included in the guideline. The final aim is to have a non-prescriptive and voluntary standard that can be used by both the clients and vendors so to improve mutual understanding and likelihood of success.

Results: The guideline describes how, at first, all clients (the party outsourcing the activity) should have at least an in-house pharmacovigilance system expert who understands regulatory requirements, workflows and risks associated with non-compliance so to mitigate them. This person should also be accustomed to manage and negotiate contracts with vendors in order to select the one that suits the needs of the client best and is empowered to take decision. It goes without saying that the client should have an overview of the outsourced tasks. Service providers, on their end, should be transparent and coherent regarding their experience, marketing positioning and overall strategy; they should always try to understand the exact clients’ needs, expectations and the support they can receive from the client in performing the outsourced tasks.

Conclusion: Setting clear, transparent and reasonable expectations is the basis for both the client and vendor to optimally design and describe their joint pharmacovigilance system.

References:

• 1. Steven S. Hoffman: Make elephants fly. The Process of Radical Innovation. Piatkus, 2017

• 2. Martin GP, et al. Making soft intelligence hard: a multi-site qualitative study of challenges relating to voice about safety concerns. BMJ Qual Saf 2018; https://doi.org/10.1136/bmjqs-2017-007579

• 3. Harper J. MHRA GPvP Inspectorate guide to marketing authorisation holder considerations for agreements with pharmacovigilance system service providers. 22 January 2018 MHRA Inspectorate Blog https://mhrainspectorate.blog.gov.uk/2018/01/22/mhra-gpvp-inspectorate-guide-to-marketing-authorisation-holder-considerations-for-agreements-with-pharmacovigilance-system-service-providers/

• 4. Furlan, G, van Leeuwen, B,Edwards, B. Pharm Med 2017;31:75

• 5. van Leeuwen BP, Prendergast C, Edwards B, Dawson B. Can pharmacovigilance learn from the oil and gas industry, which has been outsourcing for over a century? Ther Innov Regul Sci 2017;51:11–5

Disclosure of Interest: G. Furlan Employee of: pharmaceutical Industry, B. Van Leeuwen Employee of: pharmaceutical industry, M. Bertazzoli: None Declared, B. Edwards Employee of: pharma industry service provider.

41 ISoP18-1022 Fetal Effects of Serotonin-Noradrenaline Reuptake Inhibitor Use During Pregnancy

41.1 F. Martin^*1, J. L. Richardson^1,2, H. Dunstan², A. Greenall², S. Stephens², L. M. Yates^1,2,3, S. H. L. Thomas^1,2

41.1.1 ¹Institute of Cellular Medicine, Newcastle University, Newcastle-upon-Tyne, United Kingdom^{; 2}UKTIS, ³Northern Genetics Service, Newcastle-upon-Tyne, United Kingdom

Background/Introduction: The concern for women taking a medication during pregnancy is an important consideration when providing care and data is limited; previous studies have described increased risk of preterm delivery [1] for example. The aim of this study was to determine whether there is an increased risk of adverse pregnancy outcomes for women who use SNRI antidepressants during pregnancy.

Objective/Aim: To perform a prospective cohort study investigating the fetal effects of maternal Serotonin-Noradrenaline Reuptake Inhibitor (SNRI) use during pregnancy.

Methods: This study utilised a prospective comparative cohort design using data obtained by the UK Teratology Information Service (UKTIS) between 1995 and 2017. Inclusion criteria required pregnancies to be singleton, without exposure to known or suspected teratogens (excluding alcohol) and without maternal poisoning/overdose exposures. SNRI exposed pregnancies were matched 1:3 with pregnancies exposed to selective serotonin reuptake inhibitors (SSRI) and 1:5 with non-teratogen exposed (NTE) pregnancies by maternal age and year of enrolment (± 3 years). Malformations were categorised using EUROCAT specifications by two study authors blind to maternal exposure status. Maternal demographics were compared using Chi squared/Fishers exact tests, and odds ratios were computed using exact methods to compare crude pregnancy/fetal outcome rates.

Results: Of 320 pregnancies exposed to SNRIs during pregnancy, all included first trimester exposure, and were matched to 960 SSRI exposed and 1600 NTE comparator groups. We observed no significant difference in the rate of spontaneous abortion, elective termination or intrauterine death between exposed and comparator groups. However, the crude rate of live births was significantly decreased in the SNRI group compared with NTE (OR 0.690, 95% CI 0.545–0.932), but not with the SSRI group (OR 0.895, 95% CI 0.658–1.22). The risk of preterm delivery was significantly increased in SNRI pregnancies compared with the NTE group (OR 1.57, 95% CI 1.06–2.33); no such difference was observed when comparing SSRI exposed. A non-significant increase in major malformation risk was seen in the SNRI group compared with both comparator groups; half of the major malformations in the SNRI exposed were cardiac, showing a non-significant increase in comparison with both controls.

Conclusion: These data do not provide sufficient evidence of an increased risk of adverse pregnancy outcomes following SNRI exposure in human pregnancy, however due to the small sample size there is a limited ability to detect slight increases in adverse outcomes or rare events. Further research is therefore justified.

References:

• 1. Lennestål R, Källén B. Delivery outcome in relation to maternal use of some recently introduced antidepressants. J Clin Psychopharmacol. 2007;27:607–13

Disclosure of Interest: None declared.

42 ISoP18-1025 Sjs/Ten Induced by Carbamazepine and Oxcarbazepine: The Analysis of Drug Injury Relief Applications in Taiwan

42.1 M.-Y. Chien¹, P.-H. Chao^*1, W.-W. Chen¹

42.1.1 ¹Taiwan Drug Relief Foundation, Taipei, Chinese Taipei

Background/Introduction: For the prevention of carbamazepine -induced SJS/TEN (CBZ-SJS/TEN), genetic test of HLA-B*15:02 before carbamazepine (CBZ) prescription has been covered by the national health insurance in Taiwan since 2011. Since then, it is observed that the usage of carbamazepine was significantly declined, whereas the usage of oxcarbazepine (OXC) has risen substantially in the same period of time in Taiwan [1]. A recent study revealed that oxcarbazepine-induced SJS/TEN (OXC-SJS/TEN) is also associated with HLA-B*15:02 in Asian populations (Chinese and Thai), but the incidence and severity of OXC-SJS/TEN are less than that of CBZ-SJS/TEN [2].

Objective/Aim: To examine the trend of carbamazepine and oxcarbazepine associated SJS/TEN in Drug Injury Relief System in Taiwan.

Methods: This research analyzed the applications of Drug Injury Relief System from 1999 to 2017 in Taiwan. Culprit drugs and adverse drug reactions were classified according to their ATC and MedDRA codes, respectively.

Results: From 1999 to 2017, the total case number of drug-induced SJS/TEN is 1044. Among these cases, 218 cases are CBZ-SJS/TEN (17 death cases, 5 disability cases and 196 severe illness cases) and 17 cases are OXC-SJS/TEN (1 death cases, 1 disability cases and 15 severe illness cases). The majority cases of CBZ-SJS/TEN occurred before 2011, whereas the greater part of OXC-SJS/TEN occurred after 2011. It is also noted that cases of OXC-SJS/TEN were not seen until 2008, even though oxcarbazepine has been approved in Taiwan since 2002. The case number of OXC-SJS/TEN was relatively low compared to that of CBZ-SJS/TEN before 2013. However, it has exceeded that of CBZ-SJS/TEN in 2017.

Conclusion: The case number of drug injury application for CBZ-SJS/TEN was decreased significantly after 2012 which may be attributed to the prescribing patterns change due to the HLA-B*15:02 screening policy implemented in 2011. Despite the usage of oxcarbazepine was increased substantially after 2011, only a slight increase of OXC-SJS/TEN was observed from Taiwan Drug Injury Relief Database. However, the potential risk of SJS/TEN induced by oxcarbazepine still cannot be overlooked.

References:

• 1. Wang YH, et al. The medication risk of Stevens–Johnson syndrome and toxic epidermal necrolysis in Asians: the major drug causality and comparison to the USA FDA label. Clin Pharmacol Ther 2018 https://doi.org/10.1002/cpt.1071

• 2. Chen CB, et al. Risk and association of HLA with oxcarbazepine-induced cutaneous adverse reactions in Asians. Neurology 2017; 88: 78–86

Disclosure of Interest: None declared.

43 ISoP18-1032 Analysis of the Risk Management Plans Submitted to the Rational Drug Use and Pharmacovigilance Department at JFDA (2014–2017)

43.1 J. Jaber^*1

43.1.1 ¹Rational Drug Use and Pharmacovigilance Department, Jordan Food and Drug Administration, Amman, Jordan

Background/Introduction: Companies are required to submit a risk-management plan (RMP) to the Jordan Food and Drug Administration (JFDA) when applying for a marketing authorization (MA); as required by the Arab GVP [1] which was effective in July 2015. Accordingly Jordan pharmacovigilance guideline [2] was updated in July 2016, which makes an RMP submission obligatory—article No. 17—in the following situations: (1) Medicine registration; (2) Leaflet update that includes a new indication and/or a new age group and (3) upon JFDA request.

Objective/Aim: The aims of the analysis were to create a national database for the JFDA that includes all of the additional risk minimization measures (ARMM) for the registered medicines in Jordan. This database enhances JFDA decisions regarding proper implementation of these measures and the best approaches to increase their effectiveness.

Methods: Risk management plans submitted to the rational drug use and pharmacovigilance (RDU & Pv) department at JFDA from 2014 to 2017 were analyzed.

Results: The results showed that among all of the submitted and approved RMPs (n = 340), 33 RMPs (10%) had an ARMM; which contains at least one or all of the following; physicians educational materials, pharmacists educational materials, direct healthcare professional communication (DHCP), patients educational materials and controlled access program. The annual rate of RMPs increased gradually over the study period. There was about an 18-fold increase in the number of received RMPs. In year 2014 (12 RMPs) were submitted; 11 RMPs for originator medicines and one RMP for generic medicine, with only one RMP that had an ARMM. At the following year (53 RMPs) were submitted; 31 RMPs for originator medicines and 22 RMPs for generic medicines, six RMPs had an ARMM. At year 2016 (67 RMPs) were submitted; 38 RMPs for originator medicines and 29 RMPs for generic medicines with 10 RMPs having an ARMM. Finally in year 2017(216 RMPs) were submitted; 79 RMPs for originator medicines and 137 for generic medicines with 16 RMPs having an ARMM.

Conclusion: This study highlights the importance of establishing a robust pharmacovigilance system in Jordan, which is strongly needed to safe handle newly registered medicines (biosimilar, biological…) that required a special RMP design with proper risk minimization measures.

References:

• 1. Guideline on good pharmacovigilance practices (GVP) effective from July 2015 (Arab Countries). Available from http://www.jfda.jo/

• 2. Pharmacovigilance guidelines 2016 (Jordan). Available from http://www.jfda.jo/

Disclosure of Interest: J. Jaber Employee of: JFDA there is no conflict of interest.

44 ISoP18-1035 Herpes Simplex Virus Encephalitis After Deep Brain Stimulation for Epilepsy: A Case-Report

44.1 M. Gavard^*1, S. Chabardes², F. Bartolomei³, P.-M. Regis⁴

44.1.1 ¹Clinical Research and Innovation Department,; ²Neurosurgery Department, Grenoble Alpes University Hospital, Grenoble, France; ³Cerebral Epileptology and Rhythmology Department,; ⁴Neurosurgery Department, Timone University Hospital—APHM, MARSEILLE, France

Background/Introduction: Encephalitis is defined by the presence of an inflammatory process of the brain and clinical evidence of neurologic dysfunction [1]. It can be caused by a virus like Herpes simplex virus (HSV), which is the most commonly identified infectious cause of encephalitis in France [2].

Objective/Aim: Herpes simplex virus encephalitis triggered by neurosurgical interventions such as craniotomies has been previously described but it remains a rare occurrence [3, 4]; here we report a case of encephalitis due to herpes simplex virus-1 (HSV-1), following deep brain stimulation (DBS) starting for epilepsy.

Methods: Herpes simplex virus-1 infection was confirmed by cerebrospinal fluid (CSF) polymerase chain reaction (PCR).

Results: The patient was a 33-year-old man who benefited from implantation of leads for deep brain stimulation (target: anterior nucleus). Indeed, he suffered from epilepsy secondary to viral encephalitis at 7-months-old. About 1 month after leads implantation, the day after deep brain stimulation gets started, he developed a transient confusional state. Computerized tomogram did not reveal any bleeding. Confusion recurred 6 days later, with fever and headaches. The CSF PCR was positive for HSV1 and DBS was stopped. Aciclovir treatment was introduced.

Conclusion: DBS probably stimulated the herpes simplex virus that was asleep in this patient. The risk of HSV reactivation after DBS must be taken into account in patients who suffered from HSV brain infections in the past.

References:

• 1. Tunkel AR, Glaser CA, Bloch KC, Sejvar JJ, Marra CM, Roos KL, et al. The management of encephalitis: clinical practice guidelines by the Infectious Diseases Society of America. Clin Infect Dis 2008;47:303–27

• 2. Mailles A, Stahl JP. Infectious Encephalitis in France in 2007: A National Prospective Study Clin Infect Dis 2009:49:1838–47

• 3. Jaques DA, Bagetakou S, L’Huillier AG, Bartoli A, Vargas MI, Fluss J, et al. Herpes simplex encephalitis as a complication of neurosurgical procedures: report of 3 cases and review of the literature. Virol J 2016;13:83

• 4. De Almeida SM, Crippa A, Cruz C, De Paola L, De Souza LP, Noronha L, et al. Reactivation of herpes simplex virus-1 following epilepsy surgery. Epilepsy Behav Case Rep 2015;27:76–8

Disclosure of Interest: None declared.

45 ISoP18-1040 Adverse Drug Reactions Leading to Hospitalization in the Elderly Using Prospective Identification Versus Administrative Coding

45.1 N. Parameswaran Nair^*1, L. Chalmers¹, G. Peterson¹, B. Bereznicki¹, C. Curtain¹, L. Bereznicki¹

45.1.1 ¹Pharmacy, School of Medicine, College of Health and Medicine, University of Tasmania, Hobart, Australia

Background/Introduction: Adverse drug reactions (ADRs) are a major public health problem in the elderly, accounting for 6–12% of hospital admissions [1]. Effective detection and prevention are important to tackle the global burden of ADRs [2]. The use of different methods of ADR detection results in studies identifying varying rates and types of ADRs, and different drug classes responsible for ADRs [3]. There is a lack of data comparing different ADR identification methods in the same cohort of patients. 

Objective/Aim: To compare prospective identification of ADR-related hospital admissions in the elderly with administrative coding using the International Classification of Diseases 10th Revision Australian Modification (ICD-10-AM) coding system.

Methods: We linked the records of 768 enrolled patients from an earlier study (the PADR-EC study) [4], where clinical pharmacists identified ADRs using prospective data collection, to their hospital administrative data. We identified patients in the study whose admissions were coded as ADRs using ICD-10-AM codes. We then compared the prevalence and characteristics of ADR-related hospital admissions identified by the two approaches.

Results: According to ICD-10-AM coding, 2.7% of patients were admitted due to ADRs, while 15.0% of patients were deemed to have been admitted due to ADRs based on prospective identification by clinical pharmacists. Most (85.7%) patients coded as having an ADR-related admission were also identified as such prospectively. Haematological (23.1%) and metabolic reactions (23.1%) were frequent causes of ADRs identified by coding, whereas cardiovascular ADRs (27.8%) were more commonly identified prospectively by pharmacists. Antidepressants (16.7%) and cardiac glycosides (16.7%) were the most commonly implicated drug groups in ADRs identified by coding, whereas diuretics (28.8%) and renin-angiotensin system inhibitors (17.0%) were frequently implicated in the prospective identification.

Conclusion: Reliance on administrative coding potentially underestimates the extent of the problem of ADRs as a cause of hospitalization in the elderly, and more detailed prospective analysis of admissions provides additional targets for strategies to prevent ADRs. The types of ADRs identified also differ between the two approaches.

References:

• 1. Parameswaran Nair N, Chalmers L, Peterson GM, Bereznicki BJ, Castelino RL, Bereznicki LR. Hospitalization in older patients due to adverse drug reactions—the need for a prediction tool. Clin Interv Aging 2016;11:497–505

• 2. World Health Organization. International drug monitoring: the role of the hospital. Geneva: World Health Organization; 1969. Technical Report Series no. 425. http://apps.who.int/iris/handle/10665/40747. Accessed 10 April 2018

• 3. Thurmann PA. Methods and systems to detect adverse drug reactions in hospitals. Drug Saf 2001;24:961–8

• 4. Parameswaran Nair N, Chalmers L, Connolly M, Bereznicki BJ, Peterson GM, Curtain C, et al. Prediction of hospitalization due to Adverse Drug Reactions in Elderly Community-Dwelling Patients (The PADR-EC Score). PLoS One 2016;11:e0165757

Disclosure of Interest: N. Parameswaran Nair: None Declared, L. Chalmers: None Declared, G. Peterson: None Declared, B. Bereznicki Other: Bonnie Bereznicki has received funding from Boehringer Ingelheim Pty Ltd. Disclosed funding was for activities outside of this research., C. Curtain: None Declared, L. Bereznicki Other: Luke Bereznicki has received funding from Boehringer Ingelheim Pty Ltd. Disclosed funding was for activities outside of this research.

46 ISoP18-1045 Drug- or Herb- Induced Liver Injury in China

46.1 J.-B. Wang^*1,2, Y.-M. Guo¹, X.-H. Xiao^1,2

46.1.1 ¹Beijing 302 Hospital of China, Beijing, China; ²China National Advisory Council of Traditional Chinese Medicines-Induced Liver Injury, Beijing, China

Background/Introduction: Drug- or herb- induced liver injury (DILI or HILI) constitutes major concerns in drug safety issues [1, 2], which has been illustrated to cause high medical and socioeconomic burdens in developed countries [3]. However, the prevalence and burden of DILI or HIIL in developing countries are almost unknown [4, 5].

Objective/Aim: To investigate the pharmacoepidemiologic profiles and trends of DILI or HILI in the world’s largest developing country—China.

Methods: The national adverse drug reactions (ADR) monitoring data, nationwide health insurance data, million-scale community medical data and the hospital network medical data were involved in this study. 

Results: DILI has faster expansion than that of all ADRs in China. And DILI has been dramatically under-estimated in China, especially in primary health system. The under-recognition rate and the under-reporting rate of DILI (including HILI) in a typical primary healthcare system is over 80 and 90%, respectively. HILI is one of the major parts of DILI in China. Under-recognition rate and under-reporting rate of HILI were even higher than those of synthetic drugs. The proportion of severe reports of HILI increased faster than that of synthetic drugs. There are great regional differences in China regarding monitoring and controlling ability of DILI. The low-income/underdeveloped regions showed worse data than those high-income/developed regions.DILI has faster expansion than that of all ADRs in China. And DILI has been dramatically under-estimated in China, especially in primary health system. The under-recognition rate and the under-reporting rate of DILI (including HILI) in a typical primary healthcare system is over 80 and 90%, respectively. HILI is one of the major parts of DILI in China. Under-recognition rate and under-reporting rate of HILI were even higher than those of synthetic drugs. The proportion of severe reports of HILI increased faster than that of synthetic drugs. There are great regional differences in China regarding monitoring and controlling ability of DILI. The low-income/underdeveloped regions showed worse data than those high-income/developed regions.

Conclusion: There are significant under-estimation and under-recognition of DILI, especially HILI, in China, which suggest dramatically large medical and socioeconomic burdens. The income level positively related to the monitoring and controlling ability of DILI, which suggests substantial concerns in developing countries of low income and limited medical resources.

References:

• 1. Björnsson ES, Bergmann OM, Björnsson HK, et al. Incidence, presentation, and outcomes in patients with drug-induced liver injury in the general population of Iceland. Gastroenterology 2013;144:1419–25

• 2. Navarro VJ, Khan I, Björnsson E, et al. Liver injury from herbal and dietary supplements. Hepatology 2017;65:363–73

• 3. Sultana J, Cutroneo P, Trifirò G. Clinical and economic burden of adverse drug reactions. J Pharmacol Pharmacother 2013;4(Suppl 1): S73–7

• 4. Wang JB, Zhu Y, Bai ZF, et al. Guidelines for the Diagnosis and Management of Herb-Induced Liver Injury. Chin J Integr Med 2018; https://doi.org/10.1007/s11655-018-3000-8

• 5. Zhang L, Wong LY, He Y, et al. Pharmacovigilance in China: current situation, successes and challenges. Drug Saf 2014; 37: 765–70

Disclosure of Interest: None declared.

47 ISoP18-1051 Adverse Effects Due to Inappropriate Use of Topical Steroids for Common Dermatological Disorders: Need for Public Awareness

47.1 R. Jhaj^*1, D. Asati², D. Chaudhary¹, S. Jain², B. Sadasivam¹

47.1.1 ¹Pharmacology, All India Institute of Medical Sciences, Bhopal, India; ²Dermatology, All India Institute of Medical Sciences, Bhopal, India

Background/Introduction: Background: Despite a ban by the Indian regulatory authorities [1], a large number of irrational fixed-dose combinations (FDCs) of topical corticosteroids with antibacterial, antifungal and other drugs are available over the counter in India [2], facilitating their indiscriminate use for symptomatic treatment of dermatological disorders like acne, fungal infections, undiagnosed skin rash and as a fairness cream [3]. This has lead to an upsurge of adverse effects, including topical steroid dependent/damaged face (TSDF) [4] and is also responsible for drug resistant superficial fungal infections [5].

Objective/Aim: To assess the burden of adverse drug events due to topical steroid preparations reported to an Adverse Drug Monitoring Centre in a tertiary care hospital in Central India.

Methods: All adverse drug events due to topical steroid preparations reported to the Adverse Drug Monitoring Centre between January and December 2017 were analysed for suspected steroid preparations, indications of use, type of use (prescribed or self-administered), type of adverse effect and its outcome.

Results: A total of 224 reports of adverse drug events with topical steroids were received, making up 42.8% of total ADEs and 50.1% of cutaneous reactions reported during the study period. Males aged between 18 and 35 years (71%) were predominantly affected. Skin atrophy (50%), hypopigmentation (29.9%), striae (28.1%) and Tinea incognito/extensive Tinea (20.5%) were the major adverse effects reported. TDSF was reported in 18 (8%) cases. Thighs (45.1%) and face (27.2%) were the major sites of these reactions. Clobetasone propionate (69.2%) alone or in combination was the most frequent suspected agent, followed by betamethasone (26.3%). Most of the topical steroids had been purchased over the counter (88.8%) for treatment of dermatophytoes (88.4%). In more than half the patients, the reactions had not abated (47.8%) or abated only partially (13.8%) till last follow-up.

Conclusion: Topical steroid containing FDCs continue to be available over the counter and are responsible for various irreversible cutaneous adverse effects. Increased awareness of the treating physicians as well as the public is required to counter this problem.

References:

• 1. Pande S. Steroid containing fixed drug combinations banned by government of India: A big step towards dermatologic drug safety. Indian J Drugs Dermatol 2016; 2: 1–2

• 2. Kumar S, Goyal A, Gupta Y. Abuse of topical corticosteroids in India: Concerns and the way forward. J Pharmacol Pharmacother 2016; 7: 1–5

• 3. Saraswat A, Lahiri K, Chatterjee M, Barua S, Coondoo A, Mittal A, et al. Topical corticosteroid abuse on the face: A prospective, multicenter study of dermatology outpatients. Indian J Dermatol Venereol Leprol 2011; 77: 160–6

• 4. Lahiri K, Coondoo A. Topical steroid damaged/dependent face (TSDF): An entity of cutaneous pharmacodependence. Indian J Dermatol 2016; 61: 265–72

• 5. Verma S, Madhu R. The great Indian epidemic of superficial dermatophytosis: An appraisal. Indian J Dermatol 2017; 62: 227–36

Disclosure of Interest: None declared.

48 ISoP18-1053 The Moroccan Phytovigilance System: An Experience of Involving Patients from 2012 to 2016

48.1 S. Skalli^*1, R. Soulaymani Bencheikh²

48.1.1 ¹Faculty of Sciences, Mohammed V Universtity, Rabat, Morrocco; ²Centre Anti Poison et de Pharmacovigilance du Maroc, Rabat, Morocco

Background/Introduction: Patients frequently self-select Herbal medicines (HM), without the advice of a qualified health provider and they do not disclose their use of HM to their physicians or pharmacists. This prevents the reporting to national pharmacovigilance centres .

Objective/Aim: to describe how the active involvement of patients may remedy the under-reporting of adverse reactions (ARs) to herbal medicines (HMs) to the existing Moroccan phytovigilance program.

Methods: A survey was conducted once a week from January 2012 to December 2016 and consisted of a personal interview with patients, or their caregivers, in four clinical services: Medical Oncology, Paediatrics, Urology and Maternity in the Ibn Sina University Hospital in Rabat. The Moroccan reporting form used for spontaneous ARs reporting for was used for this surveyfor which participationwas voluntary.

Results: 1129 reports (70.5%) were in agreement with the inclusion criteria. These reports constitute 90% of the total of patient reports received by the pharmacovigilance of Herbal Medicines Unit during the same period, including reports received by the spontaneous reporting system. 52.3% of ARs were recorded in females and 47.7% of ARs in males. Children and adults were involved in 31.4 and 68.6% of reports, respectively. HMs were used either as crude material form in 96% of cases, or modern pharmaceutical form in 4% of cases. 29.3% of the HMs were in the form of a single plant and 70.7% of the HMs were a mixture of two or more plants. Twenty signals were detected of which two alerts were generated.

Conclusion: The additional reports from patient’s makes an important contribution to the monitoring of HM safety by detecting signals and generating.

Disclosure of Interest: None declared.

49 ISoP18-1056 Chronic Disease Sufferers’ Willingness to Participate in Spontaneous Adverse Drug Reaction Reporting System

49.1 V. Getova^*1,2, E. Naseva³, H. Lebanova⁴, I. Getov³

49.1.1 ¹Medical University of Plovdiv, Plovdiv, Bulgaria; ²Bulgarian Drug Agency, Sofia; Bulgaria; ³Medical University of Sofia, Sofia, Bulgaria; ⁴Medical University of Pleven, Pleven, Bulgaria

Background/Introduction: The tendency for high consumption of medicinal products in Europe is a prerequisite for the greater number of adverse drug reactions (ADRs) in some countries. Patients are valuable source of information on post-marketing drug safety data collection as they can thoroughly explain the direct impact of therapy on daily life activities [1, 3]. This is especially important for people whose chronic disease management requires every day medicines use. Hence their participation in spontaneous reporting systems is of special value and importance for the development of pharmacovigilance knowledge [2,4].

Objective/Aim: The aim of the study was to analyze and assess the level of patient knowledge on ADR reporting, with a special focus on chronic disease sufferers. The survey investigates participants’ past experience with ADRs and their attitude towards sources of reliable drug information.

Methods: A close-ended questionnaire study was conducted among patients and consumers. The minimum sample size was calculated to 267 people for a proportion of 50% with maximal error ± 6%. Quantitative variables were represented as median and range (minimum and maximum), and categorical ones—as absolute and relative frequencies. For analyzing the results were applied Pearson Chi square analysis to assess the relationship between categorical variables (for Table 2 × 2 Fisher’s Exact test was used) and Mann–Whitney and Kruskal–Wallis for comparison of quantitative variables in independent samples.

Results: The majority of the participants in the study who claimed to suffer from a chronic disease also confirmed everyday medicines intake. A third (36%) of all participants in the study reported having experienced ADRs, with a greater number amongst everyday medicine users. However, the level of participants’ awareness of the direct ADR reporting option remains insufficient and the number of reports sent to regulators—very small. Healthcare professionals (HCP) are the most preferred source of drug information followed by internet resources, friends and family.

Conclusion: The conducted study showed a notable lack of knowledge on spontaneous reporting and insufficient chronic disease patients’ involvement in collection of post-marketing drug safety data. The majority of participants would consult HCPs in all matters regarding ADRs which emphasizes the role they play in educating consumers on pharmacovigilance activities.

References:

• 1. Arlett R, Kurz X, New approaches to strenghten pharmacovigilance, Drug Discov Today Technol 2011; 8: e1–e42

• 2. Hazell L, Shakir S. Under-reporting of adverse drug reactions, Drug Saf, 2006;29: 385–96

• 3. Rolfes L, van Husel F, van der Linden L, Taxis K, van Puijenbroek E. The quality of clinical information in adverse drug reaction reports by patients and healthcare professionals: a retrospective comparative analysis, Drug Saf, 2017; 40: 607–14

• 4. PGEU Best Practice paper: Pharmacovigilance and risk minimization, 25/09/2017

Disclosure of Interest: None declared.

50 ISoP18-1057 Passive Enhanced Safety Surveillance in Children Receiving Fluenz^® Tetra Vaccination in England During the Early 2017–2018 Influenza Season

50.1 L. Hazell¹, S. Shakir^*1, J. Dentith², E. Hammond³, D. Brooks³

50.1.1 ¹Drug Safety Research Unit, Southampton, United Kingdom; ²AstraZeneca, Luton, United Kingdom; ³AstraZeneca, Gaithersburg, United States

Background/Introduction: Fluenz^® Tetra is a quadrivalent, live attenuated, intranasal, influenza vaccine recommended for use in children aged 2–17 years vaccinated as part of the seasonal influenza immunisation campaign in the UK [1]. Here, we present results from a third season of passive enhanced safety surveillance (ESS) for the vaccine in 2017–2018.

Objective/Aim: To measure and assess the frequencies of suspected adverse drug reactions (sADRs) in children receiving Fluenz^® Tetra during the early 2017–2018 influenza season in England.

Methods: Vaccinees or parents/guardians received a Safety Report Card (SRC) to return if children experienced sADRs after vaccination with Fluenz^® Tetra. At participating sites, 38 general practices and 36 primary schools in England, immunisation teams recorded numbers of SRCs distributed. The study was approved by an NHS Research Ethics Committee (Hampshire B South Central NRES Committee).

Results: Between 29th September 2017 and 2nd December 2017, 12,193 children were vaccinated at study sites, with 16 different vaccine lots used. In total, 10,793 SRCs were issued for children aged 2–17 including 5189 children (48.1%) aged 2–4 years, 5036 (46.7%) aged 5–10 years and 568 (5.3%) aged 11–17 years.

Of 114 SRCs returned (by 9th January 2018 datalock), 101 reported at least one sADR (0.9% of all SRCs issued). The most frequently reported sADRs were rhinorrhoea (n = 36), pyrexia (31) and cough (14). Three reported sADRs were classified as serious due to hospitalisation (0.03% of all SRCs issued) including nasal symptoms, general cold/flu symptoms and croup. The frequency of sADRs reported in the current ESS season was consistent with that reported in previous seasons.

Conclusion: Overall, the number of sADRs reported for Fluenz^® Tetra in the 2017–2018 ESS remains low. Reported sADRs were typically minor expected events with serious sADRs rarely reported. The pattern and frequency of reported sADRs in the 2017–2018 season were consistent with previous seasons with no safety signals detected. Despite the limited number of reports received, the ESS method continues to be a useful extension to routine pharmacovigilance activities in the monitoring of possible sADRs.

Project sponsored by DSRU and AstraZeneca.

References:

• 1. AstraZeneca. Fluenz^® Tetra. Summary of Product Characteristics. Last updated: 22 March 2018. Available at https://www.medicines.org.uk/emc/product/3296/smpc. Accessed 9 May 2018

Disclosure of Interest: L. Hazell Grant/Research support from: AstraZeneca for current project, S. Shakir Grant/Research support from: AstraZeneca for current project, J. Dentith Employee of: AstraZeneca, E. Hammond Employee of: AstraZeneca, D. Brooks Employee of: AstraZeneca.

51 ISoP18-1061 Utilisation and Safety of Asenapine in Primary Care in England: Results Rom a PASS

51.1 S. Dhanda^1,2, F. Coukan^1,2, L. Wise¹, S. Shakir^*1,2

51.1.1 ¹Drug Safety Research Unit, Southampton, United Kingdom; ²University of Portsmouth, Portsmouth, United Kingdom

Background/Introduction: Asenapine is approved in the EU for the treatment of manic episodes associated with bipolar I disorder in adults. A PASS was carried out to monitor the use and safety of asenapine using real-world primary care data in England; targeted safety outcomes were oral hypoaesthesia, oropharyngeal swelling, somnolence and sedation, acute allergic reactions and weight gain (defined as weight (kg) and/or BMI (kg/m²) increase of ≥ 7% at end of study vs. index).

Objective/Aim: To describe the utilisation and safety of asenapine in primary care in England.

Methods: A cohort study identified patients from dispensed prescriptions of asenapine in England from January 2012–June 2016. Patient characteristics, drug utilisation and outcome data were collected from prescribing general practitioners via questionnaires sent at ≥ 3, and ≥ 12 months after the 1st prescription issued for each patient. Summary descriptive statistics were calculated.

Results: The cohort consisted of 122 patients with 3-month data of whom 56 patients had corresponding 12-month data [median age 44 years (IQR 35–55); 66.4% female]. The most frequent indication was bipolar disorder (n = 81, 66.4%); for 10 patients (12.3%), the indication was specifically reported as bipolar I disorder. There were also a number of off-label indications, e.g., schizophrenia (n = 21, 17.2%). Asenapine was prescribed in accordance with the product label for 61 patients (50.0%) who were started on 5 mg bd; 19 patients (15.6%) were prescribed 10 mg bd.

For the 3-month cohort, incidence of targeted events was; oral hypoaesthesia n = 0 (0.0%), oropharyngeal swelling n = 1 (0.8%), sedation/somnolence n = 5 (4.1%), acute allergic reactions n = 0 (0.0%).

For the 12-month cohort, incidence of targeted events was; oral hypoaesthesia n = 0 (0.0%), oropharyngeal swelling n = 1 (1.8%), sedation/somnolence n = 6 (10.7%), acute allergic reactions n = 0 (0.0%), weight gain n = 10 (17.9%).

In addition to the targeted outcomes other events of interest included four deaths, one of which was a completed suicide, and a case of foetal spina bifida diagnosed in a pregnant patient taking multiple psychiatric medications.

Conclusion: Asenapine was largely prescribed in accordance with prescribing recommendations. Event counts were small and no new safety signals were identified. This study design allowed for the timely collection of drug utilisation and safety data directly from prescribing GPs. However, as a result of the small cohort size, any conclusions from this study should be put into context with results from other post-marketing studies.

Disclosure of Interest: S. Dhanda Other: The DSRU is an independent charity (No. 327206) which works in association with the University of Portsmouth. The DSRU has received funding from Merck, the manufacturer of Sycrest^®. The DSRU makes the final decision on the publication of external communications, F. Coukan Other: The DSRU is an independent charity (No. 327206) which works in association with the University of Portsmouth. The DSRU has received funding from Merck, the manufacturer of Sycrest^®. The DSRU makes the final decision on the publication of external communications. L. Wise Other: The DSRU is an independent charity (No. 327206) which works in association with the University of Portsmouth. The DSRU has received funding from Merck, the manufacturer of Sycrest^®. The DSRU makes the final decision on the publication of external communications., S. Shakir Other: The DSRU is an independent charity (No. 327206) which works in association with the University of Portsmouth. The DSRU has received funding from Merck, the manufacturer of Sycrest^®. The DSRU makes the final decision on the publication of external communications.

52 ISoP18-1062 Utilisation and Safety of Asenapine in Secondary Care in the UK: Results from the Observational Safety Evaluation of Asenapine (OBSERVA) Study

52.1 S. Dhanda^1,2, J. Slade¹, V. Obsorne¹, L. Wise¹, S. Shakir^*1,2

52.1.1 ¹Drug Safety Research Unit, Southampton, United Kingdom; ²University of Portsmouth, Portsmouth, United Kingdom

Background/Introduction: The OBSERVA specialist cohort event monitoring (SCEM) study, conducted in secondary care in the UK, formed part of an EU risk management plan to monitor the use and short-term safety of asenapine, licensed for the treatment of moderate to severe manic episodes associated with bipolar I disorder in adults. Targeted safety outcomes were oral hypoaesthesia, oropharyngeal swelling, somnolence and sedation, acute allergic reactions and weight gain (defined as weight (kg) and/or BMI (kg/m²) increase of ≥ 7% at end of study vs. index).

Objective/Aim: To describe the utilisation and safety of asenapine prescribed in secondary care in the UK.

Methods: An observational cohort study identified patients via a network of psychiatrists in collaboration with the Mental Health Research Network from November 2012–September 2016. Patient characteristics, drug utilisation and outcome data were collected via questionnaires completed by specialists sent at baseline and after 12 weeks of observation. Summary descriptive statistics were calculated.

Results: The final cohort consisted of 125 patients; the median age was 44 years (IQR 35–52); 63 (50.4%) were male. Asenapine was most frequently prescribed for bipolar disorder (n = 81, 64.8%); 15 of these patients (18.5%) had bipolar I disorder reported. There were also a number of off-label indications, e.g., schizophrenia (n = 18, 14.4%). Asenapine was largely prescribed in accordance with the product label; 56 patients (44.8%) were started on 5 mg bd and at 12 weeks 34 patients (27.2%) were taking 10 mg bd.

The incidence of targeted events was as follows; oral hypoaesthesia n = 9 (7.2%), oropharyngeal swelling n = 0 (0.0%), sedation n = 13 (10.4%), somnolence n = 11 (8.8%), acute allergic reactions n = 0 (0.0%), weight gain n = 1 (0.8%).

No deaths or pregnancies were reported in the OBSERVA study.

Conclusion: Asenapine was largely prescribed in accordance with prescribing recommendations; the majority of patients were taking asenapine for bipolar disorder at the recommended dose. Event counts were small and no new safety signals were identified. This study provides valuable real-word data to support the post marketing risk–benefit profile of asenapine, however as a result of the small cohort size, any conclusions from this study should be put into context with results from other research evidence.

Disclosure of Interest: S. Dhanda Other: The DSRU is an independent charity (No. 327206) which works in association with the University of Portsmouth. The DSRU has received funding from Merck, the manufacturer of Sycrest^®. The DSRU makes the final decision on the publication of external communications., J. Slade Other: The DSRU is an independent charity (No. 327206) which works in association with the University of Portsmouth. The DSRU has received funding from Merck, the manufacturer of Sycrest^®. The DSRU makes the final decision on the publication of external communications., V. Obsorne Other: The DSRU is an independent charity (No. 327206) which works in association with the University of Portsmouth. The DSRU has received funding from Merck, the manufacturer of Sycrest^®. The DSRU makes the final decision on the publication of external communications, L. Wise Other: The DSRU is an independent charity (No. 327206) which works in association with the University of Portsmouth. The DSRU has received funding from Merck, the manufacturer of Sycrest^®. The DSRU makes the final decision on the publication of external communications., S. Shakir Other: The DSRU is an independent charity (No. 327206) which works in association with the University of Portsmouth. The DSRU has received funding from Merck, the manufacturer of Sycrest^®. The DSRU makes the final decision on the publication of external communications.

53 ISoP18-1063 Risk of Major and Clinically Relevant Non-major (CRNM) Bleeding in Patients Prescribed Rivaroxaban in Primary Care in England

53.1 S. Dhanda^1,2, M. Davies^1,2, D. Roy^1,2, L. Wise¹, S. Shakir ^*1,2

53.1.1 ¹Drug Safety Research Unit, Southampton, United Kingdom; ²University of Portsmouth, Portsmouth, United Kingdom

Background/Introduction: Clinical trials and observational studies have reported bleeding risk in patients taking rivaroxaban. A PASS was carried out as part of the Risk Management Plan to monitor the safety and use of rivaroxaban using real-world primary care data in England.

Objective/Aim: To estimate the risk of Major and CRNM bleeding in patients prescribed rivaroxaban for Stroke prevention in non-valvular AF (SPAF) and for the prevention and/or treatment of Deep Vein Thrombosis and/or Pulmonary Embolism DVT/PE in primary care.

Methods: Patients identified from dispensed prescriptions in England (2012–2016). Detailed questionnaires sent to general practitioners (GPs) at ≥ 3 and ≥ 12 months of observation collected information on risk factors for bleeding (HAS-BLED) and bleeding outcomes. Summary descriptive statistics and 12-month risk estimates were calculated.

Results: Cohort = 17,546 patients: 10,225 patients with AF (58.3% of cohort, median age 78 years (IQR 70–84), 5253 (51.4%) male); 5959 patients with DVT/PE (34.0% of cohort, median age 66 years (IQR 50–78); 3197 (53.6%) female). In both groups, the median HAS-BLED score was 1 (IQR 1–2, 0–1, respectively) reflecting a low risk of major bleeding.

AF group: Risk Major + CRNM bleeding 8.3% [(95% CI 7.8, 8.9); n = 825]. Risk Major bleed 2.4% [(95% CI 2.1, 2.7); n = 239], CRNM bleeding 6.0% [(95% CI 5.5, 6.4); n = 592]. Major bleeding further stratified by site: gastrointestinal (GI) (1.2%; n = 117), urogenital (UG) (0.1%; n = 13), intracranial (IC) (0.4%; n = 42), all other critical organ (excluding IC) (0.3%; n = 26) and all non-critical organ sites (0.4%; n = 44).

DVT/PE group: Risk Major + CRNM bleeding 4.2% [(95% CI 3.7, 4.7); n = 240]. Risk Major bleed 1.4% [(95% CI 1.1, 1.7); n = 82], CRNM bleeding 2.8% [(95% CI 2.4, 3.3); n = 162]. Major bleeding further stratified by site: GI (0.7%; n = 38), UG (0.3%; n = 18), IC (0.2%; n = 12), all other critical organ (excluding IC) (0.1%; n = 4) and all non-critical organ sites (0.2%; n = 10).

Conclusion: For the primary outcome of major bleeding, the estimates of risk in the AF and DVT/PE rivaroxaban user populations were overall low and consistent with those estimated from clinical trial data. Differences in methodologies and analysed study populations prevent meaningful comparisons with other studies. This study design has unique strengths, including the collection of timely, granular data directly from prescribing GPs, however selective reporting of outcomes and selection bias might be present, and should be considered when interpreting results.

Disclosure of Interest: S. Dhanda Other: The DSRU is an independent charity (No. 327206) which works in association with the University of Portsmouth. The DSRU has received funding from Bayer, the manufacturer of Xarelto^®. The DSRU makes the final decision on the publication of external communications., M. Davies Other: The DSRU is an independent charity (No. 327206) which works in association with the University of Portsmouth. The DSRU has received funding from Bayer, the manufacturer of Xarelto^®. The DSRU makes the final decision on the publication of external communications. D. Roy Other: The DSRU is an independent charity (No. 327206) which works in association with the University of Portsmouth. The DSRU has received funding from Bayer, the manufacturer of Xarelto^®. The DSRU makes the final decision on the publication of external communications. L. Wise Other: The DSRU is an independent charity (No. 327206) which works in association with the University of Portsmouth. The DSRU has received funding from Bayer, the manufacturer of Xarelto^®. The DSRU makes the final decision on the publication of external communications. S. Shakir Other: The DSRU is an independent charity (No. 327206) which works in association with the University of Portsmouth. The DSRU has received funding from Bayer, the manufacturer of Xarelto^®. The DSRU makes the final decision on the publication of external communications.

54 ISoP18-1064 Pharmacovigilance in Emerging Markets: An Industry Initiative to Strengthen Global Engagement and Support

54.1 E. Herrero Martinez^*1

54.1.1 ¹Regulatory Policy and Intelligence, AbbVie, Maidenhead, United Kingdom

Background/Introduction: In recent years, there has been a marked increase in the development of pharmacovigilance (PV) legislation worldwide. This ranges from setting up basic frameworks, to regional coordination across multiple countries, often inspired by stringent regulatory requirements such as the EU Good Pharmacovigilance Practices, the US FDA, WHO and/or ICH.

Objective/Aim: All impacted stakeholders, including industry, have a responsibility to ensure robust PV frameworks are established to protect patients but monitoring developments, providing timely and aligned industry input to local bodies and ensuring compliance is increasingly challenging due to the speed and complexity of change.

Methods: For this reason, the European Federation of Pharmaceutical Industries and Associations (EFPIA) has set up a permanent team to focus on emerging market pharmacovigilance, which works in partnership with key industry and regulatory bodies to support robust and harmonised PV legislation development and to aid companies in compliance.

Results: This team is called the EFPIA International PV Group (IPVG) and it was set up 2 years ago. The IPVG is now very active, engaging globally via 5 regional workstreams, and progressing strategic objectives at a global level.

Conclusion: This presentation will introduce the team and its activities, focusing on where they may help support ISOP objectives going forward.

Disclosure of Interest: None declared.

55 ISoP18-1065 A Methodological Review of Observational Studies on Central Nervous System Drugs Use in Pregnancy and Outcomes in Children

55.1 Z. Wang¹, P. Ho², M. Choy², I. Wong^*1, K. Man²

55.1.1 ¹University College London, London, United Kingdom; ²University of Hong Kong, Hong Kong, Hong Kong

Background/Introduction: Various epidemiological approaches have been used to study the association between central nervous system (CNS) drugs use in pregnancy and CNS outcomes in children in the literature. 

Objective/Aim: To identify the methodological characteristics of existing studies in order to examine research gaps and recommend further research in this area.

Methods: A systematic literature search was conducted on observational studies written in English in PubMed that studied CNS drugs use in pregnancy and CNS outcomes in children up to 21 Sep 2017. Following independent screening and data extraction, a critical summary including the trend of relevant studies, differences between various data sources, methods to address bias and confounders, and the used statistical analysis methods was presented.

Results: 110 observational studies, 25 case–control studies, and 85 cohort studies were included in this review. Publications targeting antiepileptic drugs (AEDs) dominated in the early years, but a gradual increase in research on antidepressants (ADs) in recent years contributed to the vast majority of relevant studies to date. No study on antipsychotics (APs) was identified in this review.

Conclusion: Evidence suggests that multiple factors, such as different study designs and data source choices, lead to inconsistent findings in the association between the use of ADs and AEDs in pregnancy and CNS outcome. Researchers should optimise study design including reasonable exposure period, data source, confounding adjustment methodology, as well as statistical analysis to minimise underlying bias for better precision, validity and generalisability of results.

References:

Man KKC, Chan EW, Ip P, Coghill D, Simonoff E, Chan PKL, et al. Prenatal antidepressant use and risk of attention-deficit/hyperactivity disorder in offspring: population based cohort study. BMJ 2017;357:j2350

Hernandez-Diaz S, Smith CR, Shen A, Mittendorf R, Hauser WA, Yerby M, et al. Comparative safety of antiepileptic drugs during pregnancy. Neurology 2012;78:1692–9

Lao KS, Chui CS, Man KK, Lau WC, Chan EW, Wong IC. Medication safety research by observational study design. Int J Clin Pharm. 2016;38:676–84

Disclosure of Interest: None declared.

56 ISoP18-1066 What German Physicians Think About the Spontaneous Reporting System and How They Would Change It. Results of an Online Survey

56.1 T. Stammschulte^*1, M. Litwa¹, U. Köberle¹, L. Prause¹, K. Bräutigam¹, M. Pitzer², U. Gundert-Remy¹

56.1.1 ¹Drug Commission of the German Medical Association, Berlin, Germany; ²Klinik Rheinhöhe, Vitos Rheingau, Eltville, Germany

Background/Introduction: German physicians are requested by their code of conduct to report adverse drug reactions (ADRs) to the Drug Commission of the German Medical Association (DCGMA). However, underreporting is a noticed problem and strategies for motivation are required.

Objective/Aim: To elucidate the attitiude of German physicians towards spontaneous reporting and to identify their suggestions for improving reporting in clinical practice.

Methods: We conducted an online survey consisting of 12 questions. A link to the survey was distributed between November 2017 and February 2018 using six different newsletters. In theory, these newsletters reach approximately 120,000 physicians but a considerable overlap of recipients is expected.

Results: A total of 1974 replies were received (response rate 1.6%), most of which came from physicians (86.9%, pharmacists: 9.5%, others: 3.6%). For this analysis only responses from physicians (n = 1716) were included. 69.2% stated they have at least once suspected or diagnosed a serious ADR. 41.6% have never reported an ADR, while 43.9% have reported an ADR one to three times (4–10 times: 12%, > 10 times: 4.8%).

Reasons for not reporting that applied completely or to some extent for the respondents were: the required time (59.9%), lack of knowledge about the reporting system (50.6%) or the involved organisations (53.1%), uncertainty about which ADRs should be reported (56.1%) and privacy policy (47.61%). Only 2.2% considered reporting of ADRs to be useless, 17.5% missed incentives and 28.3% stated uncertainty about judicial consequences.

The majority would spend a maximum of 15 min for reporting an ADR (≤ 5 min: 43.9%, > 5 ≤ 15 min: 46.7%) and the most preferred ways to report were: online via website, via fax or directly via software systems used in private practice or hospital, respectively.

Apart from an acknowledgement of receipt (64.5%), physicians appreciated receiving a response regarding assessment of seriousness and causality (63.3%) as well as further information on the reported ADR (49.5%) and therapy advice (57.6%).

With regard to follow-up information there was willingness to provide additional documents (67.3%) or to answer precise questions (63.1%). Approximately one quarter of respondents considered further inquiries bothersome (23.5%) or didn’t answer them at all (3.3%).

Suggestions to improve the reporting system were given in 321 out of 1716 replies (18.7%). They referred to: simplification, reduction of required time, better information on what and how to report, avoidance of time-consuming follow-ups, the inclusion of reporting tools to clinical software systems, feedback to reports and incentives for reporting.

Conclusion: In principle, physicians seem to acknowledge the necessity of reporting ADRs. However, they require proper guidance and a fast and simple reporting system which should include useful feedback information. To our knowledge this has been the biggest survey relating to this topic ever done in Germany even though our response rate was low. We expect that only physicians who are interested in the topic in general completed the survey. Thus, our results may overestimate a positive attitude towards spontaneous reporting. However, the high proportion of respondents that have never reported an ADR prior to the survey demonstrates a capability to improve spontaneous reporting in Germany.

Disclosure of Interest: None declared.

57 ISoP18-1068 An Ecological Study on Consumer Adverse Event Reporting to the US Food and Drug Administration

57.1 M. Munoz^*1,2, G. Dal Pan¹, H. Xiao², C. Delcher³, J. Wei², C. Kortepeter¹, A. Winterstein^2,4

57.1.1 ¹Office of Surveillance and Epidemiology, US Food and Drug Administration, Silver Spring, Maryland, United States; ²Department of Pharmaceutical Outcomes and Policy,; ³Department of Health Outcomes and Biomedical Informatics,; ⁴Department of Epidemiology, University of Florida, Gainesville, Florida, United States

Background/Introduction: For nearly 50 years, the FDA Adverse Event Reporting System (FAERS) has been a central component of postmarketing drug safety surveillance. FAERS relies on health care professionals, consumers, and others to voluntarily report adverse events. Potential disparities in United States (US) consumer reporting have not been previously examined.

Objective/Aim: To determine if consumer reporting is uniform across the US and to evaluate potential factors associated with geographic reporting patterns.

Methods: We extracted all FAERS reports directly submitted to the FDA by US consumers with an FDA initial received date from January 1, 2011 to December 31, 2015. Reports were geocoded and aggregated by county. Population data from the US Census Bureau’s American Community Survey (ACS) was used to calculate the number of reports per 100,000 residents (i.e., reporting rate). To evaluate the correlation of county-level sociodemographic, socioeconomic, and health factors to reporting, we collapsed counties into reporting rate quartiles. Population characteristics in areas of high reporting (highest quartile) were compared to low reporting (lowest quartile). The characteristics evaluated were derived from measures in ACS and County Health Rankings and Roadmaps (CHR&R) data. Ongoing analyses include utilizing small-area estimation techniques to produce stable county-level estimates.

Results: We identified 52,403 consumer reports in the 5-year study period of which 90.3% were successfully geocoded to the county-level. One-fifth (19.5%) of the 3136 counties in the US had no reports in FAERS, with a median population of 6572 residents. Counties ranked lowest (n = 784) and highest (n = 784) in reporting had ≤ 5.5 reports per 100,000 residents and ≥ 17.6 reports per 100,000 residents, respectively, over the 5 years. Counties in the highest reporting quartile had significantly lower proportions of Hispanic (6.5 vs. 11.8%, p < 0.01), American Indian/Alaskan Native (1.5 vs. 4.0%, p < 0.01), and extent of rural area (57.7 vs. 75.4%, p < 0.01) than counties in the lowest reporting quartile. Counties in the highest reporting quartile also had higher mean household income ($50,506 vs. $45,651, p < 0.01), more primary care providers (64.2 vs. 43.7, p < 0.01), and more mental health providers (150.8 vs. 95.1, p < 0.01) per 100,000 residents. After excluding counties with < 20,000 residents, the significant patterns remained consistent.

Conclusion: Our findings suggest there are significant variations in consumer reporting across the US. Population representativeness in adverse event reporting is important given certain patient groups may be at an increased risk for adverse reactions. Opportunities exist for more rigorous evaluations of reporting disparities.

Disclosure of Interest: None declared.

58 ISoP18-1069 Multi-Level Modelling to Investigate Factors Impacting Prescribing Variability

58.1 D. Roy^1,2, L. Wise¹, S. Shakir^*1,2

58.1.1 ¹Drug Safety Research Unit, Southampton, United Kingdom; ²University of Portsmouth, Portsmouth, United Kingdom

Background/Introduction: Prescribing guidelines influence treatment choice based on patient and healthcare system factors. Multi-level modelling can provide insight into sources of variability in healthcare, especially where nested hierarchical structures exist [1]. A Specialist Cohort Event Monitoring study investigated the safety and use of rivaroxaban in clinical use, with a warfarin cohort for context.

Objective/Aim: Study to investigate prescribing variability using Multi-level modelling.

Methods: Data on NHS acute trusts in England/Wales (e.g. population size, trust type) were linked to study patient demographic and drug utilization data, and prescriber details (e.g. degree, specialty).

Using Multi-level modelling we explored the influence of patient, prescriber and trust characteristics on prescribing variability in 2106 rivaroxaban (59%) vs. 1468 warfarin (41%) adult patients nested in 780 prescribers, nested in 73 trusts. The majority of patients had an indication of DVT/PE (56.4%) or non-valvular AF (AF) (41.2%). The binary outcome was rivaroxaban or warfarin treatment.

Variance components estimate the variability accounted for by each level in the model and are expressed as: Median Odd Ratios (MOR—median relative increase in odds of rivaroxaban treatment if patient changed prescriber (PR) or trust (T)); Proportional Change in Variance (PCV) between models when successively adding fixed effects.

Results: DVT/PE group:

Adjusting for patient factors, MOR_T = 6.9 (PCV = − 0.6%); MOR_PR = 2.8 (PCV = 3.8%).

Adjusting for patient and prescriber factors, MOR_T = 6.8 (PCV = − 2.9%); MOR_PR = 2.6 (PCV = − 15.4%).

Adjusting for patient, prescriber and trust factors, MOR_T = 4.9 (PCV = − 30.2%); MOR_PR = 2.6 (PCV = 3.4%).

Differences between trusts and prescribers (in trusts) are important in treatment choice; trust being more influential. Some patient factors had a relatively large effect on odds of treatment choice although the absolute number of patient impacted was often small. Trust type, was shown to be associated with the odds of treatment choice [Final model: foundation vs acute trusts OR 4.0 (95% CI 1.5, 9.9)].

Data on AF and all indications (combined) will be included in the final presentation.

Conclusion: This study highlights the utility of Multi-level modelling in exploring patient and non-patient factors in nested hierarchical healthcare settings. Prescribing variability appears dominated by differences between trusts and prescribers (in trusts). Some patient factors were important in treatment choice, but PCV between models suggest that accounting for patient differences does not fully explain the variance between prescribers (in trusts) and between trusts.

References:

• 1. Zuidgeest MG, van Dijk L, Spreeuwenberg P, Smit HA, Brunekreef B, Arets HG, et al. What drives prescribing of asthma medication to children? A multilevel population-based study. Ann Fam Med 2009;7:32–40

Disclosure of Interest: D. Roy Other: The DSRU is an independent charity (No. 327206) which works in association with the University of Portsmouth. The DSRU has received funding from Bayer, the manufacturer of Xarelto^®. The DSRU makes the final decision on the publication of external communications. L. Wise Other: The DSRU is an independent charity (No. 327206) which works in association with the University of Portsmouth. The DSRU has received funding from Bayer, the manufacturer of Xarelto^®. The DSRU makes the final decision on the publication of external communications. S. Shakir Other: The DSRU is an independent charity (No. 327206) which works in association with the University of Portsmouth. The DSRU has received funding from Bayer, the manufacturer of Xarelto^®. The DSRU makes the final decision on the publication of external communications.

59 ISoP18-1074 Objectives and Design of the Post-Authorization Studies Evaluating the Effectiveness of the Risk Minimization Measures in the EU PAS Register

59.1 A. Farcas^*1, M. Huruba¹, C. Mogosan¹

59.1.1 ¹Drug Information Research Center, University of Medicine and Pharmacy Cluj-Napoca, Cluj-Napoca, Romania

Background/Introduction: Risk minimization measures (RMMs) are public interventions aimed to prevent the occurrence of adverse events correlated with exposure to a medicinal product. Routine or additional, these RMMs aim to optimize the safe and effective use of a medicine all throughout its lifecycle.

Objective/Aim: The aim of this review is to characterize the objectives, methods and study design used in post-authorization studies that evaluate the effectiveness of RMMs.

Methods: All studies that aimed to evaluate the effectiveness of RMMs, available in the EU Electronic Register of Post-Authorization Studies (EU PAS Register) up to 31st of April 2018, were included in our analysis.

Results: Out of the 359 effectiveness studies registered in the EU PAS Register, 31 evaluated the effectiveness of at least one RMM. 2 studies were excluded from the review due to early termination. Out of the 29 relevant studies, 18 were finalized, 6 were ongoing and 5 were planned at the time of the present analysis. A total of 9 studies evaluated routine RMMs [SPC (n = 9), PIL (n = 3) and  label updates (n = 1)], while 27 analyzed additional RMMs [patient/caregiver brochure (n = 9), prescriber guide (n = 8), Patient Alert Card (n = 8), healthcare provider brochure (n = 8), other educational materials (n = 4), Direct Healthcare Provider Communication (n = 4), prescriber checklist (n = 3), website (n = 1), Q&A (n = 1)].

With respect to study design, 24 studies used primary data sources, while 8 used secondary, analyzing chart review (n = 4), databases such as IMS (n = 2), patient based data (n = 2), EMA (n = 1), EU PAS Register (n = 1), Public Health England (n = 1), pharmaceutical companies database (n = 1), MedLINE (n = 1), NHS Business Services Authority (n = 1); 24 studies were surveys, 20 cross-sectional, 6 observational, 5 non-interventional, 2 cohort studies, 1 was a systematic review and 1 a meta-analysis.

The top 5 countries where effectiveness studies were conducted in were France (n = 20), Finland (n = 20), United Kingdom (n = 18), Spain (n = 16) and Ireland (n = 13).

Conclusion: Risk minimization programs integrate interventions designed to support safe and adequate use of medicines. A qualitative review of these studies adds to the fundament of development of the process design, and communication of the results in a standardized and transparent manner.

Disclosure of Interest: None declared.

60 ISoP18-1076 Profile of Adverse Events Related to Long-Lasting Insecticide-Treated Nets During Mass Campaigns in the Democratic Republic of the Congo

60.1 N. M. T. Mpiempie^*1, G. K. Ilombe ¹, N. N. Koka ², F. B.-M. Makila¹, M. O. Mboma¹, N. C. Nsibu³, L. G. Tona¹, M. P. Mitashi⁴, K. G. Mesia⁵

60.1.1 ¹National Pharmacovigilance Centre, Kinshasa, Congo, The Democratic Republic of the; ²National Malaria Control Program, Kongo Central, Congo, The Democratic Republic of the; ³Paediatric, Kinshasa University Hospital, Kinshasa, Congo, The Democratic Republic of the;, ⁴Tropical Medecine, University of Kinshasa, Congo, The Democratic Republic of the; ⁵National Pharmacovigilance Centre, Kinshasaa, The Democratic Republic of the Congo

Background/Introduction: The Long-Lasting Insecticide-Treated Nets (LLIN) is still used as the main tool of malaria prevention in the Democratic Republic of Congo (DRC). Approximately 23 million LLINs were distributed during mass campaigns between 2016 and 2017, with a coverage rate approaching 80%. However, the utilization rate remains low in some endemic communities. Pyrethroids group are recommended by the World Health Organization for nets impregnation. No relevant studies have investigated the potential adverse reactions of exposure to LLIN in DRC. It is therefore necessary to establish its safety profile in order to improve its use in the community.

Objective/Aim: Determine the profile of adverse events related to the use of the long-acting insecticide-treated net, their timing and duration.

Methods: This is a cross-sectional study nested on a randomized controlled LLIN effectiveness of 2 LLIN Dawa plus 2.0 (impregnated with deltamethrin) and Permanet 3.0 (impregnated with deltamethrin + PBO) in the communities of Kisantu in DRC. Thirty villages were randomly selected. An adverse event (AE) collection form was used during interviews with the heads of households. Data were entered using the Epi-info7 and analyzed with SPSS 20.0. The Pearson’s Chi square test was used to compare proportions of AEs between categories format of LLIN, with a p value < 0.05 being considered significant.

Results: Two LLIN formats were distributed: 60% Dawa plus versus 40% Permanet 3.0. Of the 505 households visited, 200 (42.2%) reported experiencing an AE. A total of 306 AEs have been reported including 1 serious adverse event (0.3%). Half of the AEs (49.7%) were respiratory (cough, cold, dyspnoea, asthma, bronchitis, choking), 36.9% cutaneous (pruritus, tingling, burning sensation) and 13.1% ocular. The occurrence of AEs is statistically related to the use of the Dawa plus format (X² = 12.845, ddl 3, p = 0.005). The onset of AE was 1.5 days while the duration was estimated to 5.15 days.

Conclusion: The use of Dawa plus LLLIN format is significantly associated with the appearance of AEs. Moreover, the occurrence of AEs may hinder the acceptability of LLIN in the community and negatively impact on malaria implementation strategies in the DRC.

References:

Sécurité d’emploi des moustiquaires imprégnées de pyrethrinoides WHO/CDS/CPE/WHOPES/99.5FR

Disclosure of Interest: None declared.

61 ISoP18-1077 Incidence of Major and Clinically-Relevant Non-major Bleeding in Patients Prescribed Rivaroxaban: Results RROM the ROSE Study

61.1 A. Evans^1,2, M. Davies^1,2, L. Wise¹, S. Shakir^*1,2

61.1.1 ¹Drug Safety Research Unit, Southampton, United Kindgom; ²University of Portsmouth, Portsmouth, United Kingdom

Background/Introduction: The ROSE (Rivaroxaban Observational Safety Evaluation) Specialist Cohort Event Monitoring (SCEM) study was conducted to monitor the safety and use of rivaroxaban for Stroke Prevention in patients with non-valvular AF (SPAF) and the prevention and/or treatment of Deep Vein Thrombosis and/or Pulmonary Embolism (DVT/PE) in the secondary care setting in England and Wales. The study complements another study conducted in primary care by focusing on the acute phase of treatment.

Objective/Aim: To estimate the incidence of major and clinically-relevant non-major (CRNM) bleeding (ISTH definition) in patients prescribed rivaroxaban for the first time for SPAF and DVT/PE indications.

Methods: Patients were identified through speciality groups (2013–2016), supported by UK Clinical Research Networks. Risk factors for bleeding (HAS-BLED) and bleeding outcomes were collected via detailed questionnaires completed by hospital specialists at baseline and ≥ 12 weeks. Summary descriptive statistics and 12 week incidence risk and rate (per 100 patient years) were calculated.

Results: The cohort consisted of 2542 patients including 965 patients with AF [38.0% of cohort; median age 76 years (IQR 69, 83); 517 (53.6%) male], median HAS-BLED score 2 (IQR 1–3) (moderate risk of major bleeding) and 1532 patients with DVT/PE [60.3% of cohort; median age 63 years (IQR 48, 73); 836 (54.6%) male], median HAS-BLED score 1 (IQR 0–2) (low risk of major bleeding).

AF group: Rate: Major + CRNM bleeding 28.2 (95% CI 21.0, 37.1; n = 51), Major Bleed 5.5 (95% CI 2.6–10.1; n = 10), CRNM bleeding 22.7 (95% CI 16.3–30.8; n = 41).

Major Bleed risk in pre-specified sites: gastrointestinal (GI) (0.2%; n = 2), urogenital (UG) (0.2%; n = 2), intracranial (IC) (0.2%; n = 2) and all other critical organ (excluding IC) (0.1%; n = 1).

DVT/PE group: Rate: Major + CRNM bleeding 36.2 (95% CI 29.4, 44.1; n = 98), Major Bleed 8.3 (95% CI 5.3, 12.5; n = 23), CRNM bleeding 27.6 (95% CI 21.7, 34.6; n = 75).

Major Bleed risk in pre-specified sites: GI (0.7%; n = 11), UG (0.3%; n = 5), IC (0.1%; n = 1) and all other critical organ (excluding IC) (0.0%; n = 0).

Conclusion: The incidence of major bleeding was low and no new safety concerns were raised. Unique aspects of this study design enabled collection of highly detailed information from cardiologists, faciliating accurate calculation of risk scores and the application of clinical trial outcome definitions, in a real-word setting. Differences in methodology between this and other study designs prevents meaningful comparison.

Disclosure of Interest: A. Evans Other: The DSRU is an independent charity (No. 327206) which works in association with the University of Portsmouth. The DSRU has received funding from Bayer, the manufacturer of Xarelto^®. The DSRU makes the final decision on the publication of external communications. M. Davies Other: The DSRU is an independent charity (No. 327206) which works in association with the University of Portsmouth. The DSRU has received funding from Bayer, the manufacturer of Xarelto^®. The DSRU makes the final decision on the publication of external communications. L. Wise Other: The DSRU is an independent charity (No. 327206) which works in association with the University of Portsmouth. The DSRU has received funding from Bayer, the manufacturer of Xarelto^®. The DSRU makes the final decision on the publication of external communications. S. Shakir Other: The DSRU is an independent charity (No. 327206) which works in association with the University of Portsmouth. The DSRU has received funding from Bayer, the manufacturer of Xarelto^®. The DSRU makes the final decision on the publication of external communications.

62 ISoP18-1078 Tamoxifen and the Risk of Parkinsonism: A Case Non-case Study

62.1 F. Montastruc¹, F. Khosrow-Khavar², J. Benevent¹, C. Renoux², J.-L. Montastruc^*1

62.1.1 ¹Medical Pharmacology, Faculté de Médecine—Centre Hospitalier Universitaire, Toulouse, France; ²Clinical Epidemiology, Jewish General Hospital and McGill University, Montreal, Quebec, Canada

Background/Introduction: Three recent epidemiological studies have suggested an association between use of tamoxifen in breast cancer and risk of Parkinsonism, mainly after long term exposure (4–6 years).

Objective/Aim: Since these studies have several limitations, we explored this potential signal, performing a case/non-case study using Vigibase^®, the World Health Organization Global Individual Case Safety Reports (ICSRs) database, between 1979 and 2018.

Methods: Among post-menopausal women aged ≥ 55 years, we measured the risk of reporting “Parkinsonism” compared with all other adverse drug reactions [as a reporting odds ratio with 95% confidence interval (ROR 95%CI)] for tamoxifen compared to all other drugs. To assess the stability of our results, we performed several sensitivity analyses, first, comparing tamoxifen to aromatase inhibitors (another drug class used in breast cancer) and, second, excluding reports with drugs known to induce Parkinsonism.

Results: Among 13,964 women aged ≥ 55 years and exposed to tamoxifen, we identified 356 cases of Parkinsonism (mainly in women aged 65–74 years, n = 146, 41%) and 13,599 non cases. Among cases exposed to tamoxifen, MedDRA Preferred Terms reported were mainly tremor (n = 125, 35.1%) and gait disturbance (n = 118, 33.1%). We failed to find a positive association between tamoxifen exposure and Parkinsonism in comparison with exposure to other drugs [ROR = 0.79 (95% CI 0.71–0.88)] or aromatase inhibitors [ROR = 0.39 (95% CI 0.33–0.46)]. No association was found after exclusion of reports with drugs known to induce Parkinsonism [ROR = 0.76 (95% CI 0.68–0.85)]. In contrast, a significant association was found with flunarizin, a drug well known to induce Parkinsonism, selected as a positive exposure control [ROR = 6.76 (95% CI 5.96–7.68)].

Conclusion: This study, performed in a real world context and in a large pharmacovigilance database, did not allow to confirm a potential signal of tamoxifen and Parkinsonism, whatever the comparator used.

Disclosure of Interest: None declared.

63 ISoP18-1080 Characterization and Disproportionality Analysis of Spontaneously Reported Cases of Medication Errors to the U S-F D A Adverse Event Reporting System

63.1 F. Mazhar^*1, C. Carnovale², M. Gentili¹, E. Clementi ¹, S. Radice^3,4

63.1.1 ¹Department of Biomedical and Clinical Sciences L. Sacco, “ Luigi Sacco”, Università di Milano, Milan, Italy; ²Department of Biomedical and Clinical Sciences L. Sacco, “ Luigi Sacco”, Milan, Italy; ³Scientific Institute, IRCCS E. Medea, Bosisio Parini LC, Italy; ⁴Department Biomedical and Clinical Sciences, CNR Institute of Neuroscience, L. Sacco University Hospital, Università di Milano, Milan, Italy, Milano, Italy

Background/Introduction: Medications errors (ME) are a leading source of unintended patient harm in worldwide. In recent past years, the World Health Organization (WHO) has also highlighted the importance of identifying ME and is now working toward expansion of existing roles of pharmacovigilance centres to include monitoring of ME. Having a strong pharmacovigilance system would also help in running and strengthening a good ME reporting and monitoring program.

Objective/Aim: To characterize adverse reactions associated with medication errors (ME) reported in US Food and Drug Administration Adverse Event Reporting System (US-FAERS), and to identify the signals of disproportionate reporting (SDR) for different drugs over the period 2004–2017.

Methods: Individual Case Study Report (ICSRs) were identified through the narrow Standardised Medical Dictionary for Regulatory Activities (MedDRA) Query for ME. ICSRs were categorized by MedDRA^® terms, patient age groups, affected stages of medication process and Anatomical Therapeutic Chemical classification system. We performed a disproportional analysis of drugs mostly reported according to age groups by using reporting odds Ratio (ROR), proportional reporting ratio (PRR) and information component (IC). A value of ROR-1.96SE > 1, PRR ≥ 2, IC-2SD > 0 were considered as the positive signal.

Results: 46,8677 ICSRs associated with ME were retrieved. An increasing trend in reporting activity were observed during the study period, with a peak in the year 2015. About 57% of ME ICSR occurred in adults, 35% in elderly and 8.2% in children. Majority of reports were submitted by consumers. Immunosuppressants (L04) and Psycholeptics (N05) were most frequently involved. Administration errors were reported most frequently (87%), followed by prescribing (5%) and dispensing errors (5%). In neonates, SDR reporting for wrong drug administration, wrong dose, and accidental overdose were associated with methylergonovine, zidovudine, and acetaminophen. In elderlies, SDR were found for dose omission and underdose error associated with etanercept and evolocumab.

Conclusion: The reporting rate of ME increased between 2004 and 2017 likely due to an increased awareness of ME in consumers and physician. Overall, a wide variety of SDR for drug-ME terms were identified, some of which in agreement with the literature, suggesting that ME recording and assessment within the existing FAERS dataset is feasible. The inclusion of relevant aspects of ME, such as risk factors that cannot be coded with MedDRA, could enhance detection of ME by using spontaneous reports.

References:

• 1. Gandhi TK, Weingart SN, Borus J, Seger AC, Peterson J, Burdick E, et al. Adverse drug events in ambulatory care. N Engl J Med 2003; 348:1556–64

• 2. WHO. Patient safety: extending the scope of pharmacovigilance centres. WHO Pharmaceuticals Newsletter No. 6. Geneva: WHO, 2007

• 3. Bencheikh RS, Benabdallah G. Medication errors: pharmacovigilance centres in detection and prevention. Br J Clin Pharmacol 2009; 67: 687–90

• 4. FDA adverse event reporting system. http://www.fda.gov/drugs/guidancecomplianceregulatoryinformation/surveillance/adversedrugeffects/default.htm. Accessed 27 Jan 2018.

Disclosure of Interest: None declared.

64 ISoP18-1082 Medication Prescribing Practices of Eating Disorder Consultants for the Treatment of Young People with Anorexia Nervosa: A Questionnaire Study

64.1 M. Yakhchi Beykloo¹, D. Nicholls², M. Simic³, R. Brauer¹, I. Wong^*1

64.1.1 ¹Research Department of Practice and Policy, University College London, London, United Kingdom; ²Great Ormand Street Hospital, ³South London and Maudsley NHS Foundation Trust, London, United Kingdom

Background/Introduction: Anorexia Nervosa (AN) is a condition that onsets in young people (YP). First line treatment consists of psychological interventions; however psychotropic medications are sometimes prescribed if there is a less than optimal treatment response [1, 2].

Objective/Aim: This study aimed to describe contemporary prescribing practices of consultant child and young people eating disorder (CAED) psychiatrists, including the most common medications they prescribe and their process of continuing pharmacotherapy among AN patients.

Methods: We developed a self-administered exploratory questionnaire which was distributed among all child and adolescent eating disorder services (CYP EDS) in England during the Health Education England national training days.

Results: From 71 CYP EDS, 40 teams represented by an ED consultant responded (response rate 56%). Most (40%) consultants estimated < 10% of patients with AN to be on psychotropic medications. 38% of consultants reported olanzapine as their most commonly prescribed medication for AN, followed by fluoxetine (29%) and sertraline (10%). The minimum olanzapine initiation dose was at 2.5 mg for 2–4 weeks, reaching a maximum dose of 5 mg. 38% ED consultants met their patients once every 2 weeks on average, whereas 30 and 22% met once weekly and once monthly, respectively. Over 40% of consultants reported to continue olanzapine prescribing with the ED service teams, in contrast to the 30% who transferred prescribing to general practitioners.

Conclusion: The study showed that despite a lack of evidence, psychotropic medications are still prescribed, most commonly olanzapine. Further evidence is needed on which patients may potentially benefit from pharmacotherapy as an adjunct to psychological interventions.

References:

• 1. Gowers S., et al. Drug prescribing in child and adolescent eating disorder services. Child Adolesc Ment Health 2010; 15: 18–22

• 2. McKnight RF, Park RJ. Atypical antipsychotics and anorexia nervosa: a review. Eur Eat Disord Rev 2010; 18:10–21

Disclosure of Interest: None declared.

65 ISoP18-1084 The Efficacy and Safety of Psychotropic Drug Treatment in Adolescents with Anorexia Nervosa: A Systematic Review

65.1 M. Yakhchi Beykloo¹, A. Naser¹, D. Nicholls², M. Simic³, R. Brauer¹, I. Wong^*1

65.1.1 ¹Research Department of Practice and Policy, University College London, London, United Kingdom; ²Great Ormand Street Hospital, London, United Kingdom; ³South London and Maudsley NHS Foundation Trust, London, United Kingdom

Background/Introduction: Anorexia Nervosa (AN) is a mental disorder that onsets in adolescents. Treatment of anorexia nervosa is mainly carried out through a multidisciplinary approach, with psychological interventions given as first line treatment. However, psychotropic medications are sometimes prescribed if initial psychotherapy does not provide an optimal treatment response [1, 2].

Objective/Aim: This systematic review aims to review the existing literature on the safety and efficacy of psychotropic medications treatment in adolescents with anorexia nervosa.

Methods: The PubMed, EMBASE, PsychINFO and Cochrane Review databases were searched for relevant studies published up to October 2017 investigating psychotropic medications and their safety and efficacy in adolescents with AN. Randomised controlled trials (RCTs) and observational studies were included with no restrictions on study design or psychotropic medication therapeutic class. The primary outcome was defined as weight change as a measurement of drug efficacy, while secondary outcomes included safety of psychotropic medications with regards to adverse events, side effects, safety measurements and death.

Results: A total of 824 citations were identified of which 10 studies were included in our narrative review. These comprised of three RCTs and seven observational studies, covering a total of 1032 individuals ranging from 10 to 21 years of age. All studies did not find a significant association between the use of psychotropic medications for adolescents with AN and weight change. 70% of the studies (7 out of 10) confirmed safety issues among users of psychotropic medications, however the safety outcomes varied even between medications from the same therapeutic class, thus prevented us from aggregating the results.

Conclusion: The findings of this review fails to provide strong evidence for the efficacy of the use of psychotropic medications treatment in adolescents with anorexia nervosa. These preliminary results demonstrate a need to characterise the behaviours that lead clinicians to prescribe, and to explore the effects of medication on those specific behaviours, in order to establish the therapeutic efficacy and safety of psychotropics in the treatment of adolescents with anorexia nervosa.

References:

• 1. Gowers S, Claxton M, Rowlands L, Inbasagaran A, Wood D, Yi I, et al. Drug prescribing in child and adolescent eating disorder services. Child Adolesc Ment Health 2010;15:18–22

• 2. McKnight RF, Park RJ. Atypical antipsychotics and anorexia nervosa: a review. Euro Eat Disord Rev 2010;18:10–21

Disclosure of Interest: None declared.

66 ISoP18-1085 Risk Minimization Measures of a Paediatric Orphan Drug for Treatment of Neonatal Diabetes

66.1 M. Exposito^*1, J. Magnaud¹

66.1.1 ¹Pharmacovigilance, AMRING, Paris, France

Background/Introduction: In the context of Risk Management Plan for the Marketing authorization, specific Risk Minimization Measures (RMM) have been developed due to particular presentations of a glibenclamide suspension (2 dosages/2 syringes in 4 different presentations).

Objective/Aim: To reduce risks in a vulnerable population.

Methods:

• 1. Risks identification

• a Safety issues identified in a very small population during clinical trial [1],

• b adverse events described in glibenclamide product monograph for adult population,

• c missing information on neonatal population.

  • 2. Risks assessment

For each important risk, evaluation was performed to determine if additional minimization measures were necessary.

Results: Hypoglycaemia due to mix ups of the different strengths and different syringes was identified as major potential risk, especially in neonates and infants, for which routine minimization measures were considered as insufficient.

Consequently, three minimization measures were implemented:

• 1. Packaging: The dosage is clearly identified on the packaging by two different colours with reverse type. One syringe is thin and small while one syringe is thick and long. An appropriate drawing of the syringe included in the pack is present on the outer carton,

• 2. specific questionnaire for hypoglycaemia in order to further characterize the risk,

• 3. prescriber guide: provided to prescribers to describe each steps and parameters to be considered for prescription of the right dose and the right syringe.

Conclusion: Although the safety profile of glibenclamide is well known due to its well-established use, the context of an orphan drug intended for very young population associated to particular presentations has led to the development of specific MMRs. The effectiveness will be monitored during pharmacovigilance activities.

References:

• 1. Bouazza N, Djerada Z, Gozalo C, Busiah K, Beltrand J, Berdugo M, et al. Evaluation of the pharmacokinetics of glibenclamide tablet given, off label, orally to children suffering from neonatal syndromic hyperglycemia. Euro J Clin Pharmacol 2016;72:1373–9

Disclosure of Interest: None declared.

67 ISoP18-1086 Process Indicators vs. Correlated Endpoints in Studies Evaluating the Effectiveness of Risk Minimization Measures in the EU PAS Register

67.1 A. Farcas^*1, M. Huruba¹, C. Mogosan¹

67.1.1 ¹Drug Information Research Center, University of Medicine and Pharmacy Cluj-Napoca, Cluj-Napoca, Romania

Background/Introduction: An important pillar of risk management is the evaluation of the risk minimization measures (RMMs) implemented in order to meet a targeted safety concern. The implementation and effectiveness of the RMMs can be evaluated through process indicators correlated with pre-established endpoints.

Objective/Aim: The aim of this review is to characterize process indicators and correlated endpoints from post-authorization studies that evaluate the effectiveness of RMMs.

Methods: All finalized studies that aimed to evaluate the effectiveness of RMMs, available in the EU Electronic Register of Post-Authorization Studies (EU PAS Register) up to 31st of April 2018, were included in the study.

Results: Out of the 359 effectiveness studies registered in the EU PAS Register, 31 evaluated the effectiveness of at least one RMM. 2 studies were excluded from the review due to early termination. Out of the 29 relevant studies 18 were finalized at the time of the present analysis. Among the finalized 18, 12 studies had both full protocol and final study results uploaded in the database, 2 had only the protocol available and 1 had only the final results uploaded.

Among the finalized studies, RMMs were evaluated through process indicators such as: knowledge (n = 18), behavior (n = 16), understanding (n = 15), utilization (n = 15), awareness (n = 10), receipt of materials (n = 9), and distribution (n = 3); 11 studies discussed the evaluation of the process indicators versus correlated endpoints, consisting of whether or not the pre-established threshold for assessed process indicator(s) was met. 7 mentioned that the study was effective, thus successfully proved the effectiveness of the analyzed RMMs by meeting their target endpoints: awareness (n = 6), knowledge (n = 5), receipt of materials (n = 4), behavior (n = 3), understanding (n = 2), utilization (n = 1), while only 1 study mentioned that the results led to discontinuation of RM tools, due to proved inefficacy.

Conclusion: Evaluation of effectiveness through process indicators versus clearly established endpoints facilitates the standardization of these studies, thus adding to development of further guidance to improve study designs and reporting.

Disclosure of Interest: None declared.

68 ISoP18-1087 A Description of Medicines Associated Safety Issues Evaluated Through a Referral Procedure at the EU Level

68.1 A. Farcas^*1, T. Balcescu¹, C. Mogosan¹

68.1.1 ¹Drug Information Research Center, University of Medicine and Pharmacy Cluj-Napoca, Cluj-Napoca, Romania

Background/Introduction: Important safety issues are evaluated through a referral procedure at the EU level, when the Pharmacovigilance Risk Assessment Committee (PRAC) within the European Medicines Agency (EMA) is requested to conduct a scientific assessment of a particular medicine/class of medicines and to make a recommendation for a harmonized position across the EU.

Objective/Aim: To describe all referrals assessed by the PRAC and having a final recommendation.

Methods: Publicly available data on safety issues assessment through a referral procedure for the period June 2012–March 2018 were collected, analyzed and classified according to predefined criteria.

Results: 46 safety-related referrals were assessed by PRAC for 42 medicines/classes of medicines/combinations in our study’s timeframe. The ATC drug classes for which most referrals were initiated were the nervous system class (n = 8), blood and blood forming organs class (n = 6), respiratory system class (n = 6), genito-urinary system and sex hormones class (n = 5) and musculo-skeletal system class (n = 5). 8 drugs were authorized in the last 5 years, the rest being well-established drugs.  Referrals were triggered for cardiac disorders (n = 14), blood and lymphatic system disorders (n = 6), hepatobiliary disorders (n = 6) and nervous system disorders (n = 6). Most referrals were triggered through article 31 (n = 27), followed by article 20 (n = 12) and fewer by article 107i, considered an urgent procedure (n = 7). Referrals were initiated mostly by the European Commission (13), France (7), United Kingdom (6), and Germany (5). In three cases (for flupirtine, hydroxyethyl-starch solutions and valproate) re-assessments of the same risks were carried out by PRAC due to the fact that previous recommended restrictions have not been sufficiently effective. Final recommendations consisted in summary of product characteristics (SmPC) and package leaflet (PIL) updates in 38 cases, in sending a Direct Healthcare Professional Communication in 28 cases, and in other additional risk minimization measures. Recommendations for withdrawal were made for 7 active substances and 6 pharmaceutical forms. The SmPC sections that was most often modified towards a safer drug use were special warnings and precautions for use (n = 35), undesirable effects (n = 26), contraindications (n = 21) and posology and method of administration (n = 20) sections.

Conclusion: In most referral procedures PRAC recommended routine or additional risk minimization measures such as amendments to the product information and communication to healthcare professionals, towards an optimized use of medicinal products.

Disclosure of Interest: None declared.

69 ISoP18-1092 Doxycycline Associated Fixed Drug Eruptions Safety Review

69.1 A. Alakeel^*1, M. Fouda¹, A. Alshahrani¹, M. Almofada¹

69.1.1 ¹Vigilance and Benefit–Risk Assessment Executive Directorate, Saudi Food and Drug Authority, Riyadh, Saudi Arabia

Background/Introduction:

Like for all medications, antibiotics are not free of adverse drug reactions (ADRs). Fixed drug eruption (FDE) is a serious special type of cutaneous reactions [1]. As a disease condition, FDE is considered a type of reactive erythema (erythema-multiform-like) [2]. Characteristically, the reaction recurs in the same site on skin/mucosa each time the offending drug is administered (hence called “fixed”). Doxycycline is a broad-spectrum bacteriostatic antibiotic belongs to tetracycline class. It is active against a wide range of microorganisms.

The Saudi Food and Drug Authority (SFDA) has initiated a safety review based on a published case report of potential association between doxycycline and the risk of FDE [1]. A 37-year-old female presented to the hospital with a history of fluid filled lesions all over her body following single dose of doxycycline. Her previous medical record shows rashes over the trunk due to former doxycycline intake as well.

Objective/Aim:

The purpose of this review is to evaluate the risk of FDE associated with doxycycline use and if required, to suggest regulatory recommendations.

Methods:

SFDA performed a search in the national ADRs database, World Health Organization (WHO) global ADRs database along with literature screening to retrieve all related information for the sake of investigating causality of drug-event combination.

Results:

The Saudi National Pharmacovigilance Center database showed two reports of doxycycline associated with FDE. One of the two cases has been excluded due to suspected duplication.

The WHO database revealed 364 international ADR reports. Of them, 356 cases were not sufficiently documented for proper medical assessment. The rest of the cases (i.e. eight cases), were reported a positive de-challenge and one case reported unknown re-challenge outcome.

During literature search, a published study evaluating the cross-reactivity of FDE with tetracycline antibiotics. The study included 16 patients who have FDE following tetracycline use. The patients challenged to doxycycline or minocycline at least 6 weeks after the last episode of FDE. Among these 16 patients, ten developed FDEs to doxycycline within 6 h after doxycycline intake [3].

Another published case-report of a 30-year-old male who has 4-day-history of erythematous patches. The patient reported that lesions developed 3 h following a single dose of doxycycline. The case reported a positive de-challenge and re-challenge [4].

Conclusion:

The SFDA investigations concluded that the current available evidence supports probable relationship between doxycycline and FDE. Therefore, the reference safety information is updated and a safety communication released.

References:

• 1. Nitya S, Deepa K, Mangaiarkkarasi A, Karthikeyan K. Doxycycline induced generalized bullous fixed drug eruption—A case report. J Young Pharm 2013; 5:195–6

• 2. Cohen BA. (2013). Reactive Erythema. In Pediatric Dermatology: Fourth Edition. Elsevier Inc.

• 3. Tham SN, Kwok YK, Chan HL. Cross-reactivity in fixed drug eruptions to tetracyclines. Arch Dermatol 1996; 132: 1134–5

• 4. Podder I, Chandra S, Das A,Gharami RC. Doxycycline induced generalized bullous fixed drug eruption. Indian J Dermatol 2016; 61:128

Disclosure of Interest: None declared.

70 ISoP18-1094 Expectedness of Antimicrobials Adverse Drug Reactions in Iraqi, 2010–2017

70.1 M. M. Younus^*1, A. AL-Charakh²

70.1.1 ¹Iraqi Pharmacovigilance Center, MOH, Baghdad, Iraq; ²Faculty of pharmacy, University of Kufa, Najaf, Iraq

Background/Introduction: Antimicrobials are among the most widely used medicines worldwide to treat or prevent different types of infections and it is not always true that their use is free from problems; in fact many studies showed different non-expected or expected adverse drug reactions (ADRs) events accompanying their use. The expectedness or Predictability of these ADRs vary widely depending on the type of antimicrobial agent, patients factors, concomitant medicines, underlying co-morbidities and many other reasons.

Objective/Aim: Analyze Iraqi pharmacovigilance center (IPC) database regarding antimicrobial ADRs expectedness. This is the first published data regarding the antimicrobials ADRs from the IPC database.

Methods: This is a retrospective study, where all antimicrobial ADRs, spontaneously reported to the IPC database using the national individual case safety report (ICSR) form, during the period 2010–1 Nov 2017 were analyzed. Data regarding patient demography, antimicrobial group (ATC code J01), and type of ADRs classified according to MedDRA by system-organ classes (SOC) and preferred terms (PT) were collected.

The expectedness was evaluated through the review of the summary of product characteristics of all the reported antimicrobials depending on the EMA definition of expectedness. The relationship between expectedness and patients characteristics (age, sex) was studied. Also, the seriousness of the reports was evaluated according to EMA criteria. The relationship between expectedness and seriousness was studied.

Statistical Package for the Social sciences (SPSS) version 17 was used to analyze the results. The data was expressed in (%). Chi square test was applied and p value < 0.05 was considered to be significant.

Results: The total number of reports included in this study were 896 reports, 45 (5.0%) of them were unexpected. Age group (2–11 years) showed 6.2% unexpected ADRs, followed by (12–17 years) with 5.9 and 5.1% in both (18–44 years) and (65+ years). Males represented 45.2% of the total number of patients and 56.9% of the unexpected ADRs were reported in males.

The most predominant SOC for antimicrobials involved in this study were; immune system disorders (347 adverse effects), gastrointestinal disorders (220 adverse effects) and skin and subcutaneous disorders (177 adverse effects). On the other hand skin and subcutaneous disorders with (25 adverse effects) was the most predominant of the unexpected adverse effects of the antimicrobials while gastrointestinal disorders showed (15 adverse effects) and the immune system disorders (11 adverse effects) was in the third place.

Antimicrobial groups most frequently involved with unexpectedness were; sulfonamides 16.7%, β-lactam 7.8%, quinolone (6.4%). Other β-lactam antibacterials (5.7%). Unexpected, serious reports represent 57.7% (26 reports) of the total unexpected reports and the results were significant at p < 0.05.

Conclusion: In this study, there were (5%) 45/896 unexpected ADRs among antimicrobials reported to the IPC. Different antimicrobial groups were involved, mostly sulfonamides, β-lactam and quinolone. Varying MedDRA terminology were reported; skin and subcutaneous, gastrointestinal and immune system disorders were the most predominant terminologies. Seriousness compared with expectedness showed significant results.

Disclosure of Interest: None declared.

71 ISoP18-1096 Snapshot of the Overprescription of Proton Pump Inhibitors in a Primary Care Hospital

71.1 C. Lenoir^*1, M. El Biali¹, C. Luthy², O. Grosgurin³, J. Desmeules¹, V. Rollason¹

71.1.1 ¹Division of Clinical Pharmacology and Toxicology,; ²Division of General Medical Rehabilitation, Geneva University Hospitals & University of Geneva, Switzerland; ³Division of Internal Medicine, Geneva University Hospitals, Geneva, Switzerland

Background/Introduction: Proton pump inhibitors (PPIs) are among the most widely prescribed drugs in the world, but more than half of the indications for prescription are unjustified [1, 2]. The misuse of this therapeutic class has heavy consequences such as additional health costs, adverse drug reactions following long-term use and gastric acid rebound when the PPI is discontinued [3–5]. The overprescription of PPIs is therefore becoming a public health problem, which led us to evaluate its use within the Geneva University Hospitals (HUG).

Objective/Aim: To characterize the PPI prescription within the HUG and in particular to determine if the PPI prescription is made according to the market authorization and known guidelines.

Methods: This prospective observational study collected data by consulting the electronic records of patients hospitalized in two divisions of the department of internal medicine of the HUG (division of general internal medicine and division of general medical rehabilitation) on a single day (instant photography) using the computerized patient record system of HUG (DPI^®).

Results: 180 (95 men and 85 women) patients were included. 54% (N = 97) of patients were on PPIs, 29% (N = 28) of whom had their treatment initiated at HUG. For patients for whom home treatment was known 46% (N = 78) of patients were already treated at home. Regardless of dosage, 72% (N = 70) of the indications for treatment were not justified, 24% (N = 23) of the indications were justified and 4% (N = 4) were eventually justified. Regarding dose, 56% of the justified indications and 75% of the possibly justified indications did not have an adequate dosage. Therefore, in all patients with a PPI at HUG, only 11% had a justified and possibly justified indication with an adequate dose. Esomeprazole was the most prescribed PPI with a rate of 97% (N = 94) at HUG and 70% (N = 54) at home. Among them, respectively 50% (N = 47) and 60% (N = 32) were prescribed 40 mg/d. Finally, 87% (N = 68) of known home prescriptions were renewed on admission and among them, 71% (N = 40) did not have a justified or possibly justified indication according to the guidelines.

Conclusion: Our study highlights the multifactorial problem surrounding the prescription of PPIs. Indication for treatment inside the hospital was not justified in 72% of patients. Moreover when taking also dose into account, only 11% of our patient had a justified indication with a correct dose. Precise guidelines with evidence-based indications and adequate daily doses would help the prescribers, in the hospital and in the community, to adequately prescribe PPIs. Moreover, patients should benefit from a thorough evaluation of their treatment at admission and on discharge.

References:

• 1. Sukhovershin RA, Cooke JP. How may proton pump inhibitors impair cardiovascular health? Am J Cardiovasc Drugs 2016;16:153-61

• 2. Rotman SR, Bishop TF. Proton pump inhibitor use in the U.S. ambulatory setting, 2002–2009. PLoS One 2013;8:e56060

• 3. Martin RM, Dunn NR, Freemantle S, Shakir S. The rates of common adverse events reported during treatment with proton pump inhibitors used in general practice in England: cohort studies. Br J Clin Pharmacol 2000;50:366-72

• 4. Nardino RJ, Vender RJ, Herbert PN. Overuse of acid-suppressive therapy in hospitalized patients. Am J Gastroenterol 2000;95:3118-22

• 5. West DS, Johnson JT, Hong SH. A 30-month evaluation of the effects on the cost and utilization of proton pump inhibitors from adding omeprazole OTC to drug benefit coverage in a state employee health plan. J Manag Care Pharm 2006;12:25-32

Disclosure of Interest: None declared.

72 ISoP18-1097 High Dose Methotrexate Delayed Elimination: A Case Series

72.1 B. Azzouz¹, C. Feliu², Q. Jarrion¹, L. Aubert¹, A. Morel¹, Z. Djerada², T. Trenque^*1

72.1.1 ¹Regional Centre of Pharmacovigilance and Pharmacoepidemiology,; ²Department of Pharmacology, Reims University Hospital, Reims, France

Background/Introduction: High dose methotrexate (HDMTX, defined as > 1 g/m²) is used to treat a variety of adult and pediatric cancers, including acute lymphoblastic leukemia, osteosarcoma and lymphoma. More than 90% of MTX is eliminated by the kidneys. Hyperhydratation, urine alkalinization, leucovorin, close monitoring of plasma MTX concentrations and kidney function, drug–drug interaction analysis are mandatory to prevent delayed MTX elimination.

Objective/Aim: To raise healthcare professionals awareness about HDMTX delayed elimination and its toxicity.

Methods: We reviewed spontaneous reports of delayed HDMTX excretion notified to the Pharmacovigilance Center of Reims within the last 6 months (December 2017 to May 2018).

Results: Despite adequate preventive measures, 5 spontaneous reports of delayed HDMTX, within the last 6 months, were notified. Mean age was 57.2 ± 9.5 years (range 44–67 years), patients were exclusively males. Indications of HDMTX were: B cell lymphoma, Burkitt’s lymphoma, blastic plasmacytoid dendritic cell neoplasia and cerebral lymphoma. MTX dose was 3 g/m² for all cases, infusion durations were: 3 h (n = 3), 24 h (n = 1) and 30 min (n = 1). Delayed MTX elimination occured at the second cycle in 4 cases. Mean CKD-EPI before the MTX regimen was: 71.6 ± 13.5 mL/min (range 59–90 mL/min). Mean serum MTX concentration at H24 (C_24H) was: 50.6 ± 37.8 µmol/L (range 20.4–114.2 µmol/L). Mean CKD-EPI at H24 was: 28.4 ± 3.5 mL/min (range 25–34 mL/min). Toxicities observed were: nephrotoxicity (renal tubular necrosis, acute kidney injury), hepatotoxicity (acute liver injury), myelosuppression (febrile aplasia, febrile neutropenia), neurotoxicity (encephalopathy), respiratory distress, epidermolysis and mucositis. Leucovorin rescue, hyperhydratation and urine alkalinization were performed for all patients. Glucarpidase was introduced in 4 cases. Despite the glucarpidase therapy, a septic shock with fatal outcome was observed in 2 cases (C_24H were: 53.2 and 114.2 µmol/L, fatal outcome occurred 9 days after MTX infusion).

Conclusion: HDMTX delayed elimination can lead to serious and life-threatening toxicities. The same MTX dose, 3 g/m², was given for patients with different kidney function levels. MTX dose should be adjusted to the kidney function. Protocol with MTX 3 g/m² during 30 min (R-MBVP) should be updated, a slower MTX infusion would be prudent. Serum MTX concentration levels should be measured at earlier stages. Serum MTX concentration assessed before 24 h would enable an early detection of HDMTX delayed elimination. Health professionals awareness is needed even if the first MTX cycle went with a good tolerance.

Disclosure of Interest: None declared.

73 ISoP18-1098 Awareness Survey on Drug–Drug Interactions Among Argentinean Health Care Professionals

73.1 G. Darderes^*1, C. Brown², J. C. Gallo³, L. D´Alessio⁴, E. Roldan⁵

73.1.1 ¹Head of Pharmacovigilance, ² GADOR S.A;, ³; ⁴ENyS-CONICET, Buenos Aires, Argentina; ⁵Scientific Direction, GADOR S.A., Buenos Aires, Argentina

Background/Introduction: Since 1994, Argentinean National Regulatory Health Authority (ANMAT) is an official member of the WHO Program for International Drug Monitoring. However, the number of adverse reactions related to drug–drug interactions reported is still very low. As surveys developed with healthcare professionals (HCP) in other countries showed that the knowledge and awareness about drug–drug interactions, and the number of reports increase, we develop a similar experience in our country.

Objective/Aim: The aim of this survey was to assess the level of knowledge about drug–drug interactions and the activity of reporting adverse events related to drug–drug interactions among local HCP. Likely, the survey will be used as an awareness tool. We hereby report on the outcomes.

Methods: A cross-sectional on-line survey was designed, using a validated self-administered questionnaire form with a set of 10 questions and their correspondent multiple choice options for answering. The questionnaires were distributed to 20,653 HCP from an own database, including medical doctors, dentists and pharmacists. The questionnaire was related to adverse events linked to drug–drug interactions and the subsequent reporting behavior of HCP in ordinary practice. Questions were chosen based on daily practice situations. Data was collected and clustered by the type of HCP, analyzed by descriptive statistic, and Chi square test and z-scores corrected with Bonferroni Test for multiple comparisons were determined using a PC software.

Results: A total of 583 HCP, answered the survey; 354 (69%) were medical doctors, 182 (63%) dentists, and 47 (43.5%) pharmacists. Most part of medical doctors 61.1% prescribe and or discharge at least one drug in more than a half of their patients, 40.4% among pharmacies but only 20.9% among dentists (p < 0.05). Inquired about a clinical intercurrence situation, the most common response was to maintain the basal treatment and only change the temporal treatment, agreeing 52.0% of medical doctors, 41.2% of dentists, and 36.2% of pharmacists (p > 0.05), and the rest react differently. Regarding the source of getting update information about adverse reactions, 30.8% of medical doctors, 28.6% of dentists and 21.3% of pharmacies, similarly answered they prefer to use electronic media (p > 0.05), and the remaining choosing other options. During a situation of a drug interaction, medical doctors choose more frequently the option about revising the medical bibliography, 59.6%, comparing dentists, 30.8% and pharmacies, 31.9% (p < 0.05).

Conclusion: In this preliminary analysis it is shown that awareness is diverse and reactions to situations from practice are variable. Hence, apart from promoting awareness between HCP on their reporting role of adverse reactions related to drug–drug interactions, there is a need of guiding health professionals on good practice reporting AE related to drug–drug interactions.

References:

• 1. Metta H, Floor-Schreudering A, Wouters H, De Smet PAGM, Bouvy ML. Preferences of patients and pharmacists with regard to the management of drug–drug interactions: a choice-based conjoint analysis. Drug Saf 2018; 41: 179–89

• 2. Floor-Schreudering A, Geerts AF, Aronson JK, Bouvy ML, Ferner RE, De Smet PA. Checklist for standardized reporting of drug–drug interaction management guidelines. Eur J Clin Pharmacol. 2014;70:313–8

• 3. Pollard S, Bansback N, Bryan S. Physician attitudes toward shared decision making: a systematic review. Patient Educ Couns 2015;98:1046–57

• 4. -Guyatt GH, Oxman AD, Kunz R, Falck-Ytter Y, Vist GE, Liberati A, et al. Going from evidence to recommendations. BMJ. 2008;336:1049–51

Disclosure of Interest: G. Darderes Grant/Research support from: Gador support the survey, C. Brown Grant/Research support from: Gador support the survey, J. C. Gallo: None Declared, L. D´Alessio: None Declared, E. Roldan Grant/Research support from: Gador support the survey.

74 ISoP18-1101 The Degree of Impact of Adverse Drug Reactions as Experienced by Patients

74.1 M. Haaksman^*1, L. Rolfes¹

74.1.1 ¹Netherlands Pharmacovigilance Centre Lareb, ‘s-Hertogenbosch, The Netherlands

Background/Introduction: Little is known about the degree by which patients experience the impact of adverse drug reactions (ADRs).

Objective/Aim: To explore the degree by which patients experience the impact of ADRs that they reported to the Netherlands Pharmacovigilance Centre Lareb. The secondary objective was to explore factors correlated to the impact-score.

Methods: A retrospective observational study about the degree of impact of ADRs. The patient ADR reporting form of the Netherlands Pharmacovigilance Centre Lareb contains a question about the degree of impact of the ADR. The answer option is a 5-point scale, ranging from ‘not at all’ (score 1) to ‘very much’ (score 5). All drug-ADR associations, reported by patients between 10 July 2017 and 5 February 2018 were included.

Impact-scores were analyzed for all included ADRs on Preferred Term (PT) level, with serious and non-serious ADRs being analyzed separately. In-depth analysis on correlated factors was done for gender (men/women), age (< 65/65 or older), outcome of the ADR (recovered/not recovered), and action taken on the drug (withdrawn/dose not changed). Data were analyzed using Pearson Chi²-test. Statistical significance was based on p < 0.05.

Results: In total 1.015 ADR reports, including 2.718 drug-ADR associations were included, of which 44 concerning serious ADRs and 971 non-serious ADRs. The average impact-score for serious ADRs was 4.8 and for non-serious ADRs 3.7 (Table 1). This difference in impact-score was statistically significant (p value < 0.001). Due to this difference and the low number of serious ADRs in our output, the in-depth analysis included only non-serious ADRs.

ADRs concerning anxiety, suicidal ideation and pain had the highest average impact-score (4.8, 4.8 and 4.6, respectively) and injection site reactions, dry eyes and drug ineffective the lowest (2.9, 3.0 and 3.0, respectively).

There was no significant difference for the impact-score between men and women (p value 0.169). Patients over 65 years experienced a higher impact compared to those under age 65 (p value < 0.001). Patients experienced a higher impact for ADRs that recovered compared to ADRs that were not recovered (p value < 0.001). The impact-score was furthermore higher for patients by whom the drug was withdrawn compared to those by whom the dose of the drug was not changed (p value < 0.001).

Conclusion: This study demonstrated the degree by which patients experience the impact of ADRs. In addition, it indicated factors that correlated the impact-score. This knowledge is useful for healthcare professionals and patients to better understand the meaning of experiencing ADRs.

Disclosure of Interest: None declared.

75 ISoP18-1102 Automatic Detection and Removal of Personal Identifiers on Case Narratives Using Deep Learning

75.1 E.-L. Meldau¹, S. Hedfors Vidlin¹, L. Gattepaille¹, H. Taavola¹, L. Sandberg^*1, Y. Aoki¹, G. N. Norén¹

75.1.1 ¹Uppsala Monitoring Centre, Uppsala, Sweden

Background/Introduction: Removing personal information (de-identification) from free-text narratives in individual case safety reports is important to preserve patient privacy. It is, however, a demanding task for human annotators [1] as well as for any automatic system that would have to face the complex challenge of deciphering natural language. The case reports provide information about the suspected adverse drug reactions experienced by patients through both structured fields and free-text narrative descriptions. Both have been found to have an impact on the causality and clinical assessment [2]. The narrative can for example provide information about the severity of the reaction or the impact on the patient’s quality of life which may not be easily conveyed in structured fields. Automatic de-identification of narratives would allow sharing these narratives between organizations to ultimately improve patient safety while preserving patient privacy.

Objective/Aim: To develop an automatic de-identification method for narratives. In this abstract, we present interim results from ongoing research.

Methods: Our de-identification method combines the prediction of several models through an ensemble method using logistic regression. Each token in the narrative is labelled “personal” or “non-personal” where names, dates and locations are considered personal identifiers for the purpose of this study. The models used are different deep neural networks, a logistic regression, a part-of-speech tagger, and rule-based methods. The neural networks and logistic regression are particularly flexible in the sense that they can be retrained on new data with different properties such as other languages, writing styles or formats. This training requires large amounts of annotated data, but at this point of our research, there is a shortage in annotated narratives. We therefore used the training data from the 2014 i2b2 de-identification challenge [3], containing 521 medical records. A validation set of 269 i2b2 records was divided; 75% were used for training the ensemble method and 25% for evaluating the method. The method was also evaluated on 300 narratives from reports in VigiBase, the World Health Organization’s global database of individual case safety reports.

Results: 95% of the personal identifiers were detected (recall), while 90% of the detected parts were actually personal identifiers (precision). For the categories names and dates the recall was higher than 95% (98 and 96%, respectively) and for locations lower (86%). From the VigiBase narratives, 90% of the personal identifiers were recalled with a precision of 45%. Parts of the annotated personal identifiers that were missed by the algorithm, for example of in “4th of May”, would not compromise the confidentiality but do decrease the recall, suggesting that the method preserves privacy better than the recall indicates.

Conclusion: Our recall on i2b2 data exceeds human recall for all three categories, while humans are more precise in their detection [1]. The high recall and intermediate precision on narratives is promising, considering that our method was only trained on medical records but not yet on the target data. With access to original narratives that can be annotated for further training and fine-tuning, we expect the performance to improve even further.

References:

• 1. South BR, Mowery D, Suo Y, Leng J, Ferrández Ó, Meystre SM, et al. Evaluating the effects of machine pre-annotation and an interactive annotation interface on manual de-identification of clinical text. J Biomed Inform 2014; 50: 162–72

• 2. Karimi G, Star K, Lindquist M, Edwards IR. Clinical stories are necessary for drug safety. Clin Med 2014; 14: 326–7

• 3. Stubbs A, Uzuner Ö. Annotating longitudinal clinical narratives for de-identification: The 2014 i2b2/UTHealth corpus. J Biomed Inform 2015; 58(Suppl): S20–9

Disclosure of Interest: None declared.

76 ISoP18-1103 Suicidal and Self-Injurious Behaviour Under Desogestrel: The Patient’s View

76.1 S. Watson^*1, L. Härmark², P. Caduff- Janosa¹

76.1.1 ¹Uppsala Monitoring Centre, Uppsala, Sweden, ²Lareb, ‘s’Hertogenbosch, The Netherlands

Background/Introduction: Altered and depressed mood are listed adverse drug reactions (ADRs) for the contraceptive drug desogestrel. More severe reactions like suicidal ideation and self-injurious behaviour are not^1,2. These ADRs were highlighted as potential signals for desogestrel in a joint signal detection workshop between Uppsala Monitoring Centre and the Netherlands Pharmacovigilance Centre Lareb, focusing on patient reports using the WHO global database of individual case safety reports, VigiBase.

Objective/Aim: To explore the evidence in VigiBase for the causal association between desogestrel and suicidal and self-injurious behaviour.

Methods: Case-by-case assessment of relevant reports in VigiBase up until October 2016.

Results: At the time of the workshop, October 2016, there were 34 reports for the MedDRA High level term Suicidal and self-injurious behaviour. Desogestrel was the only suspected drug in 27 reports. In seven reports there were more than one suspect drug; other contraceptives in two cases, topiramate in one, (es)citalopram in two and in two cases multiple drugs in two of the suicide attempts. In 19 reports desogestrel was the only reported drug and no concomitant drugs were reported. The time to onset (where this information was recorded) ranged from 1 day to 5 months. In 20 cases the patients recovered after stopping the drug and one patient mentions a rechallenge, although it is not clear if the same contraceptive drug was used.

Two patients had a medical history of depression or depressive episodes but both recovered after discontinuing desogestrel.

Conclusion: Although mood changes and depressed mood are listed as ‘common’ in the product label and patient information leaflets of desogestrel, the more severe reactions suicidal ideation and self-injurious behaviour are not mentioned at all. A literature review revealed no published case reports of desogestrel in relation to these events. Severe depressions can lead to self-harm or even suicide, which is what the reports in VigiBase describe. Although in some cases there might be confounding factors present, in 20 cases there was a positive dechallenge which strengthens the relationship between the drug and the events. These reports send a signal to both women using contraceptives and prescribers to be vigilant of the severity of the depressions that may occur while on this contraceptive. Further investigations to identify the women at risk for psychiatric disorders as ADRs for hormonal contraceptives in general are needed.

Note! In May 2018 the number of reports has increased to 45.

References:

• 1. Electronic Medicines Compendium: Summary of Product Characteristics for desogestrel (Cerazette^®) Available from https://www.medicines.org.uk/emc/medicine/10098 Accessed: May 2018

• 2. Electronic Medicines Compendium: Patient Information Leaflet for desogestrel (Cerazette^®). Available from: https://www.medicines.org.uk/emc/product/1698/pil Accessed: May 2018

Disclosure of Interest: None declared.

77 ISoP18-1104 Gender Differences in Adverse Drug Reaction Reports; How can they be Explained?

77.1 S. de Vries¹, E. van Puijenbroek^*2,3, P. Mol¹, C. Ekhart², P. Denig¹

77.1.1 ¹Dept. of Clinical Pharmacy and Pharmacology, University Medical Center Groningen, University of Groningen, Groningen, the Netherlands; ²Pharmacovigilance Centre Lareb, ‘s-Hertogenbosch, the Netherlands; ³Groningen Research Institute of Pharmacy, Pharmacotherapy, Epidemiology & -Economics, University of Groningen, Groningen, the Netherlands

Background/Introduction: In general, women have a higher risk for adverse drug reactions (ADRs) being reported than men [1]. Women also generally use more drugs as compared to men in various therapeutic areas [2]. It is unclear which differences in reporting pattern for both genders exist after correcting for drug use. Knowledge of these differences, and possible mechanisms that may explain them, may help in optimizing gender-specific pharmacotherapy.

Objective/Aim: To assess and characterize gender differences in reported drug-ADR combinations to the national pharmacovigilance centre in the Netherlands, corrected for differences in drug use.

Methods: We included drug–ADR combinations reported by healthcare professionals and patients to the Netherlands Pharmacovigilance Centre Lareb. Only combinations (ATC-MedDRA preferred term) that were reported at least 10 times between 2003 and 2016 for drugs that were used by ≥ 10,000 persons in that period were taken into account. Gender-specific ADRs, like gynaecological problems, were excluded. Data from the National Health Care Institute were used to correct for the number of users per drug. Gender differences in specific drug–ADR combinations were tested using bivariate logistic regression analyses, adjusted for the number of drug users per gender.

Results: A total number of 2482 drug–ADR combinations were analysed. Statistically significant differences (p < 0.05) were shown in 363 combinations (15%). Most of these drug–ADR combinations were reported more often for women (321 combinations, 88%). Drugs with the highest number of ADRs that were more often reported for women included thyroid hormones, and antidepressants. Some ADRs were predominantly reported for women across a range of drugs like headache and dizziness whereas other ADRs like tendon ruptures and aggression were reported more often for men.

Conclusion: Identified gender differences in reporting ADRs often referred to women. These differences may have various causes. Men and women differ in respect to their physiology, immunological processes and type and function of various hormones. This may result in differences in pharmacokinetics and pharmacodynamics. Examples are the predominance of dose depending ADRs being reported in women which may depend on pharmacokinetic properties and the predominance of ADRs associated with psychotropic drugs which may be caused by differences in pharmacodynamics. Gender differences may also result of variations in perception by reporter and patient since they are experienced as bothersome depending on a gender-specific self-image. An example is the occurrence of hair thinning and hair loss, which is more common in aging men but is likely to be considered as disturbing particularly when it occurs in women. In addition, differences in background incidence in disorders and diseases between both genders exist and may be reflected in reporting ADRs. Finally, gender differences in reporting may result from variations in reporting behaviour of the reporter, either healthcare professional or patient. Further studies are needed to assess underlying reasons, including pharmacological and behavioural factors. The results may ultimately lead to gender-specific prescribing or monitoring recommendations.

References:

• 1. Rademaker M. Do women have more adverse drug reactions? Am J Clin Dermatol 2001;2:349–5

• 2. Loikas et al., Differences in drug utilisation between men and women: a cross-sectional analysis of all dispensed drugs in Sweden. BMJ Open 2013;3:e002378

Disclosure of Interest: None declared.

78 ISoP18-1110 Who’s At Risk? Identifying Risk Groups for Adverse Drug Reactions Using Vigibase

78.1 Y. Aoki¹, L. Sandberg^*1, H. Taavola¹, R. Chandler¹, G. N. Norén¹

78.1.1 ¹Uppsala Monitoring Centre, Uppsala, Sweden

Background/Introduction: Early identification of previously unknown causal associations between medicines and adverse events has been the primary focus of pharmacovigilance. In recent years, Uppsala Monitoring Centre has initiated a shift toward signal characterisation and risk group identification in support of our vision for wise therapeutic decisions. As a first attempt at broader open-ended risk group detection, we conducted a signal screening focused on identifying risk groups for adverse drug reactions (ADR).

Objective/Aim: To explore the possibility of identifying risk groups for ADRs using VigiBase, the World Health Organization’s global database of individual case safety reports.

Methods: We considered 15.4 million reports received up until August 2017 in VigiBase. The following covariates were explored as potential risk factors for ADRs: age, sex, body mass index, pregnancy status, underlying disease, and reporting country/continent.

Disproportionality analysis was performed for all drug-adverse event (AE) pairs (1) in the entire database, and (2) across a range of data subsets defined by the selected covariates. Drug-AE pairs disproportionally reported in at least one of the subsets but not in the database as a whole were considered for initial manual assessment. To reduce the chance of highlighting spurious associations, higher confidence levels were used in the subset analyses. Further, a drug–AE-subset association was required to have at least one report received in 2012 or later and the observed-to-expected ratio in the subset needed to be at least twice that of the corresponding value in the database as a whole.

The highlighted associations were prioritised for initial manual assessment using vigiRank [1], while ensuring that a balanced number of drug–AE pairs for all covariates were assessed. The assessments were performed by physicians, pharmacists and data scientists working in teams. During this initial manual assessment, the association was determined to be a potential signal if the risk factor was not communicated in the label, there was no obvious confounder, and the reports for that association were informative enough to conduct the in-depth review. The selected potential signals were subject to in-depth clinical reviews by the domain experts and the biological evidence for the risk factor for the ADR was further investigated.

Results: The initial manual assessment covered 386 associations, of which 18 were deemed to be potential signals. A breakdown of the potential signals by covariate is tabulated below. The top 3 risk groups by the number of potential signals were females (5), underweight patients (3) and the elderly (3).

As of May 2018, in-depth clinical reviews have been completed for 14 out of 18 potential signals, resulting in seven signals describing potential risk groups for ADRs (sex: 2, age: 2, body mass index: 2, continent: 1).

Conclusion: Signals of ADRs in risk groups can be identified from a global database using subset disproportionality analysis. Continued development of statistical screening methodologies to highlight potential signals within subgroups could usher in a new era of “precision pharmacovigilance”.

References:

• 1. Caster O, Juhlin K, Watson S, Norén GN. Improved statistical signal detection in pharmacovigilance by combining multiple strength-of-evidence aspects in vigiRank. Drug Saf 2014;37:617–28

Disclosure of Interest: None declared.

79 ISoP18-1117 Pvknow: A Pharmacovigilance Knowledge Management System

79.1 J. Boer^*1, L. Härmark¹

79.1.1 ¹Netherlands Pharmacovigilance Centre Lareb, ‘s-Hertogenbosch, The Netherlands

Background/Introduction: During the daily pharmacovigilance processes at the Netherlands Pharmacovigilance Centre Lareb, multiple information sources, located in different places, are consulted by the individual assessors. The fragmentation of information leads to process inefficiencies, and as all assessors might not know the existence and location of certain information, it might lead to sub-optimal quality of the output of the processes. Clustering all the relevant information in one place would improve the quality and efficacy of internal processes.

Objective/Aim: To develop a knowledge management system in which all relevant knowledge can be stored centrally, to support the internal processes.

Methods: An internal working group was formed, to identify all the functional requirements for the system. This group consisted of Lareb staff that are involved in the knowledge intensive processes within the organisation and a Business Analyst. The system was developed using Agile methodology. This means that the building process starts before the plans are finalised, which increases flexibility. In subsequent sessions within this process additional adjustments could be made based on newly gained insights [1].

Results: The knowledge management system, called PVknow, was designed as a modular application where the core consists of knowledge pages. These pages contain the relevant and updated information about a particular drug–adverse drug reaction (ADR) association. The knowledge pages can also be shared with the public through a connection to the website. Key in designing this functionality was a search function with which you can search for a specific drug, ADR or drug–ADR combination. Other modules in PVknow are a helpdesk function, a signal management tool and a library.

Conclusion: PVknow is the result of the idea to create a knowledge management system to store pharmacovigilance related information centrally. By involving Lareb staff with in depth knowledge about the scientific processes it was ensured that the developed system would fit for purpose. By developing in an Agile way, it was possible to have maximal flexibility in adapting the system to the current needs. The experiences so far are positive; the information is well organized and easily available for assessors and thereby supporting the daily activities at Lareb. Although it was developed with the primary aim to support internal processes, it will also enable Lareb to share the content of PVknow with external partners through our own website or partner websites.

References:

• 1. Hoogendoorn S. Dit is agile. Van introductie tot implementatie. Amsterdam. Pearson Benelux. 2012.

Disclosure of Interest: None declared.

80 ISoP18-1118 Risk of Diabetes Associated with Statins Fibrates and their Association

80.1 F. Montastruc¹, J. Benevent¹, V. Rousseau¹, G. Durrieu¹, A. Sommet¹, J.-L. Montastruc^*1

80.1.1 ¹Medical and Clinical Pharmacology, Faculté de Médecine, Toulouse, France

Background/Introduction: Relations between use of hypolipidemic drugs and occurrence of diabetes are controversial.

Objective/Aim: The aim of our study was to assess the association between diabetes and use of statins, fibrates and their association using the World Health Organization (WHO) global individual case safety reports database, VigiBase^®, which records reports from over 130 countries worldwide.

Methods: We performed a disproportionality analysis including all reports until the 31st December 2017 to measure the risk of reporting “new onset of diabetes” compared with all other reports [as a reporting odds ratio (ROR 95% CI)] for statins, fibrates and their association. The likelihood that diabetes resulted from statin–fibrate pharmacodynamic interaction was also estimated. According to the interaction additive model, a ROR value for coexposure exceeding the sum of the RORs estimated for each individual class of drug supports a potential drug–drug interaction (DDI). To assess the stability of our results, we performed several sensitivity analyses, according to outcome definition and excluding reports with drugs known to induce hyperglycaemia or diabetes (putative competitors).

Results: Among the 177,323 reports with statins (without fibrates), 7633 were diabetes (mainly with atorvastatin). Among the 19,149 reports with fibrates (without statins), 189 were diabetes (mainly with fenofibrate). Sixty other diabetes reports involved the combination statin+fibrates. Statin use alone was associated with an increase in the ROR of diabetes [ROR = 1.75 (1.72–1.78)], but not with fibrate use alone [ROR = 0.76 (0.71–0.82)]. The combination statin + fibrate was also associated with an increase in the ROR of diabetes [ROR = 1.46 (1.28–1.67)]. Similar results were found in sensitivity analyses, using different outcome definitions and after exclusion of putative competitors.

Conclusion: Using the WHO pharmacovigilance database, our study strengthens the signal of diabetes with statins but failed to find any potential signal with the use of fibrates alone. Furthermore, our data did not suggest a pharmacodynamic DDI with the use of statin + fibrate.

Disclosure of Interest: None declared.

81 ISoP18-1119 Fluoroquinolones and Aortic Aneurysms: What are the Fluoroquinolones at Risk?

81.1 A. Sommet¹, V. Rousseau¹, L. Chebane¹, G. Durrieu¹, F. Montastruc¹, J.-L. Montastruc^*1

81.1.1 ¹Medical and Clinical Pharmacology, Faculte de Medecine, Toulouse, France

Background/Introduction: Association between fluoroquinolone use and risk of aortic aneurysm and dissection has been previously suggested from observational data (2 cohort studies and 1 case–control analysis).

Objective/Aim: The present study was performed to investigate this putative association in Vigibase^®, the World Health Organization Global Individual Case Safety Reports (ICSRs) database. We also studied the fluoroquinolones most at risk.

Methods: We performed a case non-case study where cases were ICSRs of “aortic aneurysms and dissections” (HLGT) with fluoroquinolones (J01MA) registered until the 31th December 2017 as “suspected” or “concomitant” in patients ≥ 50 years. In order to reduce possible confounding, comparison was made with amoxicillin. Similar analyses were made for “tendonitis”, “tendon injury, pain and rupture” (PT terms), a well-known adverse drug reaction of fluoroquinolones, comparing each fluoroquinolone versus all others. Only fluoroquinolones with more than 3 reports were included. Results are shown as Reporting Odds Ratios (ROR) of the exposure odds among reported cases of aortic aneurysms to the exposure odds among reported non-cases.

Results: Among the 172,588 ICSRs with fluoroquinolones, 113 were aortic aneurysms. Most of the patients were females (54.9%) with a mean age of 67.6 ± 11.1 years. Aortic aneurysms were reported with levofloxacin (n = 78), ciprofloxacin (26), moxifloxacin (15). The first ICSR was reported in 2000. There were 8 aortic aneurysms among the 40,954 ICSRs with amoxicillin. The ROR value for fluoroquinolones versus amoxicillin was 3.13 (1.53–6.43) after adjustment on age and sex. Compared with use of other fluoroquinolones, use of levofloxacin was associated with a significant increase in the ROR of aortic aneurysms [RORa = 4.44 (2.94–6.72)] and also of tendinopathy [RORa = 3.02 (2.89–3.15)].

Conclusion: The present study strengthens the signal of aortic aneurysms with fluoroquinolones and shows for the first time that, among the different fluoroquinolones, the most at risk is levofloxacin. Moreover, the similarity of results concerning the most at risk fluoroquinolones could suggest the involvement of some common pharmacological mechanisms for aneurysms and tendinopathy with these drugs, like, for example, chelating properties and upregulation of metalloproteinases resulting in extracellular matrix alterations.

Disclosure of Interest: None declared.

82 ISoP18-1120 Pharmacovigilance: Analysis of Medication Errors Reports Received by Argentine Regulatory Authority (Anmat) During the Period 2013–2017

82.1 M. G. Papazian^*1, S. Schiaffino², R. Papale³, R. Heredia³

82.1.1 ¹Italian Hospital University of Buenos Aires, Buenos Aires, Argentina; ²School of Medicine, Buenos Aires University, Buenos Aires, Argentina; ³Pharmacovigilance, ANMAT, Buenos Aires, Argentina

Background/Introduction: Drug safety has undergone significant changes and proper drugs use are currently priorities for developed countries. For improving patient safety, it’s not only necessary to develop safer medicines, but also create error-proof use systems to minimize medication error (ME) and guarantee that if it occurs, it doesn´t reach and determine negative consequences on patients. Therefore, it´s essential to identify and have a close understanding of this topic. The detection of ME contributes to effective practices for a rational medicine´s use, thus increasing patient safety [1]. In Argentina this problem hasn´t been studied and it´s necessary to know the characteristics of ME, mainly studying the frequency and impact, in order to establish local strategies to prevent the occurrence. Argentine Health Authority—ANMAT [2]—has a National Pharmacovigilance System in which ME reports are considered as part of Medications Related Problems. Hence, a validated ME Notification Form [3] is available since 2013. The rate of ME reports since the implementation, has increased significantly. According to ANMAT´s Annual Reports [2], in 2012-prior to the GPV implementation there were just 210 reports received, increasing to 1299 in 2017.

Objective/Aim: Analyze the ME reporting in Argentina from 01 Jan 2013 to 31 Dec 2017 received by ANMAT.

Methods: The study is a transversal, retrospective and observational. It´s descriptive, since it´ll determine the frequency of the results in terms of ME reported to ANMAT. Subsequently, the frequency that could have occurred since the implementation of the Form[3] will be analyzed, from 2013 to 2017.

Results: 3090 ME reports were included. Each variable was analyzed according to the data available. Regarding ATC code, the number for each category has increased for all the classifications. The “L” has the highest ME reports with 34.9%, followed by “C” with 15.8%. 88.1% of the patients received the drug. 94.6% of the reports belong to the Pharma Industry and only 3.8% to others reporters (hospitals and universities). Female patient reports lead with 60.8%. Regarding age, 27.6% patients were adults, followed by the elderly. 75.8% of the reporters were HCP. According to the error category 76.3% were errors that reached the patient, but fortunately did not cause harm. 70.3% of ME occurred at home.51% of ME occurred during the administration phase, and 25.8% in prescription. Incorrect interval frequency was the main ME identified, 27.3%. From 2013 to 2017, there was a fourfold increase in the ME reports.

Conclusion: The analysis of ME provides the basic knowledge about them. Beside the real importance lies in investigating the behavior of them, in order to avoid future errors or to take measures that properly interposed in the system of use of drugs as security filters. Strengthen education on the safe use of medicines, both the patient and the treating health professional is an indispensable tool to achieve a more appropriate and safe use of them, the solution to the problem lies in prevention. It is essential to continue with policies to reduce and prevent morbidity related to medication errors in both outpatient and inpatient patients and to further disseminate the current ANMAT Medication Errors notification program. Encouraging this program by intensifying the analysis will contribute to the greater knowledge of this problem.

References:

• 1. Santos LD, Winkler N, Santos MAD, Martinbiancho JK. Descripción de los errores de medicación detectados en un centro de información sobre medicamentos en el sur de Brasil. Pharmacy Pract (Granada) [Internet]. 2015 [cited 2017 Jun 29];13(1):0–0. Available from: http://scielo.isciii.es/scielo.php?script=sci_arttext&pid=S1885-642X2015000100007&lang=pt

• 2. ANMAT. A.N.M.A.T [Internet]. [cited 2017 Jun 29]. Available from http://www.anmat.gov.ar/principal.asp

• 3. ANMAT. Ficha de Notificación de Errores de Medicación [Internet]. Available from: http://www.anmat.gov.ar/farmacovigilancia/notificaciones/FichaERRORES.pdf

Disclosure of Interest: M. G. Papazian Employee of: Boehringer Ingelheim, S. Schiaffino Employee of: Boehringer Ingelheim, R. Papale: None Declared, R. Heredia: None Declared.

83 ISoP18-1123 Drospirenone and Its Association with Gall Bladder Diseases: A Retrospective Study of Canadian Adverse Drug Reaction Monitoring Program, 1965–2017

83.1 M. Shah^*1, C. Kothari¹

83.1.1 ¹Institute of Pharmacy, Nirma University, Ahmedabad, India

Background/Introduction: Drospirenone, a fourth generation progestin combined with ethinyl estradiol is one of the most prescribed oral contraceptives in North America. In past, various concerns have been raised in media as well as published reports for the association of Drospirenone and Gallbladder diseases [1–4].

Objective/Aim: The objective of the study was to identify possible signal induced by drospirenone and its significance with gall bladder diseases by searching database from Canadian Adverse Reaction Monitoring Program (CADRMP).

Methods: A total of 9938 reports of patients from January 1965 to December 2017 were downloaded from CADRMP website. These reports contained information of adverse events associated with all other drugs inclusive of Drospirenone. Signal detection were determined by proportional reporting ratio (PRR), reporting odds ratio (ROR), PRR calculated by Chi square statistics, 95% confidence interval of PRR, observed to expected (O/E) ratio and De Mouchel method calculated PRR. Information component (IC) was given by Bayesian confidence propagation neural network. (As per regulatory criteria, PRR ≥ 2, ROR ≥ 1, Chi square statistics calculated PRR ≥ 4 and lower bound of 95% CI of PRR ≥ 1 to consider particular adverse drug reaction as a signal. Further by BCPNN method, if IC-2SD ≤ 0 then that drug-ADR pair considered has no signal; if 0 < IC-2SD ≤ 1.5, then that drug-ADR pair considered as weak signal; if 1.5 < IC-2SD ≤ 3.0, then that drug-ADR pair considered as middle signal; if IC-2SD > 3.0, then that drug-ADR pair considered as strong signal).

Results: A total of 120 reports of Drospirenone induced gallbladder diseases were reported in CADRMP database. The PRR was found to be 37.73 and by the Du Mouchel method it was 2.18. Further, the PRR calculated by Chi square statistics was 140.61 (p value < 0.0001). The lower and upper limits of 95% CI of PRR was found to be 2.17 and 4.16, respectively The O/E ratio was found to be 2.19 and ROR was found to be 40. The value of PRR > 2, ROR > 1, Chi square statistics calculated PRR > 4 (p value < 0.0001) and lower limit of 95% CI of PRR ≥ 1, also IC-2 SD is 1.12 indicating a significant weak signal associated with drospirenone and gall bladder diseases.

Conclusion: Study shows a low risk associated with drospirenone and gallbladder diseases. These data will enhance the information available to healthcare professional and may be found useful in management of women’s health.

References:

• 1. Wang F, Shen X, Guo X, Peng Y. Oral contraceptives and risk of gallbladder disease. CMAJ 2011;183:1517; author reply 1517.

• 2. Etminan M, Delaney JAC, Bressler B, Brophy JM. Oral contraceptives and the risk of gallbladder disease: a comparative safety study. CMAJ 2011;183:899–904

• 3. Jick S, Pennap D. Drospirenone- and levonorgestrel-containing oral contraceptives and the risk of gallbladder disease. Contraception2012;86:220–3Bayer to Pay $24 M to Settle Yaz Gallbladder Lawsuits [Internet]. 2013 [cited 2018 Jun 2]. Available from: https://topclassactions.com/lawsuit-settlements/prescription/3808-bayer-to-pay-24m-to-settle-yaz-gallbladder-lawsuits/

Disclosure of Interest: M. Shah Other: No conflicts of interest, C. Kothari Other: No conflicts of interest.

84 ISoP18-1124 Statistics Trend: Vaccine Reports Regarding Pediatric People in Korea

84.1 M. Kim^*1, S. Chung¹, Y. Woo¹, J. Park¹, S. Kim¹

84.1.1 ¹Korea Institute of Drug Safety and Risk Management, Anyang, Korea, Republic Of

Background/Introduction: Since the nature of immunization is mainly for healthy children, monitoring of adverse events after vaccination is important. We analyze the status of vaccine spontaneous reports in pediatric population for which factors might need to be considered when monitoring vaccine safety.

Objective/Aim: Understanding the current trend of spontaneous adverse event (AE) reports about vaccine in Korea.

Methods: We used dataset from Korea Adverse Event Reporting System, which is for collecting and managing ICRSs. We used ATC code for drugs and WHO-ART code for AEs, to analyze the comparison between number of reports regarding vaccines and non-vaccines, from 2013 to 2017. Only initial reports and the drug type of suspect or drug–drug interaction were analyzed. Ages are divided into 2 groups; infants/children/teens (pediatric group) birth through age 18, and adults age 19 and older. J07 were selected for the vaccine, and ATC codes except J07 were for non-vaccines.

Results: Total of 873,314 ICSRs were reported to the KAERS for the last 5 years. Non-vaccines (98.1%) were reported more than vaccines (1.9%). The No. of vaccine reports were 7873 (14.8%) among reports for pediatric group and 8755 (1.1%) among reports for adults. The reporting proportion of vaccine in pediatric group was much higher than that in adults group.

In the pediatric group, the most frequently reported vaccine type was followed in order by influenza (inactivated) (27.80%), rota virus (13.54%), tuberculosis (9.62%), hemophilus influenzae B/combinations with meningococcus C/conjugated (9.54%), and hemophilus influenzae B/combinations with toxoids (9.02%).

The most commonly reported AEs among pediatric group included fever (18.19%), vomiting (12.35%), lymphadenopathy (7.19%), pharyngitis (7.02%), headache (6.48%), injection site pain (5.27%), and injection site inflammation (4.85%). The most frequent reported SAEs included fever (2.10%), headache (1.00%), asthenia (0.80%), convulsions (0.75%), myalgia (0.72%), dizziness (0.58%), dyspnoea (0.52%), pneumonia (0.47%), and paralysis (0.32%).

Conclusion: As Korea has implemented national immunization program (NIP), the most frequent reported vaccine types were also influenced by NIP [1]. Injection site reaction or well-known AEs were most frequently reported for total vaccine reports among pediatric group. AEs which need careful reviewing and monitoring were identified among the SAE reports.

References:

• 1. Lee JK, Choi WS. Immunization policy in Korea. Infect Chemother 2008;40:14–23

Disclosure of Interest: None declared.

85 ISoP18-1125 The Impact of a Restrictive Regulatory Action on the Utilization Of Pregabalin Containing Products in Saudi Arabia

85.1 A. Alshatri Almotiry^*1, W. AlGhannam¹, N. AlJasser¹, A. AlShahrani¹

85.1.1 ¹Vigilance and Benefit–Risk Assessment Executive Directorate, Saudi Food and Drug Authority (SFDA), Riyadh, Saudi Arabia

Background/Introduction: Pregabalin, a drug indicated in epilepsy, neuropathic pain, and Generalized Anxiety Disorder known to be associated with drug abuse and misuse [1, 2]. In Saudi Arabia, the Saudi Food and Drug Authority (SFDA) received numerous cases of misuse and product access through community pharmacies without prescription. This provoked the SFDA to issue a circular in May 2015 banning pregabalin products from community pharmacies and restricting its dispensing to hospital and primary care centers’ pharmacies only [3].

Objective/Aim: To display the impact of pregabalin dispensing restrictions on the utilization of pregabalin products in Saudi Arabia.

Methods: The Intercontinental Marketing Statistics (IMS Health) database was accessed to determine the sales data of four pregabalin products in Saudi Arabia in the period within 2 years before and 2 years after executing the pregabalin restriction rules in May 2015 [4]. Data collected were in terms of number of packs sold to all dispensing channels. The four pregabalin products were the only registered pregabalin products at the time of the application of pregabalin restriction.

Results: In the period before the restriction of pregabalin, from 2013 to the first half of 2015, we found that the total utilization of the four pregabalin products have peaked in the 2^nd half of 2014, reaching 1,119,228 packs (Table 1). After restricting the product in mid-2015, the sales dropped from 984,731 in the first half of the year to 601,786 packs in the remaining months. This drop was followed by a slight increase in 2016, and a rebound in the first half of 2017.

The average utilization of the selected products prior to the implementation of the restriction was 173,892 packs per month. Whilst post the restriction, the average utilization decreased to 114,002 per month (34.4% decrease).

Conclusion: After restricting pregabalin products to hospital and primary care centers’ pharmacies in Saudi Arabia, the utilization notably decreased. This roughly suggests the effectiveness of this regulatory action in restricting the access of pregabalin to abusers and misusers. More structured Drug Utilization Studies are required to assure that it does not influence the access of pregabalin in legitimate patients.

References:

• 1. LYRICA^® (Pregabalin) Summary of Product Characteristics. 4th ed. Saudi Arabia: Pfizer; 2016.

• 2. Schjerning O, Rosenzweig M,Pottegård A,Damkier P,Nielsen J. Abuse potential of pregabalin. CNS Drugs 2016; 30:9–25

• 3. SFDA Strengthens Prescribing Rules on Products Containing Pregabalin. (2015). Saudi Arabia: Saudi Food and Drug Authority. Available at: https://www.sfda.gov.sa/en/drug/about/sector_departments/national_pharmacovigilance_center/Documents/Pregabalin_Safety_Communication.pdf [Accessed 22 May 2018].

• 4. IMS Health, IMS National Sales Perspectives™. January 2013–September 2017.  Extracted May 2018.

Disclosure of Interest: None declared.

86 ISoP18-1126 Outcomes of Drug Exposition During Pregnancy: Analysis from a Teratology Information Service

86.1 C. Lenoir^*1, S. Boumaïza¹, M. Boulvain², K. R. Ing Lorenzini¹, J. Desmeules¹, V. Rollason¹

86.1.1 ¹Division of Clinical Pharmacology and Toxicology, Geneva University Hospitals & Geneva University, Switzerland; ²Department of Obstetrics & Gynaecology, Geneva University Hospitals, Geneva, Switzerland

Background/Introduction: Many pregnant women are exposed to drugs and this often in early pregnancy because 50% of pregnancies are unplanned and/or detected after the 6th to 8th week of amenorrhea [1]. Studies show that the percentage of pregnant women exposed to drugs is substantial with a number of drugs being also high [2–3]. Drugs can be teratogenic, with occurrence of insufficient/abnormal fetal development, spontaneous miscarriages, in-uterine deaths, and malformations [4]. The risk of malformation in the general population is estimated at 3–5% and drugs explain approximately 4–5% of this percentage [5–6]. When a drug arrives on the market, in vitro, in vivo and animal studies show part of the potential teratogenic/foetotoxic effects, but epidemiological studies and case studies are needed to further highlight the risks.

Objective/Aim: To characterize drug exposures during pregnancy where the outcome a birth was known that came from our counselling service between 1994 and 2016.

Methods: This observational study analysed all data collected through the drug exposures during pregnancy counselling. Age, numbers and type of drug taken, timing of drug exposition and outcome of the pregnancy were retrieved. Data was analysed descriptively.

Results: A total of 1378 pregnant women were analysed. Women were between 30 and 34 years old in 31.6% (N = 435) of the cases with a global mean age of 32 years. These women were exposed to one drug in 49.1% (N = 676) of cases and more than ten drugs in 1.4% (N = 19) of cases, with a mean drug intake of 2. Analysis of the drugs altogether (N = 3137) showed that FDA Pregnancy Category C drugs represented 42.8% (N = 1344) of drugs and ATC code N (nervous system) represented 36.5% (N = 1144). The onset of drug exposure was during the first trimester of pregnancy in 95.2% (N = 2987) of patients. Regarding outcomes, the rate of induced abortion was 10.8% (N = 151), of pregnancy complications was 19.1% (N = 268) and of malformations was 3.7% (N = 52).

Conclusion: Pregnant women counselled by our information service take a mean of 2 drugs, ranging from one to 17. Drugs are from FDA Pregnancy Category C and ATC N drugs in most cases, 42.8 and 36.5%, respectively. Almost all exposures begin in the first trimester probably because women are not yet aware of their pregnancy. The rate of malformation of our cohort was of 3.7%, close to the estimated spontaneous rate of malformation. This data gives a reassuring aspect of drug exposure in pregnancy but takes into account the outcome at birth only.

References:

• 1. Rousso P. [Drug exposure during pregnancy: a differentiated approach]. Praxis 2004;93:943–7

• 2. Lacroix I, Damase-Michel C, Lapeyre-Mestre M, Montastruc JL. Prescription of drugs during pregnancy in France. Lancet. 2000;356:1735–6

• 3. Malm H, Martikainen J, Klaukka T, Neuvonen PJ; Finnish register-based study. Prescription drugs during pregnancy and lactation—a Finnish register-based study. Eur J Clin Pharmacol 2003;59:127–33

• 4. Audus KL. Controlling drug delivery across the placenta. Eur J Pharm Sci 1999;8:161–5

• 5. European Network Teratology Information Service (ENTIS) [Internet]. [23 Mai 2018]. Available on https://www.entis-org.eu/

• 6. Stevenson R, Hall J. Human malformations and related anomalies. 2nd ed. Oxford University Press. Oxford. New York. 2006

Disclosure of Interest: None declared.

87 ISoP18-1128 Monitor the Benefit and Risk of Medicines in Primary Care; Building a Bridge Between Daily Practice and Research

87.1 L. Rolfes¹, K. Hek², R. Verheij², L. van Dijk², E. van Puijenbroek^*1,3

87.1.1 ¹Netherlands Pharmacovigilance Centre Lareb, Den Bosch, the Netherlands; ²NIVEL, Netherlands Institute for Health Services Research, Utrecht, the Netherlands; ³University of Groningen, Groningen Research Institute of Pharmacy, Unit of PharmacoTherapy, Epidemiology & -Economics, Groningen, the Netherlands

Background/Introduction: Monitoring the benefits and risks of medicines in the post-marketing phase is important for a safe and effective treatment of patients. The aim of this project is to create an infrastructure in which data from the patient’s electronic health records from general practitioners (GPs) can be linked to patient reported outcomes (PROs) about their perspective of the effect and adverse drug reactions (ADRs) of medicines. In this way information about (cost) effectiveness and ADRs of medicines can be acquired. The system is designed to be applied to any new medication on the market and has several advantages above more traditional monitoring instruments like larger cohorts, longer follow-up, and a real life setting.

Objective/Aim: To evaluate the feasibility of acquiring real world data to monitor the (cost) effectiveness and ADRs of medicines.

Methods: For this pilot study, two GP practices in the Netherlands were included. Patients with an overactive bladder (OAB) were used as a use case. OAB patients were selected from electronic health records from the two GPs as obtained in the Nivel Primary Care Database. The GP indicated whether the patient was suitable to be included in the monitor.

PROs were obtained by an electronic questionnaire, using the Lareb Intensive Monitoring system. Questions were posed about patient characteristics, bladder complaints, treatment (physiotherapeutic interventions and medicines), ADRs, therapy compliance, and quality of life.

Results: 109 OAB patients were selected from the electronic health records. After revision by the GPs, an invitation to contribute to the monitor was send to 82 (75%) patients. Reasons for excluding patients were: patient cannot work with a computer or questionnaire (10×), cognitive/mental not able to fill in questionnaire (10×), patient in hospital (3×), patient died/moved (2×). GPs were enthusiastic about the method.

Nineteen patients (23%) filled in the questionnaire. Average age was 63 years (38–79 years) and it concerned 14 women. Patients reported an average score of 4.0 for urogenital complaints, on a scale from 1 (none) to 10 (very much). Eleven patients received physiotherapeutic interventions and 5 received medicines to treat the OAB. Of these, 3 patients mentioned that they experience ADRs; dry mouth (3×), dry eyes (1×) and dry vagina (1×). The OAB had an average influence of 3.0 on the quality of life, on a scale from 1 (none) to 10 (very much). These data indicate that patients are willing and able to give proper information about the effect and ADRs of their OAB treatment. After analysis, the wording of some questions was revised to make the question clearer.

Conclusion: GPs are enthusiastic and a sufficient number of patients per GP practice have been included, providing clear information about their perspective of the effect and ADRs of their treatment. A next step is to expand the registry to a greater number of patients and to develop an automatic feedback module. This concerns individualized feedback for patients which will give them insight in their own situation. In addition, there will be feedback for patients and GPs to get a general insight in the patient’s perspectives of effects and risk of medicines. This could result in better treatment, tailored to patients’ needs and experiences.

Disclosure of Interest: None declared.

88 ISoP18-1129 Spontaneous Reporting of Suspected Adverse Drug ReactionsiIn Patients with Arthritis: A Scoping Literature and Data Review

88.1 G. Greer^*1, A. Sabokbar¹, C. Anderson², J. Aronson³

88.1.1 ¹Nuffield Department of Orthopaedics, Rheumatology & Musculoskeletal Sciences, University of Oxford, Oxford, United Kingdom; ²School of Pharmacy, University of Nottingham, Nottingham, United Kingdom; ³Centre for Evidence Based Medicine, Nuffield Department of Primary Care Health Sciences, University of Oxford, Oxford, United Kingdom

Background/Introduction: Biological products (biologics) are medicines whose active principles derive from biological sources or are produced by biotechnology [1]. The types of biologics most commonly used to treat rheumatoid arthritis (RA) [2] and ankylosing spondylitis (AS) [2] are TNF inhibitors, anti-IL-1/6/17, and JAK inhibitors. The most commonly used TNF inhibitors are adalimumab, certolizumab, etanercept, golimumab, and infliximab. Their main adverse effects include skin reactions and infections, including reactivation of tuberculosis. Suspected adverse drug reactions (ADRs) are reported in various ways (e.g. as case reports and spontaneous reports to schemes such as the UK’s Yellow Card Scheme and in randomized trials, observational studies, and Biologics Registers). However, under-reporting is common and continues to be a global problem [3, 4], with a risk of worldwide mortality and morbidity [5]. The extent of under-reporting to worldwide schemes of suspected ADRs in patients taking TNF inhibitors for RA & AS is unknown.

Objective/Aim: To review the scientific literature and available data on suspected ADRs in response to biologics for treating patients with RAor AS, seeking information on reporting rates.

Methods: We searched for papers using 54 search engines from the University of Oxford’s online libraries (e.g. PubMed, Medline). We imposed no restrictions, apart from the period surveyed (2012–2017). We also searched other online information sources (e.g. the website of the National Institute of Health and Care Excellence). We generated a series of tables describing examples of the classes and types of biologics used and examples of the associated ADRs. We also generated figures to describe suspected ADRs reported to the UK Yellow Card Scheme for five biologics used in the treatment of RA and AS (adalimumab, certolizumab, etanercept, golimumab, and infliximab).

Results: We retrieved 120 papers, of which 30 passed screening of titles; of those, 14 were published during 2012–2017, of which we were able to retrieve 11. None of those dealt with any aspects of international spontaneous reporting schemes, including the extent of reporting/under-reporting. The MHRA received 22,773 Yellow Card reports (1999–2017) for the five selected biologics, including 3226 reports of musculoskeletaland connective tissue disorders (six fatal) and 1637 reports of joint disorders.

Conclusion: Compared with other biologics, the rate of Yellow Card reporting of suspected ADRs associated with biologics used to treat RA and AS is relatively high, although the extent of under-reporting for these drugs is not known and has not been studied.

References:

• 1. Giezen TJ, Mantel-Teeuwisse AK, Straus SM, Schellekens H, Leufkens HG, Egberts AC. Safety-related regulatory actions for biologicals approved in the United States and the European Union. JAMA 2008; 300: 1887–96

• 2. Scherer K, Spoerl D, Bircher AJ. Adverse drug reactions to biologics. J Dtsch Dermatol Ges 2010; 8: 411–26

• 3. Hazell L,Shakir SAW. Under-reporting of adverse drug reactions: a systematic review. Drug Saf 2006; 29: 385–96

• 4. Lopez Gonzalez E, Herdeiro M, Figueiras A. Determinants of under-reporting of adverse drug reactions: a systematic review. Drug Saf 2009; 32: 19–31

• 5. World Health Organization (WHO). Safety of Medicines—A Guide to Detecting and Reporting Adverse Drug Reactions—Why Healthcare Professionals Need to Take Action. Geneva; WHO: 2002; http://apps.who.int/medicinedocs/en/d/Jh2992e/

Disclosure of Interest: None declared.

89 ISoP18-1130 Assessment Of Cyclosporine Blood Concentrations In Adults With Aplastic Anemia

89.1 S. Ben Tekaya¹, M. Ben sassi^1,2, R. Charfi^1,2, E. Gaies^1,2, I. Salouage¹, N. Jebabli², H. El Jebari², R. Daghfous^*1,2, S. Trabelsi^1,2

89.1.1 ¹National Pharmacovigilance center; ²Clinical and Experimental Pharmacology Research Laboratory LR16SP02, Tunis, Tunisia

Background/Introduction: Aplastic anemia is defined as peripheral blood pancytopenia caused by bone marrow failure [1]. Patients are treated successfully with a combinaison of antilymphocyte globulin and cyclosporine (CsA), especially in case of lack of a histocompatible donor. CsA is an immunosuppressive agent with a narrow therapeutic index and a high variability between individuals [2]. That’s why therapeutic drug monitoring (TDM) of CsA is important to optimize therapy and to avoid toxicity.

Objective/Aim: To assess CsA blood concentrations in adults with aplastic anemia.

Methods: It consisted in a retrospective study made between February 2016 and December 2017 in the Department of Clinical Pharmacology. We included 130 samples collected from 17 patients who were treated by CsA for aplastic anemia. We measured CsA trough blood concentration (C0) by an immunological method. C0 should be maintained between 200 and 400 ng/mL.

Results: Median age of our population was 26 years. The sex ratio (M/W) was 0,74 and average weight was 67 kg. Mean C0 was 223 ± 140  ng/mL. Among samples, 59 were from oral administration and 71 were from intravenous administration. Median dose of our population was 2 mg/kg/day. More than half of measured C0 (53%) was infratherapeutic. The results showed that 39% of measured C0 was in therapeutic range and 8% was supratherapeutic. Median dose and concentration among three each group are showed in Table 1.

Table 1 Median dose and C0 of CsA

	Infratherapeutic (< 200 ng/mL)	In therapeutic range	Spratherapeutic (> 400 ng/mL)
Dose (mg/kg/day)*	2.06	2.05	1.71
C0 (ng/mL)**	132.72	279.54	553.12

1. *p = 0.5; **p = 003

Conclusion: In our study, despite similar dose, there is a significant difference between CsA concentrations. Because of inter and intra-variability of CsA pharmacokinetic, TDM is required to adjust the individual doses of the drug.

References:

• 1. Ni S, Zhao W, Wang J, Zeng S, Chen S, Jacqz-Aigrain E, et al. Population pharmacokinetics of ciclosporin in Chinese children with aplastic anemia: effects of weight, renal function and stanozolol administration. Acta Pharmacol Sin. 2013;34(7):969–75.

• 2. Xue L, Zhang W, Ding X, Zhang J, Miao L, Bao J. Population Pharmacokinetics and Individualized Dosage Prediction of Cyclosporine in Allogeneic Hematopoietic Stem Cell Transplant Patients. Am J Med Sci. 2014;348(6):448–54.

Disclosure of Interest: None declared.

90 ISoP18-1131 Facilitating Safety Signal Triage and Prioritization Through Literature Based Matrix Factorization Ranking of Drug and Designated Medical Event Pairs

90.1 K. Bannout¹, K. Shen², W. Wallis¹, S. Spangler^*3

90.1.1 ¹Celgene, Summit, New Jersey, NY, United States; ²IBM Watson Health, New York, NY, United States; ³IBM Research-Almaden, San Jose, CA United States

Background/Introduction: Adverse drug reactions, especially those which are Designated Medical Events (DMEs), are a serious concern for patients, physicians, regulatory authorities, pharmaceutical companies, drug safety professionals and pharmacoepidemiologists. A number of methods have been employed to try and predict drug-adverse event (AE) associations [1], but those methods do not tap into hidden underlying patterns in literature text for their predictions.

Objective/Aim: We sought to predict potential drug-AE associations to support prioritization of AE reports in safety signal evaluation.

Methods: We employed matrix factorization (MF) within a network of drug and AE relationships for the drug-AE prediction. The network relationships were derived through natural language processing on Medline abstracts and PubMed Central Open Access (PMCOA) articles. Matrix factorization is a computational technique for collaborative filtering that creates a dense matrix from a sparse one by creating two product matrices which when multiplied approximate the original network [2]. It is commonly used in recommender systems (e.g. suggesting movies to watch) [3], but has not been employed in drug safety.

We selected 9 AE types from the 2016 EMA DME list [4] that were also included in the 2016 FDA DME list [5] and for which a single term was considered likely to capture the majority of events. Medications that included the selected DME in both the Warnings & Precautions section and ADR section of the drug label, but not in the Indications section, were selected as positive controls for MF scoring using Dr. Evidence^® prescribing information analytics on both US and UK medications. Negative controls for MF scoring were medications where the selected DME did not appear in any part of the label. The positive drug–AE pairs were zeroed out (removed) from the input matrix and then the matrix factored to yield a drug–AE score. From these observed scores, Z-scores were calculated using a simulated expected score and standard deviation for each pair.

Results: Our model achieved an average AUC (the area under the receiver operating characteristic curve) of 0.78 over the the 9 DMEs. On average for a given DME, the mean true positive Z-score was 2.97 and the mean true negative Z-score was − 1.35.

Conclusion: Matrix factorization of DME-drug pairs shows promise as a useful tool in supporting safety signal triage and warrants further evaluation with other drug-AE pairings and in prospective studies.

References:

• 1. Huang LC, Wu X, Chen JY. Predicting adverse side effects of drugs. BMC Genomics 2011; 12(Suppl 5): S11

• 2. Lee DD, Seung HS. Learning the parts of objects by non-negative matrix factorization. Nature 1999; 401: 788–791

• 3. Koren Y. The BellKor Solution to the Netflix Grand Prize [Internet]. 2009. Available from: http://www.netflixprize.com/assets/GrandPrize2009_BPC_BellKor.pdf

• 4. Designated Medical Event (DME) list [Internet]. European Medicines Agency; 2016. Available from: http://www.ema.europa.eu/docs/en_GB/document_library/Other/2016/08/WC500212079.xls

• 5. 2016 FDA OSE DMEs and Associated MedDRA Preferred Terms [Internet]. US Food and Drug Administration; 2016. Available from: https://www.accessdata.fda.gov/drugsatfda_docs/nda/2016/206843Orig1s001,s003OtherR.pdf

Disclosure of Interest: None declared.

91 ISoP18-1132 Product Confusion Errors: A Pharmacovigilance Analysis of Vigibase™ and Canada Vigilance Case Narratives

91.1 G. Castillon^*1, Z. Gesztesi², Y. Moride^1,3

91.1.1 ¹YOLARX Consultants, Montreal, Canada; ²AstraZeneca, Mississauga, Canada; ³Faculty of Pharmacy, Université de Montréal, Montreal, Canada

Background/Introduction: Product confusion errors (associated with drug name, dosage, label, packaging) are a common cause of medication errors that are preventable and may lead to inappropriate medication use or patient harm. The types of product confusion errors that occur most frequently, as well as their consequences, have not been extensively examined to date.

Objective/Aim: (1) To describe the product confusion errors that have been reported worldwide; (2) To assess the consequences of confusion errors [serious or non-serious adverse events (AE), lack of effectiveness]; (3) To characterize the populations affected by these errors.

Methods: We conducted an analysis of VigiBase™ (an international spontaneous reporting database that covers over 110 countries) over the period 2007–2016. In order to identify the relevant cases, we identified four MedDRA Preferred Terms (PTs) related to product confusion errors. As VigiBase™ does not collect information on the party responsible for the confusion error, we performed a qualitative review of individual case safety reports (ICSRs) recorded in Canada Vigilance spontaneous reporting system.

Results: A total of 1769 individual cases of product confusion errors were found in VigiBase™, corresponding to 2409 PTs related to confusion errors (average of 1.4 per ICSR). Product label confusion (e.g., confusing or vague instructions) was the most common error reported (54.1%), followed by product name confusion (e.g., similar names of drugs with different indications) (26.3%). Patients or consumers accounted for the most frequent reporting source (39.4%), and the number of reports has been increasing dramatically over time. Among the 30% reports providing patient age, median age was 57 years [34.8–70 (Q1–Q3)]. There appeared to be a small female predominance (58.0%). Of the reported confusion errors, 76.5% resulted in non-serious AEs, 22.0% in serious AEs, 4.6% in lack of effectiveness, and 3.2% had no adverse effect. Antithrombotic agents were often involved in the serious AEs related to packaging confusion. Insulins were the most frequently involved in name or label confusions, followed by drugs to treat peptic ulcer and gastro-oesophageal reflux disease. The qualitative review of the 52 case narratives of product confusion errors identified in Canada Vigilance revealed that 53.8% of errors were made by patients or their families, 11.5%, respectively, by pharmacists and unspecified healthcare professionals, and 9.6%, respectively by physicians and nurses.

Conclusion: Although regulatory agencies have issued guidelines to avoid look-alike/sound-alike drug names, confusion errors still occur, with some leading to serious AEs. Study showed that errors are made principally by patients or their families, and are also mainly reported by consumers. These findings highlight the opportunity to minimize the occurrence of product confusion errors through patient-centered interventions.

Disclosure of Interest: G. Castillon Grant/Research support from: AstraZeneca Canada, Z. Gesztesi Employee of: AstraZeneca Canada, Y. Moride Grant/Research support from: AstraZeneca Canada.

92 ISoP18-1133 Framework for Analyzing Consumer Health Questions

92.1 J. Cao^*1

92.1.1 ¹Global Post Market Safety Surveillance, Herbalife Nutrition, Torrance, California, United States

Background/Introduction: Consumer health questions glean valuable insight into consumer attitudes and behaviors, and provide opportunities for educational initiatives. However, consumer health questions are typically lengthy, complex, and often utilize open-domain language rather than medical terminology [1, 2].

Objective/Aim: To develop a framework for analyzing consumer health questions in the knowledge management database.

Methods: We extracted consumer health questions created in the internal knowledge management system from January 1, 2017 to December 31, 2017 and limited to only English language submissions. The data was analyzed using R and the tm package [3, 4]. Singularly occurring words were excluded from any analyses, common English stopwords and additional stopwords assembled from company specific terms (e.g., department names, brand names) were removed, and related words in the top 250 underwent word stemming as part of the data preprocessing. With the final dataset, a word cloud was constructed from the top 250 frequently occurring terms. Additionally, word associations for key terms were calculated using a correlation limit of 0.2.

Results: 1636 consumer health questions were created in the knowledge management database in 2017. There were 2278 unique terms in the consumer health questions dataset after preprocessing of the data. The most frequently occurring terms were: product, shake, weight, use, and loss. Other key terms that were identified were: protein, safe, and allergy. We analyzed the key terms for word associations to identify other key terms that appear to be correlated. The results are presented in Table 1.

Conclusion: Text mining techniques can be used as a framework for analyzing consumer health questions. Frequently occurring words elucidate the topics of greatest interest to consumers, and word associations provide general context to the questions being asked. This framework can be used to guide the development of educational materials related to health and safe product use for consumers.

References:

• 1. Kilicoglu H, Abacha AB, Mrabet Y, Shooshan SE, Rodriguez L, Masterton K, et al. Semantic annotation of consumer health questions. BMC Bioinformatics 2018;19:1–28

• 2. Roberts K, Demner-Fushman D. Interactive use of online health resources: a comparison of consumer and professional questions. J Am Med Inform Assoc 2016;23:801–11

• 3. R Core Team. R: A language and environment for statistical computing. Vienna, Austria: R Foundation for Statistical Computing; 2017

• 4. Feinerer I, Hornik K. tm: Text Mining Package. 2017

Disclosure of Interest: J. Cao Employee of: Herbalife Nutrition.

93 ISoP18-1134 Paritaprevir/Ritonavir/Ombitasvir and Dasabuvir Related Liver Injury in Real-World Setting: Data from Taiwan National ADR Reporting System

93.1 P.-H. Chao^*1,2, W.-I. Huang^1,2, W.-W. Chen^1,2

93.1.1 ¹Taiwan Drug Relief Foundation, Taipei City, Taiwan; ²Taiwan National ADR Reporting Center, Taipei City, Chinese Taipei

Background/Introduction: In 2015, U.S. FDA has announced that paritaprevir/ritonavir/ombitasvir and dasabuvir (PrOD) regimen, which is used for hepatitis C virus (HCV) infection, can cause serious liver injury mostly in patients with underlying advanced liver disease [1, 2]. PrOD has been approved for the treatment of genotype 1 and 4 HCV infection in Taiwan and was reimbursed by Taiwan National Health Insurance since January 2017 for patients with advanced hepatic fibrosis or compensated cirrhosis who had experienced treatment failure with interferon and ribavirin therapy [3].

Objective/Aim: To evaluate the liver injury risks associated with PrOD in real clinical practice in Taiwan.

Methods: We reviewed all PrOD-associated reports collected in Taiwan National Adverse Drug Reaction (ADR) Reporting System from 2015 to 2017. Reported adverse events were categorized by MedDRA System Organ Class (SOC). Liver injury cases were identified by the search of MedDRA Preferred Term (PT) under drug related hepatic disorders of Standardised MedDRA Query (SMQ) version 21.0. Baseline characteristics and the patterns of liver injuries of these cases were analyzed.

Results: 59 PrOD-associated reports were derived from the database, and among them 41 cases were related to liver injury. The mean age of those cases was 67.6 ± 17.8 years old and females accounted for 61%. 40 cases (97.5%) were classified as serious and 15 cases (6 duplicate reports) were led to death. 28 cases using WHO-UMC causality assessment were classified as possible or above. In these cases, 14 and 6 patients were with underlying liver cirrhosis and previous/concurrent hepatocellular carcinoma, respectively. Two cases were reported with HBV reactivation and 1 case was resulting in death. All of the above cases were reported with elevated serum bilirubin. Substantial serum total bilirubin (t-bil) and direct bilirubin (d-bil) elevation without concomitant significant increase of liver enzymes within 7 days after initiating PrOD regimen were found in 12 cases, of which t-bil level ranged from 2.2 to 14.2 mg/dl and 80% were greater than 4 mg/dl. The majority of the cases developed initial t-bil elevation within 6 weeks of regimen introduction, except cases with HBV reactivation was found to be around 90 days.

Conclusion: Our findings suggest that possible risk of liver injury with the use of PrOD regimen in Taiwan and patients with moderate to severe cirrhosis prior to the therapy may be at risk. Close monitoring for clinical signs and symptoms of hepatic decompensation as well as intensively performing hepatic laboratory testing including bilirubin levels during the first 4 weeks of treatment are strongly recommended to prevent such risks.

References:

• 1. DA Drug Safety Communication. FDA warns of serious liver injury risk with hepatitis C treatments Viekira Pak and Technivie (2015 Oct.). https://www.fda.gov/Drugs/DrugSafety/ucm468634.htm?source=govdelivery&utm_medium=email&utm_source=govdelivery

• 2. Armen MP, Corneliu PP, Corneliu B, et al. Real-world efficacy and safety of ombitasvir, paritaprevir/r + dasabuvir + ribavirin in genotype 1b patients with hepatitis C virus cirrhosis. Liver Int 2018;38:602–10

• 3. Liu CH, Liu CJ, Su TH, et al. Real-world effectiveness and safety of paritaprevir/ritonavir, ombitasvir, and dasabuvir with or without ribavirin for patients with chronic hepatitis C virus genotype 1b infection in Taiwan. J Gastroenterol Hepatol 2018;33:710–7

Disclosure of Interest: None declared.

94 ISoP18-1135 Hepatotoxicity with a Natural Dietary Supplement, Artemisia Annua L. Extract in Grapeseed Oil. New Zealand Pharmacovigilance Centre Reports

94.1 R. L. Savage^*1, M. Tatley¹, J. Barnes²

94.1.1 ¹New Zealand Pharmacovigilance Centre, Dunedin School of Medicine, University of Otago, Dunedin, New Zealand; ²School of Pharmacy, Faculty of Medical and Health Sciences, University of Auckland, Auckland, New Zealand

Background/Introduction: Artemisia annua L. (Qing Hao, sweet wormwood; family: Asteraceae) has been used in traditional Chinese medicine for two millenia [1]. Recently, an extract of A. annua in grapeseed oil (Arthrem^®) has been marketed in New Zealand (NZ) as a natural dietary supplement for maintaining and supporting joint health. Between February 2016–December 2017, the NZ Pharmacovigilance Centre (NZPhvC) received 14 reports of hepatic adverse drug reactions (ADRs) associated with Arthrem^®.

Objective/Aim: To analyse reports of Arthrem^®-associated hepatotoxicity for causality, seriousness and patterns of hepatotoxicity.

Methods: Qualitative anaysis of NZPhvC reports for patient characteristics, patterns of hepatotoxicity, temporality, seriousness and outcome.

Results: The 14 reports originated from several reporter groups and geographic areas in NZ. Cases (ten females, four males) were aged 48-77 years. Six reports were classified’serious’ as the patients involved were hospitalised. Where stated, the indication for use was osteoarthritis or joint pain. Arthrem^® was the sole medicine in nine reports and the sole suspect in four. There were no preceding clinical conditions or concomitant medicines considered a more likely cause of the hepatic reactions. Increased hepatic enzyme concentrations were reported for 10 cases and hepatitis for four. A prominent feature was jaundice (ten patients), in some cases symptomatic and/or the presenting manifestation. Reports with sufficient information indicated that two patients had hepatocellular reactions, both with jaundice, one confirmed by biopsy; three had mixed hepatic reactions. One other biopsy confirmed toxicity. Hepatobiliary scans, serology and alcohol intake were reported for a minority and did not provide alternative explanations. The time to ADR onset from Arthrem^® initiation (12 patients) was typically within 2 months. Twelve patients were recovered/recovering following Arthrem^® dechallenge, most within 2 months (8).

Conclusion: A case series of suspected ADRs associated with an extract of A. annua in grapeseed oil provides evidence of hepatotoxicity sometimes leading to hospitalisation. The reaction appears reversible on dechallenge. The finding is unexpected since there is little reference to hepatoxicity with A. annua in the literature or VigiBase, the WHO Global Database of Individual Case Safety Reports although hepatotoxicity associated with the constituent artemisinin has been described [2]. This raises several questions, including whether these reports reflect increased exposure following the product’s wide promotion, and/or whether product formulation (e.g. type of extract) or quality (e.g. source of raw material) issues are involved.

References:

• 1. Hsu E. The history of qing hao in the Chinese materia medica. Trans R Soc Trop Med Hyg 2006;100:505–50

• 2. National Institutes of Health. LiverTox. Clinical and Research Information on Drug-Induced Liver Injury. Artemisinin derivatives. Available at https://livertox.nih.gov/ArtemisininDerivatives.htm [last accessed June 3, 2018]

Disclosure of Interest: None declared.

95 ISoP18-1137 Using WHO Recommendations Antibiotic Consumption Evaluation in Kazakhstan for the Period 2012–2017

95.1 G. Zhussupova¹, S. Zhaldybayeva¹, D. Utepova^*1, L. Yesbatyrova¹

95.1.1 ¹RSE Republican Center for Healthcare Development of the Ministry of Health of the Republic of Kazakhstan, Astana, Kazakhstan

Background/Introduction: Antimicrobial resistance is a global problem, one of the reasons of which is the inadequate antibiotic prescribing and use practice. Regular monitoring of antibiotic consumption will reveal the reasons of antibiotic resistance and improper use.

Objective/Aim: Antibacterial drugs consumption estimation for identifying the problems of using antibiotics at a stationary level in Kazakhstan.

Methods: (1) ATC/DDD-methodology;

(2) Analytical.

ATC/DDD-methodology—consuption indicators determination using DDD—defined daily dose, recommended by WHO and medicines’ATC code. The measure—DID [1].

A comparative analysis of antibacterial preparations (ABP) consumption was conducted by pharmacological groups and routes of administration.

Results: The total ABP for systemic use consumption in 2012-2016 was systematically decreasing, but in 2017 there is significantly increased by 30%, and amounted 3.0 DID compared with 2016 (2.3).

Consumer leader is cephalosporin subgroup: consumption has increased to 1.1 in 2017, compared with 0.9 in 2012.

The most frequently consumed antibiotics in 2017 were I- and III-generation cephalosporins—48 and 42%, respectively. 82% of the total consumption of III-generation cephalosporins in 2017 was ceftriaxone.

A comparative analysis of amoxicillin and amoxicillin with clavulanic acid consumption showed an consumption increase of the second to 76% in 2017, while the share of amoxicillin fell steadily to 24% in 2017, compared to 57% in 2012.

Analysis of the ABP consumption by the routes of administration also revealed a gradual decrease of oral and parenteral forms consumption in 2012–2016, but a significant increase of both forms consumption in 2017—by 27 and 35%, respectively. 

Conclusion: Cephalosporins (I- and III-generation representatives) excessive consumption without results of sensitivity analysis may lead to the development of resistance of microorganisms to antibiotics, including hospital strains.

Cephalosporins are the most “unreliable” antibiotics, the use of which is accompanied by a pronounced “parallel damage”—selection of multiresistant microorganisms not only among strains of pathogens, but also among not etiologically significant microorganisms [2].

Amoxicillin with clavulanic acid more actively suppresses microorganisms that produce β-lactamases. However, hepatotoxic effects of clavulanic acid, especially among children should be taken into account, and the daily dose (no more than 6 mg/kg) should be monitored.

References:

• 1. WHO Collaborating Centre for Drug Statistics, Norwegian Institute of Public Health; Irene Little Scare, Seminar on ATC/DDD, Copenhagen, 26.02.2015

• 2. Kozlov RS. Selection of resistant organisms when using antimicrobial drugs: the concept of parallel damage. Antimicrobials, 2010, Edition 12, №4, P. 284

Disclosure of Interest: None declared.

96 ISoP18-1138 Pharmacovigilance Awareness in Healthcare Professionals

96.1 E. D. Aral Karakulak^*1, H. Barışkaner²

96.1.1 ¹Konya Beyhekim State Hospital, Konya, Turkey; ²Medical Faculty Pharmacology Department, Selcuk University, Konya, Turkey

Background/Introduction: Pharmacovigilance is the pharmacological science relating to the collection, detection, assessment, monitoring, and prevention of adverse effects with pharmaceutical product [1]. In this study a questionnaire was conducted to 200 healthcare professionals in Beyhekim State Hospital to evaluate knowledge and attitude about pharmacovigilance and adverse effects. 200 healthcare professionals consist of 60 doctors (57 specialists, 3 general practitioners), 5 pharmacists, 100 nurses, 10 midwifes, 8 health officers, 5 first aid care technicians, 12 anesthesia technicians. For statistical analysis respondents evaluated in 3 groups: doctors group n = 60, nurses group n = 110 and the other healthcare professionals group n = 30.

Objective/Aim: The aim is to increase pharmacovigilance awareness in healthcare professionals in Turkey by measuring the knowledge, attitude and practices towards pharmacovigilance and adverse drug reactions by conducting a questionnarie in healthcare professionals. 

Methods: A-15 question survey was directed to 200 healthcare professionals. The statistical evaluation of the study was made by using SPSS 21 statistical package program. For the analysis of the obtained data, Chi square test, Pearson Chi Square test, Fisher’s Exact Test was used (statistically significant when p < 0.05).

Results: 71 of 200 respondents (35.5%) had received pharmacovigilance education. 21.7% of 60 doctors, 43.6% of 100 nurses, 33.3% of 30 other healthcare professionals had received pharmacovigilance education. In comparison with the Chi square test, it was determined that there is a significant relationship between health professionals occupations and their education status.

63% of healthcare professionals experienced an adverse drug reaction (ADR) at least. 37% of them had never experienced any ADRs. 15% of the respondents had previously reported an ADR. 85% had never reported any ADRs. Respondents’ main reason (26.5%) for not reporting is not knowing where to report.

73.5% of respondents selected the correct answer when asked who should report ADRs. 68.2% of nurses provided the correct answer of definition of ADRs. 91.7% of doctors provided the correct answer of definition of ADRs. Doctors rate for this question is higher than the other groups. There is a significant relationship between health professionals occupations and their knowledge about definition of ADRs (p < 0.05).

41% of respondents selected correct answer when asked where to report ADRs in Turkey and 6.5% has no idea.

Conclusion: The survey showed that the awareness of pharmacovigilance of health professionals is not sufficient. The awareness of pharmacovigilance among healthcare professionals in other hospitals in our country can give similar results. Therefore, the inclusion of pharmacovigilance in the curricula of pharmacology education in faculties can increase the awareness of pharmacovigilance in healthcare professionals through training, seminars and in hospitals. The formation of accurate information about pharmacovigilance in health workers can be provided by central and/or regional training, seminars, congresses, workshops and so on.

References:

• 1. World Health Organization Definitions [document on the Internet]. 09.10.2017. Available from: http://www.who.int/medicines/areas/quality_safety/safety_efficacy/trainingcourses/definitions.pdf

Disclosure of Interest: None declared.

97 ISoP18-1140 Factors Influencing Low Adverse Drug Reaction Reporting Among Healthcare Professionals in Ghana

97.1 W.-R. Nagumo^*1, R. Cooper¹, R. Akparibo¹

97.1.1 ¹ScHARR, University of Sheffield, Sheffield, United Kingdom

Background/Introduction: Adverse Drug Reaction (ADR) is one of the leading causes of morbidity, mortality and increased healthcare cost. For instance, recent studies in the UK show that ADR costs the NHS £98.5 million and 712 deaths yearly [1]. Elsewhere in the USA, it is estimated that ADRs have led to over 2 million hospitalizations and over 100,000 fatalities making it the fourth leading cause of death [2]. Given this burden, the reporting of ADRs is essential and requires attention from Health Care Professionals (HCPs) to save costs and improve patient safety.

In Ghana, studies are scant and ADR reporting rates are very low. For instance, only 30 reports were received in 2016, which is below the WHO recommendation of 200–250 reports per million inhabitants [3].

Objective/Aim: The study objective is to assess the perceived factors affecting low spontaneous adverse drug reaction reporting among HCPs in Ghana.

Methods: This was a cross-sectional survey, designed based on previous studies with modifications to suit the purpose of the study. 450 self-administered questionnaires were distributed randomly to healthcare professionals. 10 departments in four hospitals (3 secondary care facilities and 1 teaching hospital) were purposively selected for the study. The questionnaire items collected data on their backgrounds, knowledge, attitudes, perceptions, and practice of the reporting system. We analyse the data using SPSS version 22.

Results: The overall response rate was 85.7%. There were 386 respondents (277 nurses, 18 midwives, 40 doctors, 8 Physician Assistant, 19 pharmacists, 5 pharmacist technicians and 9 missing data) who return their questionnaires.

The results show that there was generally a positive attitude to ADR reporting among HCPs. Knowledge about ADR reporting was however very low with merely 18% responding correctly to the five questions assessing knowledge. For instance, only 31.9% of HCPs were aware of the national pharmacovigilance center while 0.5% knew the recommended number of days to send any type ADR report. 20% knew the required number of days to submit a serious suspected and unexpected ADR and 5.2% knew the FDA SMS short code to submit a report to. 32.9% knew the basic information required on the ADR form. 75.6% had not reported any ADRs in the year and 45.9% strongly agreed that they were ignorant about the reporting system. A majority of HCPs (68.4%) either disagreed or strongly disagreed that the following factors affected reporting; guilt of wrong doing, fear of litigation, uncertainty about causality, ambition to keep ADRs for cases series and carrier progression, lack of time and complacency that, all drugs are safe.

Conclusion: These results are preliminary with data showing that there is lack of knowledge but a positive attitude towards reporting. There is a need for further inferential statistics on the significance and association between the variables. The qualitative part of this study will also further explain and understand the findings.

References:

• 1. Elliott R, Camacho E, Campbell F, Jankovic D, Martyn St James M, et al. (2018). Prevalence and Economic Burden of Medication Errors in The NHS in England. Rapid evidence synthesis and economic analysis of the prevalence and burden of medication error in the UK. Policy Research Unit in Economic Evaluation of Health and Care Interventions. Universities of Sheffield and York

• 2. Davies EC, Green CF, Taylor S, Williamson PR, Mottram DR, Pirmohamed M. Adverse drug reactions in hospital in-patients: a prospective analysis of 3695 patient-episodes. PLoS One 2009;4:e4439

• 3. Sabblah GT, Akweongo P, Darko D, Dodoo ANO, Sulley AM. Adverse drug reaction reporting by doctors in a developing country: a case study from Ghana. Ghana Med J 2014;48:189–93

Disclosure of Interest: None declared.

98 ISoP18-1141 Characteristics of Patients Started on a Reduced Dose of Rivaroxaban for Atrial Fibrillation—Results from the ROSE Study

98.1 M. Davies^1,2, A. Evans^1,2, F. Coukan^1,2, L. Wise^1,2, S. Shakir^*1,2

98.1.1 ¹Drug Safety Research Unit, Southampton, United Kingdom; ²University of Portsmouth, Portsmouth, United Kingdom

Background/Introduction: Rivaroxaban is used to treat a range of conditions, including prevention of stroke and systemic embolism in patients with atrial fibrillation (AF). This indication requires a fixed once-daily (od) dose (20 mg) with dosage reduction to 15 mg once daily only recommended in patients (pts) with renal impairment (creatinine clearance 15–49 ml/min). However other clinical factors might be considered by the prescriber when choosing an initial dose.

Objective/Aim: To describe the clinical characteristics of pts prescribed an initial dose of (≥ 20 and < 20 mg daily) for treatment of AF in secondary care.

Methods: The Rivaroxaban Observational Safety Evaluation (ROSE) study was a specialist cohort event monitoring study of pts prescribed rivaroxaban. Specialists provided information from medical records via questionnaires at baseline and ≥ 12 weeks. Information on the prescribed initial dose and baseline characteristics including risk factors for bleeding (HAS-BLED criteria) was collected. For pts prescribed a reduced initial dose (< 20 mg daily), we assessed clinical characteristics vs. pts prescribed ≥ 20 mg daily.

Results: The cohort consisted of 965 pts with AF; of whom (18.3%, n = 177), were prescribed an initial dose of < 20 mg daily [15 mg od (n = 173); 10 mg od (n = 3); 4 mg od (n = 1)].

The majority were female (57.6%, median age 84 years) vs. pts prescribed ≥ 20 mg daily [20 mg od (n = 750); 30 mg od (n = 16)], where the majority were male (56.4%, median age 75 years). 36/177 (20.3%) pts had a history of CKD 3–4 or 5. For 22 patients the start dose was unknown.

Frequency of HAS-BLED criteria in pts prescribed < 20 mg daily vs. ≥ 20 mg daily was: Hypertension (37.3 vs. 38.5%), Abnormal renal function (Chronic dialysis, renal transplant, serum creatinine ≥ 2.3 mg/dL or 200 µmol/L) (3.4 vs. 1.2%), Abnormal liver function (0.6 vs. 1.0%), History of Stroke (31.1 vs. 30.4%), History of Bleeding/predisposition (18.6 vs. 16.3%), Labile INR (N/A), age ≥ 65 years (96.1 vs. 81.2%), Drug therapy (52.0 vs. 51.4%), Alcohol (0 vs. 4.8%).

Conclusion: In pts prescribed a reduced total daily dose of rivaroxaban < 20 mg od, there were a higher proportion of females and those aged ≥ 65 years. Patients prescribed this lower dose were also more likely to have abnormal renal function, defined as per HAS-BLED, than those patients prescribed ≥ 20 mg daily. Although the label recommends dose reduction in AF pts with renal impairment, other factors seem to impact the choice of the initial dose in the clinical setting.

Disclosure of Interest: M. Davies Other: The DSRU is an independent charity (No. 327206) which works in association with the University of Portsmouth. The DSRU has received funding from Bayer, the manufacturer of Xarelto^®. The DSRU makes the final decision on the publication of external communications. A. Evans Other: The DSRU is an independent charity (No. 327206) which works in association with the University of Portsmouth. The DSRU has received funding from Bayer, the manufacturer of Xarelto^®. The DSRU makes the final decision on the publication of external communications. F. Coukan Other: The DSRU is an independent charity (No. 327206) which works in association with the University of Portsmouth. The DSRU has received funding from Bayer, the manufacturer of Xarelto^®. The DSRU makes the final decision on the publication of external communications. L. Wise Other: The DSRU is an independent charity (No. 327206) which works in association with the University of Portsmouth. The DSRU has received funding from Bayer, the manufacturer of Xarelto^®. The DSRU makes the final decision on the publication of external communications. S. Shakir Other: The DSRU is an independent charity (No. 327206) which works in association with the University of Portsmouth. The DSRU has received funding from Bayer, the manufacturer of Xarelto^®. The DSRU makes the final decision on the publication of external communications.

99 ISoP18-1142 Assessment of Start Dose in Patients Prescribed Rivaroxaban for Atrial Fibrillation with Chronic Kidney Disease-Results from the ROSE Study

99.1 M. Davies^1,2, A. Evans^1,2, F. Coukan^1,2, L. Wise^1,2, S. Shakir^*1,2

99.1.1 ¹Drug Safety Research Unit, Southampton, United Kingdom; ²University of Portsmouth, Portsmouth, United Kingdom

Background/Introduction: Rivaroxaban is used to treat patients with non-valvular atrial fibrillation (AF) for the prevention of stroke and systemic embolism. This requires a fixed daily dose (20 mg) with dosage reduction only recommended in patients with a reduced creatinine clearance (15–49 ml/min), a range encompassed within chronic kidney disease (CKD) stage 3–4 (eGFR 15–59) although CKD 3a (eGFR 45–59) is largely outside this range. Rivaroxaban is not recommended in patients with severe renal impairment (creatinine clearance < 15 ml/min).

Objective/Aim: To assess starting dose amongst patients with chronic kidney disease stage 3–4 or 5.

Methods: The Rivaroxaban Observational Safety Evaluation (ROSE) study was a specialist cohort event monitoring study of patients prescribed rivaroxaban. Specialists provided information via detailed questionnaires at baseline (and ≥ 12 weeks). Baseline characteristics included starting dose and presence of either CKD stage 3–4 (eGFR 15–59) or CKD stage 5 (eGFR < 15). Starting dose was examined amongst patients with CKD stage 3–4 and 5 to assess how many patients had a reduced starting dose of less than 20 mg od.

Results: The cohort consisted of 965 patients with AF: 75 patients with history of either stage 3–4 CKD (n = 73, 7.6%) or stage 5 CKD (n = 2, 0.2%). Of the patients with CKD stage 3–4, 36 (49.3%) were started on < 20 mg daily [15 mg od (n = 35); 10 mg od (n = 1)]. 35 patients (48.0%) were started on ≥ 20 mg [20 mg od (n = 34); 30 mg od (n = 1)]. Starting dose was missing for two patients (2.7%). Neither patient with CKD stage 5 had a dose reduction.

Conclusion: Our results suggest that amongst patients with CKD stage 3–4, approximately half were started on the recommended reduced dose of < 20 mg od. Not all patients with CKD stage 3–4 would be recommended a dose reduction as per the product label. A UK audit suggests that approximately 20% of patients in the UK with CKD 3 are 3b rather than 3a although the frequency may be higher in a hospital cohort.

Disclosure of Interest: M. Davies Other: The DSRU is an independent charity (No. 327206) which works in association with the University of Portsmouth. The DSRU has received funding from Bayer, the manufacturer of Xarelto^®. The DSRU makes the final decision on the publication of external communications. A. Evans Other: The DSRU is an independent charity (No. 327206) which works in association with the University of Portsmouth. The DSRU has received funding from Bayer, the manufacturer of Xarelto^®. The DSRU makes the final decision on the publication of external communications. F. Coukan Other: The DSRU is an independent charity (No. 327206) which works in association with the University of Portsmouth. The DSRU has received funding from Bayer, the manufacturer of Xarelto^®. The DSRU makes the final decision on the publication of external communications. L. Wise Other: The DSRU is an independent charity (No. 327206) which works in association with the University of Portsmouth. The DSRU has received funding from Bayer, the manufacturer of Xarelto^®. The DSRU makes the final decision on the publication of external communications. S. Shakir Other: The DSRU is an independent charity (No. 327206) which works in association with the University of Portsmouth. The DSRU has received funding from Bayer, the manufacturer of Xarelto^®. The DSRU makes the final decision on the publication of external communications.

100 ISoP18-1143 Distribution of Cha2ds2-Vasc Scores in Patients with Atrial Fibrillation Treated with Rivaroxaban in Primary vs. Secondary Care Settings

100.1 M. Davies^1,2, L. Wise^1,2, S. Shakir^*1,2

100.1.1 ¹Drug Safety Research Unit, Southampton, United Kingdom; ²University of Portsmouth, Portsmouth, United Kingdom

Background/Introduction: Since its incorporation in the European Society of Cardiology guidelines in 2010, the CHA2DS2-VASc score is widely used to characterise the risk of stroke in patients (pts) with atrial fibrillation (AF). It is frequently calculated in pharmacoepidemiological studies through retrospective application of the component criteria included in the risk score. CHA2DS2-VASc Scores of 0, 1, or ≥ 2 indicate low, moderate, or high stroke risk, respectively. 

Objective/Aim: To describe the distribution of CHA₂DS₂-VASc scores and individual stroke risk components in two cohorts of pts prescribed rivaroxaban for AF in primary (1°) vs. secondary (2°) care.

Methods: Two PASS were conducted to investigate the safety of rivaroxaban in pts with AF in 1° care (Modified-Prescription Event Monitoring) and 2° care (Specialist Cohort Event Monitoring) (2012–2016). Baseline characteristics were provided by general practitioners (1° care) and specialist prescribers (2° care) using customised questionnaires. An algorithm was used to compute a CHA₂DS₂-VASc score (0–9) for each pt from fixed response options or open questions, according to published guidelines.

Results: The response rate for baseline questionnaires was lower in 1° care vs. 2° care (22.3 vs. 99.7%). The 1° care cohort consisted of 10,225 pts with a primary indication of AF [median age 78 years (IQR 70–84); 5253 (51.4%) male]. The median CHA₂DS₂-VASc score was 4 (IQR 3–5) reflecting a high risk of stroke. The 2° care cohort consisted of 965 pts with a primary indication of AF [median age 76 years (IQR 69–83); 517 (53.6%) male], with a median CHA₂DS₂-VASc score of 4 (IQR 3–6). There were a higher proportion of pts with a score of 6–9 in 2° care vs. 1° care. The proportions of pts in 1° care vs. 2° care with each criterion were:

• Age 65–74 years (25.7 vs. 25.8%)

• ≥ 75 years (61.8 vs. 58.0%)

• Female (48.6 vs. 46.5%)

• Congestive heart failure/left ventricular dysfunction (14.3 vs. 14.6%)

• History hypertension (82.6 vs. 73.2%)

• History stroke, transient ischaemic attack (TIA) or thromboembolism (TE) (19.8 vs. 46.9%)

• Vascular disease (11.3 vs. 26.9%)

• Diabetes mellitus (17.2 vs. 18.8%)

Conclusion: These results suggest that pts initiated rivaroxaban for AF in 2° care are more likely to have a history of stroke, TIA, TE or vascular disease than pts treated in 1° care. This may mean that pts considered to be at ‘higher risk’ with greater co-morbidities are more likely to be managed in 2° care, or that co-morbidities may be less well recorded in 1° care.

Disclosure of Interest: M. Davies Other: The DSRU is an independent charity (No. 327206) which works in association with the University of Portsmouth. The DSRU has received funding from Bayer, the manufacturer of Xarelto^®. The DSRU makes the final decision on the publication of external communications. L. Wise Other: The DSRU is an independent charity (No. 327206) which works in association with the University of Portsmouth. The DSRU has received funding from Bayer, the manufacturer of Xarelto^®. The DSRU makes the final decision on the publication of external communications. S. Shakir Other: The DSRU is an independent charity (No. 327206) which works in association with the University of Portsmouth. The DSRU has received funding from Bayer, the manufacturer of Xarelto^®. The DSRU makes the final decision on the publication of external communications.

101 ISoP18-1144 An Example of Qualitative Signal Detection Within the French Signal Management Process

101.1 M. Abou Taam¹, R. Youdarene¹, M.-L. Veyries¹, M. Benkebil¹, C. Ferard^*1, P. Maison¹

101.1.1 ¹ANSM, Saint-Denis, France

Background/Introduction: According to Guideline on good pharmacovigilance practices, a signal is an information that is judged to be of sufficient likelihood to justify verificatory action. ANSM management signal process includes among others a qualitative signal detection based on ICSR registered in the french pharmacovigilance database (FPD). Those signals are subject to confirmation during the monthly pharmacovigilance Committee between ANSM and the Regional pharmacovigilance network. The relevance of those ICSR is decided according to a decision algorithm (e.g., unexpected, expected with other criteria, cluster of cases, high-risk population) and a risk analysis. In several cases, the signal should be consolidated. In all cases, the signal should be assessed considering all drug related data. If the signal is validated, ANSM or EMA decide then to make actions or measure to consolidate the signal and/or to reduce the risk.

Objective/Aim: To analyze the type of actions or measures which were proposed following a potential signal.

Methods: All actions/measures proposed after validated signal in 2017 were analyzed.

Results: Among the overall 73,991 ICSRs registered in FPD in 2017, 818 (1%) were considered as potential signals. From those signals, the risk analysis allow to detect 531 validated potential signals (65%). In relation to recommendations stemming from those collegial assessments, the majority of actions/measures were mainly to review the data of next PSUR or to discuss the signal through PRAC procedures. In some cases, it was to decide label changes for national authorization procedures and/or additional monitoring including expertise solicitation and all-drug related data retrieval. Less frequently, the decision was to communicate with healthcare professionals and with patients about the risk, to trigger an European signal procedure and to decide a reassessment of the benefit–risk balance at national level.

Conclusion: The signal management of validated signals has led to a number of additional actions or measures. Some potential signals were not validated. However, spontaneous reporting of adverse events continues to play a major role in regulatory decision making. Several other sources of signals are also used including the statistical signal detection. The patient and all stakeholders play an important role in the signal detection.

Disclosure of Interest: None declared.

102 ISoP18-1145 Drug–Drug Interactions Between ß Lactam (Amoxicillin/Clavulanic Acid and Imipenem) and Valproic Acid in the Same Patient

102.1 E. Gaies¹, I. Fazaa¹, R. CharfiI¹, M. Ben SassiI¹, N. Jebabli¹, H. El Jebari¹, I. Salouage¹, S. TrabelsiI¹, R. Daghfous^*1

102.1.1 ¹Clinical Pharmacology, National Centre of Pharmacovigilance, Tunis, Tunisia

Background/Introduction: Valproic acid (VPA) is a commonly used agent for the treatment of both partial and generalized seizure. It can be administrated alone or associated to other antiepileptic drugs (AEDs) and it is known to interact with the most used AEDs and non-AEDs. Concomitant administration of imipenem and VPA leads to a decline in plasma concentrations of VPA, which can sometimes induce seizures [1]. This interaction was previously reported but the mechanism remains unclear. However, interaction with amoxicillin/clavulanic acid (AMC) is reported only in one study evaluating the interaction in healthy volunteers [2].

Objective/Aim: In this context we report a case of drug–drug interactions between AMC and VPA and between imipenem and VPA in the same patient.

Methods: This report describes a 2-year-old male child treated since January 2017 by VPA (Dépakine^®), 400 mg daily (33 mg/kg) for encephalopathy. Therapeutic drug monitoring of VPA was made regularly from January to August 2017 by measuring through plasma concentration (C₀).

Results: VPA concentrations (C₀) varied between 40.61 and 58.16 µg/mL (mean C₀ was 46.61 µg/mL and C₀/Dose ratio = 1.27).

On August 2017, the patient was diagnosed with pneumonia and he was treated initially by AMC (Augmentin^®). Few days later, AMC was changed to imipenem.

One day after AMC administration, the plasma level of VPA was rapidly decreased to 32.88 µg/mL. VPA dose was adjusted to 660 mg daily but the concentration remained low. With the administration of imipenem, VPA dose was adjusted to 800 mg daily (66 mg/kg). Plasma level of VPA were monitored more closely during the imipenem treatment period, and showed C₀ mean of VPA near the therapeutic range (51.65 µg/mL) but a low Co/Dose ratios (Mean ratio 0.87).

Conclusion:

There is a complex interaction between VPA and carbapenem antibiotics due to a combination of different mechanisms, including absorption, distribution, and metabolism [1]. The interaction between AMC and VPA seems rare but its mechanism is clear and involves interruption of the enteroheptic circulation by AMC. So, if concomitant ß lactam therapy with VPA is needed and cannot be avoided, VPA concentration should be monitored more frequently.

References:

• 1. Mancl EE, Gidal BE. The effect of carbapenem antibiotics on plasma concentrations of valproic acid. Ann Pharmacother 2009;43:2082–7

• 2. Lee S-Y, Huh W, Jung JA, Yoo HM, Ko J-W, Kim J-R. Effects of amoxicillin/clavulanic acid on the pharmacokinetics of valproic acid. Drug Des Devel Ther 2015;9:4559–63

Disclosure of Interest: None declared.

103 ISoP18-1146 Alopecia Occurred With Infliximab: What Link? Clinical Case

103.1 I. Bennani¹, R. Soulaymani², A. Tebaa^*2

103.1.1 ¹Faculty of Medicine and Pharmacy of Rabat, University Mohamed V Rabat, Morocco; ²Pharmacovigilance Departement, Poison and Pharmacovigilance center of Morocco, Rabat, Morocco

Background/Introduction: The increased use of biologic drugs has been revealing new adverse effects. The cutaneous reactions described include eczema, erythema, urticaria, lupus-like syndrome and, paradoxically, psoriasis. One cases of alopecia associated with anti-TNF therapy were reported, which resulted in cutaneous psoriasiform lesions. One case of alopecia associated with anti-TNF therapy has been reported at the National Center for Pharmacovigilance in Morocco, which is covered in this communication.

Objective/Aim: The aim of this work is to determine the causality assessment between alopecia and the administration of the Influximab.

Methods: It is a women of 60-year-old, hospitalized for ankylosing spondylitis on 2016, treated by Sulfasalazine and AINS in front of the therapeutical echec with sulfasalazine, she was switched to Infliximab. After the 5th cure/cycle, the patient developed alopecia mainly involving the scalp. The drug was stopped and all etiological causes were eliminated with a positif rechallenge. The relationship between alopecia and Infliximab was suspected and evaluated using the French method of imputability.

Results: Biotherapy represents therapeutic advances in the treatment of immunological and inflammatory diseases.

The accountability score was I5B4 with an informativity level of 2 using the French method In our patient, we were tempted to retain Infliximab as responsible for alopecia. Alopecia is a notorious effect already described in the literature.

However, the Cases notified at the level of the national pharmacovigilance center concerning infliximab which are 48 cases; Our case is the first case of alopecia caused by infiliximab.

According to the literature, the development of alopecia related to anti-TNF is a possible although seldom reported collateral effect. Alopecia is a rare cutaneous reaction to tumor necrosis factor alpha antagonists. This reaction often reverses with discontinuation of the offending drug and initiation of topical treatments; however, irreversible hair loss may occur if a scarring alopecia develops.

Conclusion: the decision to discontinue anti-TNF therapy following adverse skin reactions should be based on a case-by-case assessment of the benefit/risk and benefit/cost ratio of the prolonged treatment.

Disclosure of Interest: None declared.

104 ISoP18-1149 Impact Analyses of European Pharmacovigilance Interventions on Public Health Burden

104.1 S. Lane¹, E. Lynn^1,2, J. Slattery³, S. Shakir^*1,2

104.1.1 ¹Drug Safety Research Unit, Southampton, United Kindgom; ²University of Portsmouth, Portsmouth, United Kingdom; ³European Medicines Agency, London, United Kingdom

Background/Introduction: Adverse drug reactions (ADRs) account for 5% hospital admissions and 197,000 deaths each year in the EU, amounting to a societal cost of €79billion [1, 2]. These ADRs have large impacts on public health in terms of morbidity and mortality. Since the implementation of the 2012 pharmacovigilance (PV) legislation, monitoring success of pharmacovigilance (PV) has become commonplace; however impact of PV regulatory interventions on public health remains mostly unquantified [3]. There is a current need for research identifying the best technique for measuring impact of these actions, and guidelines on reporting impact analyses [3].

Objective/Aim: Interim analysis to assess the feasibility of predictive modelling techniques for estimation of public health impact of PV regulatory interventions.

Methods: Prescription products used in primary care whose marketing authorisations were withdrawn, revoked or suspended in France, Germany or the UK between 2012 and 2016 were previously identified [4]. Databases from France, Germany and the British Health Improvement Network (THIN) were used to estimate usage of the identified products for the year prior to PV action. Systematic searches of the PubMed/MEDLINE database, and European Medicines Agency and national competent authority websites and documents were conducted to identify quantitative studies for the product and ADR of interest. Study data were meta-analysed to produce an overall weighted odds ratio for each ADR relating to each product. Background risk was estimated using THIN data. Modelling of usage figures, odds ratios and background risk allowed the public health impact of the intervention to be estimated in terms of morbidity attributable to the withdrawn, revoked or suspended product.

Results: 18 products were considered for impact analysis; 9 were excluded as no quantitative studies were identified for the respective ADRs. This interim analysis on a subset of these products provides a prediction of the number of ADRs avoided as a result of the included products’ marketing authorisation withdrawal, revocation or suspension, and an estimation of the public health impact of these PV actions using changes in morbidity as an indicator.

Conclusion: This study identifies a method for predicting public health impact of PV interventions, based on drug utilisation data and expected changes in morbidity. This interim analysis tested the method on a subset of products. Further validation of the predictive modelling method will be undertaken before completion of the study.

References:

• 1. Mammi M, Citraro R, Torcasio G, Cusato G, Palleria C, di Paola ED. Pharmacovigilance in pharmaceutical companies: an overview. J Pharmacol Pharmacother 2013;4(Suppl 1):S33–7

• 2. Mazzitello C, Esposito S, De Francesco AE, Capuano A, Russo E, De Sarro G. Pharmacovigilance in Italy: An overview. J Pharmacol Pharmacother 2013;4(Suppl 1):S20–8

• 3. Goedecke T, Morales D, Pacurariu A, Kurz X. Measuring the impact of medicines regulatory interventions—systematic review and methodological considerations. Br J Clin Pharmacol 2018;84:419–33

• 4. Lane S, Lynn E, Shakir S. Investigation assessing the publicly available evidence supporting postmarketing withdrawals, revocations and suspensions of marketing authorisations in the EU since 2012. BMJ Open. 2018;8:e019759

Disclosure of Interest: None declared.

105 ISoP18-1150 Changes in Evidence Supporting Post-marketing Withdrawal of Marketing Authorisation in the EU

105.1 S. Lane¹, E. Lynn^1,2, S. Shakir^*1,2

105.1.1 ¹Drug Safety Research Unit, Southampton, United Kingdom; ²University of Portsmouth, Portsmouth, United Kingdom

Background/Introduction: Market withdrawal of a product affects patients, healthcare professionals, regulators and manufacturers, therefore these decisions should be based on the most robust evidence available [1]. EU pharmacovigilance (PV) legislation was updated in 2012 and led to modernisation of PV systems in Europe. It is necessary to explore whether evolution of PV leads to changes in evidence supporting PV actions (e.g. withdrawals), and how the use of supporting evidence has adapted over time.

Objective/Aim: To evaluate changes in evidence supporting post-marketing product withdrawals from 1999 to 2016.

Methods: Using 3 previously published studies, it was possible to compare evidence used to support post-marketing PV actions in the EU, spanning the period 1999–2016 [1–3]. The number of withdrawals per year was calculated. Study designs supporting post-marketing withdrawals in 1999–2016 were evaluated. Changes to evidence supporting product withdrawals were assessed.

Results: A total of 31 market withdrawals occurred in the EU between 1999 and 2016 and were included in these publications, with an average of 1.72 per year. The number of withdrawals increased between 1999 to its peak in 2010 (5 withdrawals in 2010), and decreased from 2011 to 2016. In general the number of different study designs used to justify withdrawals increased between 1999 and 2016. Spontaneous and published case reports were the most common evidence source, involved in 29 withdrawals (93.5%). This evidence alone led to 4 (12.9%) withdrawals. The least common sources of supporting evidence were epidemiological studies (case–control n = 6; cohort n = 7). Other sources included randomised clinical trials (n = 20), animal studies (n = 13) and meta-analyses (n = 7). The average number of distinct evidence types used was 3.1 per withdrawal. Six (19.4%) withdrawals were justified by just one source of evidence, with the remaining 25 (80.6%) withdrawals supported by at least 2 different types of evidence.

Conclusion: The number of products withdrawn from the EU market appears to have increased until 2010, and has since decreased. Although there is an apparent increase in different types of evidence being used to justify post-marketing withdrawals, spontaneous and published case reports remain a vital source of evidence for regulators.

References:

• 1. Clarke A, Deeks JJ, Shakir SA. An assessment of the publicly disseminated evidence of safety used in decisions to withdraw medicinal products from the UK and US markets. Drug Saf 2006;29:175–81

• 2. Lane S, Lynn E, Shakir SAW. Measuring the impact of product withdrawals and other major pharmacovigilance actions on public health burden in the EU: methodological considerations. Drug Saf 2017; 40: 953–4

• 3. McNaughton R, Huet G, Shakir S. An investigation into drug products withdrawn from the EU market between 2002 and 2011 for safety reasons and the evidence used to support the decision-making. BMJ Open. 2014;4:e004221.

Disclosure of Interest: None declared.

106 ISoP18-1151 Psychiatric Adverse Drug Reactions with Anti-Hypertensive Treatment: Review and Analysis of SmPCs

106.1 P. Biswas^*1, H. Biswas¹

106.1.1 ¹Symogen Limited, Bourne End, United Kingdom

Background/Introduction: Prescription drugs result in myriad of ADRs, including psychiatric ADRs. Psychiatric effects of drugs impact sensorium, attention, concentration, memory and higher cognitive functions resulting in distressing physical illness and psychiatric consultation. In regards to psychiatric effects associated with drug therapy, complications such as depression and suicidality are those most frequently reported. 

Objective/Aim: To review the SmPCs of Antihypertensive drugs used commonly in medical practice and understand the common, uncommon, rare, very rare psychiatric ADRs and risks associated with these drugs.

Methods: List of most common antihypertensives prescribed in medical practice was prepared and review of Section 4.8 Undesireable effects of the SmPCs undertaken. Three drugs each from treatment categories were included for analysis: Angiotensin Receptor Blockers (ARBs):Losartan, Valsartan, Candesartan; ACE Inhibitors: Enalapril, Lisinopril, Ramipril; Beta Blockers; Atenolol, Metoprolol, Carvedilol; Calcium Channel Blockers (CCB): Amlodipine, Diltiazem, Verapamil; Diuretics: Furosemide, Bumetanide, Spironolactone; Alpha Blockers : Prazosin, Terzosin, and others: Clonidine and Minoxidil.

Results:

Analysis involved review of common, uncommon, rare and very rare psychiatric ADRs based on the Pyschiatric System Organ Class (SOC) ranked under headings of frequency as described in Table 1. Almost all the drugs had psychiatric ADRs ranging from common to very rare listed in SmPCs except Valsartan, Candesartan in ACE Inhibitors; Verapamil in CCBs; Bumetanide in Diuretics and Minoxidil. Depression, depressed mood, anxiety, nervousness were common for Alpha and Beta blockers which are the most prescribed antihypertensives. Insomia, sleep disturbances, confusion, psychosis were uncommon, while hallucinations, mental confusion, disturbance in attention, depression and personality disorders were rare and very rare ADRs. None of the drug categories listed suicidality or suicide attempt as an ADR.

Conclusion: Identification of psychiatric ADRs related to drug therapy in real world post-marketing setting is difficult. Reporting rates determined on the basis of spontaneously reported post-marketing ADRs are generally presumed to underestimate the risks associated with treatment. As a medical student, this exercise will help me to assess patients starting therapy with antihypertensives, identify and report ADRs to the MHRA, monitor ongoing ADRs, and lastly read the SmPC as guidance to good medical practice.

Disclosure of Interest: None declared.

107 ISoP18-1152 Is the Risk of Linezolid to Cause Serotonin Syndrome Real in Routine Clinical Practice?

107.1 H. Thirot¹, X. Holemans², F. Jacobs³, C. Briquet⁴, F. Frippiat⁵, A. Spinewine⁶, F. Van Bambeke¹, P. M. Tulkens^*1

107.1.1 ¹Louvain Drug Research Institute, Université catholique de Louvain, Bruxelles, Belgium; ²Médecine Interne générale et infectiologie, Grand Hôpital de Charleroi, Charleroi, Belgium; ³Infectiologie, CUB-Hôpital Erasme, Bruxelles Belgium; ⁴Pharmacie, Cliniques universitaires St-Luc, Bruxelles, Belgium; ⁵Infectiologie, Centre Hospitalo-Universitaire de Liège, Liège, Belgium; ⁶Clinical Pharmacy, Cliniques Universitaires St-Luc, Bruxelles, Belgium

Background/Introduction: Because of structural similarities, linezolid is an inhibitor of monoamine oxidases [1], which may lead to a serotonin syndrome [2, 3] and toxicities in case of co-administration of drugs inhibiting the recapture of serotonin (such as several antidepressants or antihypertensives), drugs undergoing catabolism through monoamine oxidase activity, or direct inhibitors of monoamine oxidase, all of which are then considered as contra-indicated [4, 5]. However, the occurrence of real cases of serotonin syndrome and/or of drug–drug interaction toxicities during routine linezolid treatment of patients receiving contra-indicated drugs remains unsettled [5], especially if considering that linezolid is often prescribed for longer periods and for other indications than recommended in the SmPC.

Objective/Aim: To better document the risk of developing a serotonin syndrome in patients receiving routine linezolid treatments (including off-label uses [6]) and prescribed drugs considered as contra-indicated because of potential toxicities due to linezolid-induced inhibition of monoamine oxidase(s).

Methods: We retrospectively analysed the medical files of hospitalized adult (> 18 years) patients receiving linezolid for treatment of documented infection(s) and assessed their exposure to other drugs based on analysis of pharmacy records. The study covered 4 large Belgian hospitals for a period of 1 year (2016).

Results: We enrolled 236 patients (248 treatments) who received linezolid at the recommended dose (600 mg BID) during a mean period of 14.8 days (median: 10 days; extremes: 1–90 days). Among those, 100 patients were prescribed one or several drugs susceptible to cause serotonin syndrome [tramadol (n = 53), escitalopram (n = 19), trazodone (n = 15), mirtazapine (n = 11), duloxetine (n = 8), amitriptyline (n = 4), venlafaxine (n = 3), sertraline (n = 2), paroxetine (n = 2), melitracene (n = 1)] but no drug acting through monoamine oxidase inhibition. Of all these patients, only one [receiving trazodone (100 mg/day) and duloxetine (60 mg/day)] developed a clinically confirmed serotonin syndrome (with agitation and delirium) after 7 days of treatment with linezolid, leading to the antibiotic withdrawal and recovery within a week.

Conclusion: The documentation of a serotonin syndrome was very low in this cohort of patients receiving linezolid with co-prescription of one or several drugs commonly considered as contra-indicated. The co-administration of linezolid with these drugs may therefore be safer than anticipated in routine clinical practice. However the retrospective design of the study and the common under-diagnosis of serotonin syndrome should trigger the launching of a prospective study where actual drug exposures, clinical status, and risk factors could be controlled and assessed.

References:

• 1. Taylor JJ, Wilson JW and Estes LL. Linezolid and serotonergic drug interactions: a retrospective survey. Clin Infect Dis 2006;43:180–7

• 2. Frykberg RG, Gordon S, Tierney E and Banks J. Linezolid-associated serotonin syndrome. A report of two cases. J Am Podiatr Med Assoc 2015;105:244–8

• 3. Sutton J, Stroup J, Som M. Linezolid-induced serotonin toxicity in a patient not taking monoamine oxidase inhibitors or serotonin receptor antagonists. Proc (Bayl Univ Med Cent) 2016;29:214–5

• 4. Preston CLe (2016) Stockley’s Drug Interactions. Pharmaceutical Press, London, UK

• 5. Linezolid SmPC. Linézolide: Résumé des caractéristiques du produit. Agence fédérale des médicaments et produits de santé. Available from: http://bijsluiters.fagg-afmps.be/DownloadLeafletServlet?id=115863 Last updated: 2014; last accessed: 25-05-2018

• 6. Thirot H, Briquet C, Frippiat F, Jacobs F, Holemans X, Tulkens PM, et al. Review of linezolid (LZD) use and onset of toxicity in 4 belgian hospital centers: a retrospective study. Abstract submitted to the 2018 IDweek, San Diego, CA (3–7 Oct 2018)

Disclosure of Interest: H. Thirot: None Declared, X. Holemans: None Declared, F. Jacobs: None Declared, C. Briquet: None Declared, F. Frippiat: None Declared, A. Spinewine: None Declared, F. Van Bambeke: None declared, P. Tulkens Speaker bureau of: Merck.

108 ISoP18-1153 A Prospective Evaluation of MCEM Method for Drug Safety Signal Detection in Spontaneous Reporting Systems

108.1 Y. Li^*1, C. Xiao², K. Shen³, K. Bannout⁴, W. Wallis⁴, F. Wang^1,5

108.1.1 ¹Center for Computational Health, IBM Research, Yorktown Heights, NY, United States; ²AI for Healthcare, IBM Research, Cambridge, MA, United States; ³IBM Watson Health, Life Science Division, New York, NY, United States; ⁴Global Safety and Risk Management, Celgene, Summit, NJ, United States; ⁵Department of Healthcare Policy and Research, Cornell University, New York, United States

Background/Introduction: Modern signal detection and statistical data-mining methods are used to identify new drug safety signals based on spontaneous reporting systems (SRS). Bayesian methods, such as Multi-Item Gamma Poisson Shrinker (MPGS) [1], are frequently used on such databases to correct for the sampling variance. Although a significant advance, the Bayesian methods do not address other limitations intrinsic to SRS such as confounding effect. We have developed the Monte-Carlo Expectation Maximization (MCEM) method [2] that reduces sampling variance by leveraging the MGPS technique [1], and filters out concomitant confounders in each case report by leveraging a Monte Carlo sampling procedure that could assign each ADR with its major associated drug determined by drugs’ contribution to that ADR [2]. We have previously demonstrated the effectiveness of MCEM according to a reference standard developed by the Observational Medical Outcomes Partnership (OMOP) [2–4]. This evaluation is considered retrospective as most drug-ADR positive pairings in the reference standard were well established at the time of evaluation. In this study, we sought to test the method’s performance in the prospective detection of emerging and unknown ADRs.

Objective/Aim: To test whether MCEM applied prospectively is able to detect the unknown ADRs earlier than the traditional MGPS method.

Methods: We tested the MCEM and the MGPS method using the de-duplicated and standardized (e.g., drug names mapped to RxNorm concepts and ADR outcomes mapped to MedDRA concepts) FDA adverse event reporting system (FAERS) [5]. We divided the data into six cumulative yearly subsets based on the reporting time of each case report from 2007 through 2012. Each successive set consists of reports from that year plus those from preceding years. For example, the 2007 data set consists of reports from only 2007, and 2012 data set consists of all reports from 2007 through 2012. We used a time-indexed reference standard involving 44 drugs and 38 ADRs, which resulted in 62 positive test cases time-stamped with the date on which the relevant ADR was added to the label by the FDA throughout the calendar year 2013, and 75 negative controls [6]. Positive signals were those whose signal scores greater than or equal to the particular threshold that was determined by the pre-specified false discovery rate. Two performance metrics were evaluated: (1) area under the Receiver Operating Characteristics (ROC) curve (AUC) for the six cumulative yearly data sets, and (2) the mean lead time-to-signals (MLT2D) for both MCEM and MGPS.

Results: The MCEM achieved better AUCs from 2008 to 2011, and worse AUCs in 2007 and 2012 than the MGPS. The MCEM was able to identify 19, 24 and 24 true positive signals before 2013 at false discovery rates of 5, 10 and 15%, respectively, and the MLT2D were 2.20, 2.66 and 2.73 years. This compared with that the MGPS that identified 18, 19 and 20 true positive signals at the same levels of false discovery rates, and the MLT2D were 1.95, 2.07 and 2.18 years.

Conclusion: Within the framework of this evaluation, our findings suggest that MCEM leads to earlier signal detection than MGPS. Acknowledging the limitations associated with the data and reference standard used in this work, the MCEM is a promising alternative to MGPS that merits further investigation and validation.

References:

• 1. DuMouchel W. Bayesian data mining in large frequency tables, with an application to the FDA spontaneous reporting system. Am Stat 1999;53:177–90

• 2. Xiao C, Li Y, Baytas IM, Zhou J, Wang F. An MCEM framework for drug safety signal detection and combination from heterogeneous real world evidence. Sci Rep 2018;8:1806

• 3. Ryan PB, Schuemie MJ, Welebob E, Duke J, Valentine S, Hartzema AG. Defining a reference set to support methodological research in drug safety. Drug Saf 2013;36:33–47

• 4. Xiao C, Li Y, Argentinis E, Zhou J, Wang. An MCEM-MTL framework for drug safety signal filtering and detection in spontaneous reporting systems. Drug Saf 2017; 40: 944–5

• 5. Banda JM, Evans L, Vanguri RS, Tatonetti NP, Ryan PB, Shah NH. A curated and standardized adverse drug event resource to accelerate drug safety research. Sci Data 2016;3:160026

• 6. Harpaz R, Odgers D, Gaskin G, et al. A time-indexed reference standard of adverse drug reactions. Sci Data 2014;1:140043

Disclosure of Interest: Y. Li Employee of: IBM Research, C. Xiao Employee of: Celgene, K. Shen Employee of: IBM Watson Health, K. Bannout Employee of: Celgene, W. Wallis Employee of: Celgene, F. Wang Employee of: IBM Research, Other: Assistant Professor at Cornell University.

109 ISoP18-1154 Amusing Ourselves to Death: How Science and Reason are Losing Their Grip in Public Discourse

109.1 B.P.J. Hugman^*1

109.1.1 ¹Global Communications, Uppsala Monitoring Centre, Sweden, Oxford, United Kingdom

Background/Introduction: This new work builds on last year’s study of alternative facts and the threats to the authority of evidence, particularly in medicine. The politicisation of scientific and medical issues and those associated with women’s health and welfare, for example, have led to partisan battles and polarisation which preclude engagement, negotiation and consensus. These problems reflect the wider social and political landscape in which measured debate across tribal boundaries is becoming increasingly rare, if not impossible, and in which policy is driven by idiosyncratic preferences and emotion, not evidence. The title of this proposal is an acknowledgement of Neil Postman’s prescient 1985 book, Amusing Ourselves to Death: Public Discourse in the Age of Show Business. He argued that television was bringing about a realisation of Huxley’s nightmare society in which populations were oppressed and infantilised by their addiction to amusement; in the twenty-first century, popular technology has dramatically intensified those processes, while opening up unprecedented access to human attention.

Objective/Aim: (1) To analyse some of the major current political and cultural trends across the world, including tribalism, polarisation, partisanship, binary thinking, and public cynicism about experts, and (2) To suggest novel ways in which actors in science, medicine and public health, and activists, might respond in order to be heard and for their messages to influence opinion and behaviour.

Methods: The work is part of a larger project examining how medicines information and risk communication can be deployed to stem the tide of irrationality and recover from their failure to deliver full benefits in an increasingly perverse political, social and philosophical context. It has involved extensive reading and research and a degree of original thinking and synthesis.

Results: The analysis reveals the depth of the problems we face and the extent to which they are inter-related with profound social and political currents in recent years. Scientists themselves have contributed through failures and shortcomings in scientific practice and public relations. The solutions are radical and demanding and require commitment and creativity in broad alliances.

Conclusion: It is clear that ‘the institutional structure of science advocacy has created a conservative bystander culture that has failed science abysmally’ [2]; that our methods are often sluggish and bureaucratic; that only an era of radical new communication skills and engagement will begin to diminish the power of superstition, division and unreason. Some ideas for starting this reform, and some speculative methods for pushing it forward, in collaboration with others, will be offered.

References:

• 1. Postman N. Amusing ourselves to death: public discourse in the age of show business. Penguin Books; Anniversary edition, 2005

• 2. Tran L. Why March for science? Because when it is attacked, only the elite benefit. The Guardian, UK, 22 Apr 2017

Disclosure of Interest: None declared.

110 ISoP18-1155 Pharmacovigilance: Inspection and Audit—Are they the Same or Different?

110.1 M. Yaman^*1, E. Canak², S. Dagistanli^3,4

110.1.1 ¹Pharmacovigilance, ²Business Development, ³DeltaPV Ilac Danismanlik Saglik Urun ve Hizmet A.S., Istanbul, Turkey; ⁴Turkish Pharmacovigilance Foundation, Istanbul, Turkey

Background/Introduction: Pharmacovigilance (PV) personnel should always be ready for health authorithy inspections, as well as audits by internal auditors, headquarter PV functions, licensor companies, independent audit companies and others at any time. Pharmacovigilance World currently has many guidelines and written procedures released by different countries to be followed by pharmaceutical companies. They are released in order to manage PV operational system and improve input on benefit/harm perceptions from obtained safety data. PV audits and inspections are conducted to review and analyze the compliance of the systems with these guidelines and procedures which in turn support patient safety.

Objective/Aim: To identify similarities and differences between PV inspections and audits, by sharing experiences from a Contracted Pharmacovigilance Service Company’s view.

Methods: Within the time frame starting from 01 April 2014 to 01 June 2018, a total of 7 inspections conducted by Turkish Health Authorithy and 58 audits were included in our analysis; 16 of these were local affiliate (Contracted Company’s Local PV function) audits, 42 were global (Contracted Company’s Headquarter PV function) and licensor company audits. In terms of methodological approach, a total of 41 face to face audits, 11 questionnaire based audits via e-mail and 6 remote connection audits via videoconference were evaluated in our study. In terms of categories of findings, a total of 1 critical, 26 major and 115 minor finding were identified during this period and they were all closed within 1 year at the latest from inspection or audit date.

Results: All inspections conducted by the Turkish Health Authority Inspection Division were face to face and the longest inspection conducted lasted 5 days by five inspectors in 2016. Majority of other audits and inspections lasted between one and 2 days. In this study, the major observation is that some confusion may occur about the purpose of inspections and audits or the terms may be used interchangeably without realizing that they have many different aspects. Audits are considered to guide a “self check” to ensure compliance in PV system. However, the inspections are considered to be a check list of “have to do’s” of the specified compliance obligations have been met. In Turkey, in cases of noncompliance, various enforcement actions can be forced in the result of Turkish Health Authority Inspections on the pharmaceutical companies and contracted pharmacovigilance service companies by the Turkish Health Authority. Commonly, inspectors and auditors review similar tasks of PV operations; such as quality management systems, agreements, internal and external trainings, procedures and all other PV related documents which are available for data collection.

Conclusion: In this reporting period the impact of the revised pharmacovigilance legislation on pharmaceutical companies and contracted pharmacovigilance service companies in Turkey became more apparent. After the legislation and guidance has been strengthened, an increased number of inspection and audits have been recorded. The PV personnel should always be ready for an inspection and/or audit, since patient safety is considered as priority along with Good Pharmacovigilance Practices.

References:

• 1. Khurana A, Rastogi R, Gamperl HJ. Ready for Pharmacovigilance Inspection—USFDA. Int. J Pharm Sci Rev Res 2015;35:210–7

• 2. EMA. Guideline on good pharmacovigilance practices (GVP), Module III—Pharmacovigilance Inspections, EMA/119871/2012 Rev 1*; 08 September 2014. http://www.ema.europa.eu/docs/en_GB/document_library/Scientific_guideline/2014/09/WC500172400.pdf

• 3. EMA. Guideline on good pharmacovigilance practices (GVP), Module IV—Pharmacovigilance audits, EMA/228028/2012 Rev 1*; 03 August 2015. http://www.ema.europa.eu/docs/en_GB/document_library/Scientific_guideline/2015/08/WC500191778.pdf

• 4. MHRA. Pharmacovigilance Inspection Metrics Report. April 2013–March 2014

• 5. https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/418323/Pharmacovigilance_Inspection_Metrics_Report_2013-2014.pdf

Disclosure of Interest: M. Yaman Employee of: DeltaPV Ilac Danismanlik Saglik Urun ve Hizmet A.S., E. Canak Employee of: DeltaPV Ilac Danismanlik Saglik Urun ve Hizmet A.S., S. Dagistanli Shareholder of: DeltaPV Ilac Danismanlik Saglik Urun ve Hizmet A.S.

111 ISoP18-1156 From Database to Diagnosis: ‘Intelligent Query’, a Tool to Help With Safety Signal Evaluation

111.1 V. Samal^*1, O. Mahaux¹

111.1.1 ¹GSK, Wavre, Belgium

Background/Introduction: Safety signal evaluation is a core activity in pharmacovigilance. The methods and tools rely on adverse events coded using the Medical Dictionary for Regulatory Activities (MedDRA). Standardized MedDRA Queries (SMQs) represent the current practice to identify groups of adverse event terms (e.g. signs and symptoms) for a particular medical condition. However, missing patient history in spontaneous report databases is an issue. We have developed an algorithm, the ‘Intelligent Query Bayesian Network’ (IQ BayNet), to ascertain the diagnosis of adverse event reports with more specificity and sensitivity by using computed probabilities of natural history of a disease to predict and complement missing case information.

Objective/Aim: To validate the IQ BayNet as a reliable diagnostic tool for spontaneous adverse event reports of Diabetes Mellitus Type II (DMII).

Methods: A relevant list of variables was identified from GSK’s spontaneous safety database, including laboratory results, clinical features, risk factors and medical history. A Bayesian network software (Hugin) was used to develop a graphical model representing the probabilistic relationships between those variables and DMII. Variable-specific probabilities were attributed based on their frequency of occurrence using published peer-reviewed articles, standard algorithms and GSK physicians’ experience that determines their degree of influence on DMII. The IQ BayNet uses specific variables from the safety database as input and provides a probability for diagnosis as output. A sample of 50 DMII cases was output using IQ BayNet and qualitatively reviewed. The output DMII probabilities of those 50 cases were compared with the results of MedDRA SMQ Hyperglycaemia/new onset diabetes mellitus.

Results: The IQ BayNet sensitivity and specificity are observed to be increased when compared to SMQ by 48 and 22.6%, respectively. Eighteen DMII positive cases were detected by the IQ BayNet, all being true positive cases, whereas 13 true positive cases were identified among the 25 positive cases detected by SMQ. Further 32 negative cases were detected by the IQ BayNet, 29 being true negative cases, while 17 true negative cases were identified in the 25 negative cases detected by SMQ.

Conclusion: The IQ BayNet has shown higher sensitivity and specificity in automatic detection of true cases of DMII cases as compared to the SMQ. Future work will focus on developing similar algorithms tailored to other medical conditions.

Funding source: GlaxoSmithKline Biologicals SA.

Disclosure of Interest: V. Samal Shareholder of: GSK group of companies, Employee of: GSK group of companies, O. Mahaux Employee of: GSK group of companies.

112 ISoP18-1158 Adverse Event Recognition in Tweets: Results from a WEB-RADAR Project

112.1 L. Gattepaille^*1, S. Hedfors Vidlin¹, T. Bergvall¹, J. Ellenius¹

112.1.1 ¹Uppsala Monitoring Centre, Uppsala, Sweden

Background/Introduction: Thanks to the large patient coverage and instant availability of open data, pharmacovigilance in Twitter could complement more traditional tools such as spontaneous case reports and electronic health records [1]. In 2014, The WEB-RADR consortium [2] was established to, among other aims, provide an answer to the open question: can social media data be harnessed and reliably utilized for pharmacovigilance purposes?

Objective/Aim: One year after completion of the WEB-RADR consortium, we report the results of a processing pipeline aiming at recognizing Adverse Events in Tweets, developed in the course of the project.

Methods: First, a relevance filter was applied to remove tweets with low probability of containing an adverse event (AE), using a previously published method [3]. Medicinal product detection used a dictionary lookup method based on WHODrug, a global dictionary of medicinal products. Medical event detection used a mix of dictionary lookups, natural language processing and logistic regression classifiers to map the identified text to the Medical Dictionary for Regulatory Activities preferred terms. Finally, a logistic regression classifier was trained to identify AE relationships.

This pipeline was trained on a proprietary dataset of 196,533 tweets manually annotated by MedDRA coders at Epidemico, a WEB-RADR consortium member, and evaluated on an independent benchmark Reference dataset of 57,481 manually annotated tweets produced by WEB-RADR. We present comparative performance results of the pipeline on the WEB-RADR Reference dataset with a previously published method [3] that has reported good performance on similar data.

Results: The Reference dataset contains 1058 posts with at least one medicinal product—AE pair (MP/AE pair) and a total of 1398 MP/AE pairs. Our pipeline could recognize and correctly map 23% of the 1398 gold standard pairs (recall) while 37% of putative MP/AE pairs identified by the pipeline were indeed gold standard pairs (precision), leading to an F-score of 0.28. 36% of the gold standard pairs were lost in the relevance filter module, 54% of the remaining pairs were lost in the medicinal product and medical event recognition modules combined and 35% of the remaining pairs were lost in the AE relation classification module, explaining the low overall recall. Comparatively, the previously published method achieved a recall of 0.32 and a precision of 0.18 (F-score 0.23) on the same dataset, despite a good published performance on another dataset (0.86 recall, 0.72 precision and 0.78 F-score).

Conclusion: This study highlights two major challenges with developing methods of automatic recognition of adverse events in Twitter posts: (1) detecting and normalizing medicinal products and medical events is difficult in Twitter posts, most likely owing to the noisiness of the data [e.g. abbreviations, misspellings, slang and laymen expressions (TB1)], (2) the transferability of models outside the universe of the training data to external datasets is poor, despite the use of a training/validation/test setup. The latter difficulty is poorly understood and should be the object of more research, to investigate the true ability of AE recognition algorithms to harness social media data.

References:

• 1. Sarker A, Ginn R, Nikfarjam A, O’Connor K, Smith K, Jayaraman S, et al. Utilizing social media data for pharmacovigilance: A review. J Biomed Inform 2015;54:202–2

• 2. https://web-radr.eu/about-us/

• 3. Freifeld CC, Brownstein JS, Menone CM, Bao W, Filice R, Kass-Hout T, et al. Digital drug safety surveillance: monitoring pharmaceutical products in twitter. Drug Saf 2014;37:343–50

Disclosure of Interest: None declared.

113 ISoP18-1160 A Cross-Sectional Study of Product and Batch Traceability for Biologics in Clinical Practice and Adverse Drug Reaction Reporting in the United Kingdom

113.1 K. Klein¹, L. Hazell², P. Stolk¹, S. Shakir^*2

113.1.1 ¹Lygature, Utrecht, The Netherlands; ²Drug Safety Research Unit, Southampton, United Kingdom

Background/Introduction: Biological medical products (‘biologics’) are derived from living cells or organisms. Due to their complex molecular structures and sensitivity to changes in the manufacturing process, variations may exist between biologics with the same active substance from different manufactures (e.g. ‘biosimilars’) or between different batches of the same product. It is of vital importance that biologics are traceable by brand name and batch number when reporting an adverse drug reaction (ADR) but such product details are not always provided [1]. In a recent study in the Netherlands, the brand name was identifiable in 76% of ADR reports submitted for biologics, whereas only 5% contained a batch number [2]. Shortcomings in the recording and tracing of product and batch information in clinical practice are believed to be associated with the limited traceability of biologics found in ADR reports.

Objective/Aim: The UK BIO-TRAC study was initiated to assess the current status of traceability of brand names and batch numbers for (recombinant) biologics in UK clinical practice within the hospital setting and in the UK ADR reporting database (Yellow Card Scheme).

Methods: An on-line survey has been set up for hospital pharmacists in England to capture information on recording practices that contribute to the traceability of brand names and batch numbers for biologics in UK clinical practice and to gather the views and experiences of hospital pharmacists on current bottlenecks and how to improve the current systems. In addition, a comprehensive analysis of 9036 ADR reports for biologics submitted to the UK Yellow Card Scheme between 1st January 2009 and 30th September 2017 will be performed to assess the extent to which brand name or batch number details are included in these reports. This includes a stratified analysis by reporter type, therapeutic product classes, calendar year and other variables.

Results: The results of the UK BIO-TRAC study will be presented in the context of both objectives, the assessment of the recording practices for product specific details in routine clinical practice and the ADR data analysis of the MHRA Yellow Card database.

Conclusion: On completion, this study will provide important insights as to the traceability of biologics in the UK.

Disclaimer:

The study is conducted by DSRU Education and Research Ltd in collaboration with Lygature and is funded by a grant from ABPI. This study is based in part on data obtained under licence from the UK Medicines and Healthcare Products Regulatory Agency. However, the interpretation and conclusions contained in this study are those of the author’s alone.

References:

• 1. Directive 2010/84/EU of the European parliament and of the council of 15 December 2010, amending as regards pharmacovigilance, Directive 2001/83/EC on the Community code relating to medicinal products for human use. Available from: http://ec.europa.eu/health/files/eudralex/vol-1/dir_2010_84/dir_2010_84_en.pdf

• 2. Klein K, Scholl JHG, Vermeer NS, Broekmans AW, Van Puijenbroek EP, De Bruin ML, et al. Traceability of biologics in the Netherlands: an analysis of information-recording systems in clinical practice and spontaneous ADR reports. Drug Saf 2016;39:185–92

Disclosure of Interest: None declared.

114 ISoP18-1161 Online Availability of Regulatory Documents, Safety Information and Adverse Drug Reaction Reporting n African Countries

114.1 H. Dominicus^*1,2, J. Doua^2,3, T. Verstraeten^2,4

114.1.1 ¹Dominicus Medicus Consultancy, Voorburg, the Netherlands; ²Consortium for African Regulatory expertise Development, Voorburg, The Netherlands; ³Johnson & Johnson, Breda, The Netherlands; ⁴P-95, Heverlee, Belgium

Background/Introduction: All African countries, except one, have established a Drug Regulatory Authority (DRA); together with the African Medicines Regulatory Harmonisation Programme (AMRH), WHO and ISOP [1]. This is an important step. The Internet being the greatest source of information nowadays, it is important that stakeholders have online access to regulatory information, and that safety information can be reported and accessed online.

Objective/Aim: To assess the free accessibility of regulatory documents on DRAs websites; to assess the availability of online safety information; further to assess the possibility for adverse drug reaction (ADR) reporting online; and to delineate gaps and provide recommendations for regulatory measures.

Methods: All African countries are included. A checklist to collect data from each DRA for analysis was developed based on information from the Regulatory Resources for Africa, WHO and the New Partnership for Africa’s Development (NEPAD). Countries were assessed on the online availability (without logging-in) of documents on global drug regulatory legislation and regulations, pharmacovigilance (PV) guidelines, ADR reporting tools, safety communications, and SmPCs. Three investigators scrutinized the internet and completed the list for their part of countries and one of the others re-checked the other’s country data. In case of discrepancy, investigators reached consensus.

Results: For this review, DRA websites of 55 countries were found and assessed. Libya could not be assessed because none of the investigators could read Arabic. Documents on regulatory legislation in general are online accessible in 51% of the countries, pharmacovigilance guidelines in 31%. Of the websites, 27% provides ADR reporting online, and 24% publish online safety warnings (DHPCs and/or other communications). Some countries have limited or no online information. A list of registered products is accessible in 27%; only 1 country (2%) provides summaries of product characteristics.

Conclusion: Most countries do not have pharmacovigilance guidelines or ADR reporting options or safety information online. For the safe use of medicines, our advice is to at least facilitate the online reporting of adverse drug reactions, and to provide online safety information. A list of registered products could be a way to diminish wrong medicine use. Based on our results, African countries, NEPAD, ISOP, and WHO can decide on quick regulatory wins by ameliorating authorities’ websites.

References:

• 1. Ndomondo-Sigonda M, Miot J, Naidoo S, Dodoo A, Kaale E. Medicines regulation in Africa: current state and opportunities Pharm Med 2017;31:383–397

Disclosure of Interest: None declared.

115 ISoP18-1162 Ecstasy-Induced Hepatitis in a Young Adult

115.1 W. Kaabi¹, I. Hamza¹, F. Zgolli¹, S. ben Hammamia¹, N. Alaya¹, R. Daghfous^*1, S. El Aidli¹, S. Kastalli¹

115.1.1 ¹Service de recueil et d’analyses des effets indésirables des médicaments, Centre National de Pharmacovigilance, Tunis, Tunisia

Background/Introduction: Ecstasy (methylenedioxymethamphetamine, MDMA) is widespread drug among youngs. It is used as a recreational drug in festive settings for its stimulating and hallucinogenic properties. Hepatotoxicity is a rare side effect that can range from asymptomatic liver injury to acute hepatic failure.

Objective/Aim: We present a case of hepatotoxicity following exposure to MDMA with favorable clinical outcome.

Methods: This case was notified on March 13th, 2018 to Tunisian National Centre of Pharmacovigilance and validated by Daman et al. method of imputability.

Results: A 35-year old male patient with no medical history was referred to us for jaundice. He was a regular user of “Ecstasy” over 4 months. One week after the last use of “Ecstasy” he reported abdominal pain, nausea, vomiting, weakness and dark urine. Blood tests were as follows: aspartate aminotransferase (AST), 1771 U/L (reference range 10–37 U/L); alanine aminotransferase (ALT), 1301 U/L (10–40 U/L); alkaline phosphatase, 299 U/L (0–270 U/L); total bilirubin, 7.83 mg/dL (0.2–1 mg/dL); direct bilirubin, 7.54 mg/dL (0–0.3 mg/dL); prothrombin rate 45%. Liver enzymes and bilirubin levels were found to be extremely high (AST = 48 × normal, ALT = 33 × normal, and bilirubin = 8× normal). Viral, autoimmune, and metabolic causes were excluded. Serologic tests for hepatitis A, B, and C viruses were all negative. A diagnosis of ecstasy-induced toxic hepatitis was retained. The patient has recovered after a month and laboratory findings has returned to normal values.

Conclusion: Practitioners might need to be alert to the possible adverse events of ecstasy especially liver injury in young adults. Fulminant hepatic failure associated with MDMA has also been described and should warrant attention in case of hepatic failure in young people.

Disclosure of Interest: None declared.

116 ISoP18-1163 Ototoxic Adverse Drug Reactions: A Pharmacovigilance Study Using the Italian Spontaneous Reporting Database

116.1 M. A. Barbieri^*1, G. Cicala¹, P. M. Cutroneo², E. Mocciaro¹, L. Sottosanti³, F. Freni⁴, F. Galletti⁴, V. Arcoraci¹, E. Spina¹

116.1.1 ¹Department of Clinical and Experimental Medicine, University of Messina, Messina, Italy; ²Regional Pharmacovigilance Centre, University Hospital of Messina, Messina, Italy; ³Pharmacovigilance Office, Italian Medicines Agency, Rome, Italy; ⁴Department of Adult and Developmental Human Pathology “Gaetano Barresi”, University of Messina, Messina, Italy

Background/Introduction: Drug ototoxicity can be defined as a temporary or permanent hearing impairment that occurs after a pharmacological treatment [1]. The earliest documentation of ototoxicity regards antimalarials, non–steroidal antiinflammatory drugs, aminoglycosides and other antimicrobials, loop diuretics and antineoplastics [2]. Clinical trials are relevant to evaluate new drugs capable of being ototoxic [3–5]. Nevertheless, the frequency of ear adverse drug reactions (ADRs) and the safety profile of some therapeutic classes remain largely unclear.

Objective/Aim: To perform a descriptive analysis of ototoxic reports using the Italian Spontaneous Reporting System (SRS) Database and to detect any unexpected drug—related ear side effects.

Methods: This study included all reports having at least one ADR related to the System Organ Class (SOC) “ear and labyrinth disorders” recorded in SRS database up to December 31, 2017. All reports with ADRs including the following words “ear”, “hearing” or “auricular” not belonging to the selected SOC were considered. Reports with Preferred Terms (PT) ‘vertigo’ not associated with other auricular diseases were excluded from the analysis because vestibular vertigo could not be discerned from the central ones. Descriptive analyses were carried out to evaluate patient’s characteristics and drug—related variables of selected reports versus the rest of the SRS database. Subsequently, case/non—case method analysis was performed calculating the reporting odds ratio (ROR) as a measure of signals disproportionate reporting (SDRs).

Results: The analysis included 919 reports out of the 381.548 collected into the SRS database. The median age was higher for patients with ototoxic ADRs than for patients with other ADRs (59 vs. 57 years). Reports were almost equally distributed by sex. Overall, 27.2% of ototoxic reports regarded serious ADRs (n = 250). The median number of drugs had no effect in both settings. ADRs (PTs) mostly reported were tinnitus (38.9%) and hypoacusis (23.3%). Antineoplastic and immunomodulating agents and antiinfectives for systemic use were the most involved drugs (n = 260; 28.3% and n = 188; 20.5%, respectively). Among the most reported drugs (n > 20), statistically significantly higher ROR values were observed for antithrombotics, other antineoplastics, quinolones, plant alkaloids, antidepressants, macrolides, antiepileptics, immunostimulants and aminoglycosides.

Conclusion: This study has identified drugs that contributed to developing ototoxic ADRs. In some cases, pharmaceutical classes with significant ROR fully confirm what is reported in the product information while in other cases, they show ADRs whose frequency or causality is not exactly known. Therefore, additional analyses are necessary to better outline the safety profile of ear side effects.

References:

• 1. Yorgason JG, Fayad JN, Kalinec F. Understanding drug ototoxicity: molecular insights for prevention and clinical management. Expert Opin Drug Saf 2006;5:383–99

• 2. Schacht J1, Hawkins JE. Sketches of otohistory: part 11: ototoxicity: drug-induced hearing loss. Audiol Neurootol 2006;11:1–6

• 3. King KA, Brewer CC. Clinical trials, ototoxicity grading scales and the audiologist’s role in the therapeutic decision making. Int J Audiol 2017;23:1–10

• 4. Lanvers-Kaminsky C, Zehnhoff-Dinnesen AA, Parfitt R, Ciarimboli G. Drug-induced ototoxicity: mechanisms, pharmacogenetics, and protective strategies. Clin Pharmacol Ther 2017;101:491–500

• 5. Ganesan P, Schmiedge J, Manchaiah V, Swapna S, Dhandayutham S, Kothandaraman PP. Ototoxicity: a challenge in diagnosis and treatment. J Audiol Otol 2018;22(2):59–68

Disclosure of Interest: None declared.

117 ISoP18-1165 Social Media: the Battleground for Public Opinion on Medicines Safety

117.1 A. Hoegberg^*1

117.1.1 ¹Global Communications, Uppsala Monitoring Centre, Uppsala, Sweden

Background/Introduction: With 2.2 billion users on Facebook, the world’s largest social media platform, and considering the potential for rapid “viral” spread of information, social media has a huge potential to influence public opinion on various topics—particularly issues where opinions are easily polarised, such as recurring debates on the safety of vaccines and medicines. Social media is therefore increasingly important to health communication and to pharmacovigilance advocates.

Objective/Aim: The objective of this work is to outline the challenges faced by pharmacovigilance advocates and healthcare actors when using social media to address various aspects of medicines safety, such as improving the public’s understanding for the benefits and risks of medicines, fostering patient adherence and trust in health authorities, and enhancing adverse drug reaction reporting rates. This also includes the particular challenges posed by strong voices spreading messages that are counterproductive to the work of pharmacovigilance actors.

Methods: This work builds on a review of academic papers; on lectures previously given by the author on the topic of social media management for pharmacovigilance actors; and on observations of user behaviour on social media and of the efficacy of various types of social media messages.

Results: Social media is widely used to search for health information, and the information such searches generate may have a profound impact on the users’ perception of the health issue in question and inform their course of action [1]. Due to the self-publishing nature of social media, every user, whether private or organisational, have the opportunity to share and widely disseminate their messages. In effect, social media platforms can and are used as mouthpieces for the full spectra of voices present online, from healthcare authorities to laymen, from public health actors to anti-vaccine advocates, and every imaginable creed in between.

The traditional top–down model of information dissemination, which is employed by many health authorities, is ineffective on social media where audiences need to be actively engaged and listened to [2, 3].

Conclusion: Many health authorities are poorly equipped both in terms of crafting effective health messages on social media, and in terms of addressing rumors and inaccurate information circulating online. However, with the right insights into audience behavior and with strategies in place to foster trust in one’s own information, social media can be used to proactively build a knowledge-base of medicine safety, and as a vehicle for trustworthy and timely safety information in times of crises [4]. This has the potential to improve public health and to facilitate discussions about the benefits and risks of medicines—whether between individual healthcare providers and patients, or between authorities and the public—and to counteract the flow of inaccurate information that may be harmful to the bottom-line of medicines safety advocates.

References:

• 1. Orr D, et al. Social media as a platform for health-related public debates and discussions: the Polio vaccine on Facebook, Isr Health Policy Res 2016;5:34

• 2. Heldman AB, et al. Social media engagement and public health communication: implications for public health organizations being truly “social”. Public Health Rev 2013;35:13

• 3. Moorhead SA, et al. A New Dimension of Health Care: Systematic Review of the Uses, Benefits, and Limitations of social media for health communication. J Med Internet Res 2013;15:e85

• 4. Novillo-Ortiz D, Hernández-Pérez T, Social media in public health: an analysis of national health authorities and leading causes of death in Spanish-speaking Latin American and Caribbean countries BMC Med Inform Decis Mak 2017;17:16

Disclosure of Interest: None declared.

118 ISoP18-1166 Safety and Tolerability of Antipsychotic Drugs in Pediatric Patients: Data from an Ongoing Active Pharmacovigilance Study in Sicily

118.1 G. Cicala^*1, M.A. Barbieri¹, V. Santoro¹, A. Gagliano², E. Germanò², C. Tata³, V. Colucci⁴, F. Vanadia⁵, F. Drago⁵, C. Russo⁶, P. M. Cutroneo⁷, E. Spina^1,7

118.1.1 ¹Department of Clinical and Experimental Medicine, ²Department of Adult and Developmental Human Pathology “Gaetano Barresi”, University of Messina, Messina, Italy; ³Childhood and Adolescence Neuropsychiatry, Provincial Health Unit 8, Syracuse,Italy; ⁴Complex Operive Unit of Neurology for Mental Disabilities, I.R.C.C.S. Oasi Maria SS., Enna, ⁵Developmental Age Neuropsychiaetry, A.R.N.A.S Civico di Cristina Benfratelli, Palermo, Italy; ⁶Developmental Age Neuropsychiaetry, S. Marta and S. Venera Hospital, Catania, Italy; ⁷Regional Pharmacovigilance Centre, University Hospital of Messina, Messina, Italy

Background/Introduction: Antipsychotic drugs are increasingly used to treat a variety of psychiatric disorders in children and adolescents. However, these medications are frequently used outside of their therapeutic indications. For this reason, a higher level of attention regarding the safety and tolerability profile of these drugs is necessary.

Objective/Aim: To perform an integrated assessment of the risk and tolerability profile associated with the use of antipsychotic drugs (both typical and atypical) in pediatric patients, with a focus on the identification of suspected adverse drug reactions (ADRs).

Methods: Starting from April 2017 pediatric patients followed by 5 units of developmental age neuropsychiatry who initiated a treatment with at least an antipsychotic drug (ATC class N05A) where included into the study, regardless of diagnosis and concomitant administration of other drugs. Data relative to patient’s characteristics, therapy, use indications, clinical and laboratory examinations have been collected at the time of treatment initiation and regularly thereafter into a dedicated database.

Results: To date, 129 patients (84 males and 45 females, mean age ± SD of 13.6 ± 3.4 years) were enrolled into the study, 99 (76.7%) of them were assisted asoutpatients, 25 (19.4%) as inpatients and 5 (3.8%) in day care regimen. Fifty-six patients were initially treated with risperidone, 46 with aripiprazole, 10 with olanzapine, 5 with clozapine, 3 with levomepromazine, 3 with quetiapine, 3 with haloperidol, 1 with asenapine, 1 with clotiapine and 1 with periciazine. The most represented diagnoses, classified according to the ICD10, were: attention-deficit hyperactivity disorder (n = 31), autistic disorder (n = 30), mild intellectual disabilities (n = 16), Tourette’s disorder (n = 14), oppositional defiant disorder (n = 13) and obsessive–compulsive disorder (n = 12). In total, we observed 207 ADRs with the most frequent being weight increase (n = 31), hyperfagia (n = 21), hyperprolactinemia (n = 17) and hypercholesterolaemia (n = 14). Among the 48 cases of ADR reported into the Italian pharmacovigilance network only 7 (14.6%) were serious and were represented by elevated creatinine phosphokinase (n = 2), QTc interval prolongation (n = 2), dystonia (n = 1), leucopenia (n = 1) and hyperprolactinemia (n = 1).

Conclusion: These preliminary datatrace a safety profile for antipsychotics in pediatric patients similar to the one already laid down for adults. However, more data are necessary in order to evaluate risk and tolerability of these drugs when used in pediatric patients.

Disclosure of Interest: None declared.

119 ISoP18-1167 Hepatic Adverse Drug Reactions Associated with Methylprednisolone: Analysis of the French Pharmacovigilance Database

119.1 J. Cottin¹, S. Pierre^*1, V. Pizzoglio¹, C. Simon², G. Durieu³, A. Gouraud¹, B. Kassaï-koupaï¹, J. Dumortier⁴

119.1.1 ¹Centre Régional de Pharmacovigilance, Service Hospitalo-universitaire de Pharmacotoxicologie, Hospices Civils de Lyon, CHU-Lyon, Lyon, France; ²Service de Pharmacosurveillance, Centre Régional de Pharmacovigilance, CHRU Tours, Tours, France; ³Service de Pharmacologie Médicale et Clinique, Centre Midi-Pyrénées de Pharmacovigilance, Pharmacoépidémiologie et Informations sur le Médicament, Pharmacopôle Centre Hospitalier Universitaire et Faculté de Médecine de Toulouse, Toulouse, France; ⁴Service d’Hépato-Gastro-Entérologie, Hôpital Edouard Herriot, Hospices Civils de Lyon, Lyon, France

Background/Introduction: Hepatotoxicity associated with methylprednisolone (MP) is rarely described with only few case-reports in the literature [1].

Objective/Aim: Our aim was to review the characteristics of acute hepatic disorders associated with intravenous (IV) or oral MP administration and registered in the French pharmacovigilance database (FPD) since 1985.

Methods: All cases with MP coded as suspected, concomitant or interacting drug associated with hepatic disorders and reported up to May 2016 were extracted from the FPD. “Drug related hepatic disorders” cases were defined using the Standard Medical Query (SMQ).

Results: After exclusion of 317 poorly informative cases and cases more probably related to another drug or non-drug cause, 97 cases of hepatic disorders associated with MP were analyzed. 58.8% (57/97) of patients were women with a median age of 46 years (1–91 years). MP was used for an autoimmune disease in 47.6% (46/97) of cases which including 27% cases of multiple sclerosis. MP was given intravenously in 79.4%. 72.2% of patients had a hepatocellular type of injury with a severity mainly evaluated as mild (45%) or moderate (31%). For 40 patients (52%) liver injury occurred after MP discontinuation. In these cases the median time to onset of the liver injury after the last MP administration was 12.5 days (range 1–191 days). The median cumulative MP dose was 2 g (range 0.06–42.7 g). Most patients (92%) spontaneously and fully recovered within an average of 38.4 days. A rechallenge using the IV route was performed in 13 patients and 10 (76.9%) of them experienced single or multiple positive recurrence of liver injury, which is in favor of MP responsibility. In all of these cases, the initial type of injury was hepatocellular. Regarding IV route of administration (n = 77), MP was coded as the only suspected drug in 22% of cases.

Conclusion: Our results suggest that intravenous MP responsibility should be considered in case of acute liver injury. By contrast, data on oral MP (n = 20) are generally insufficient to conclude and the evidence of a link is not compelling. The EU summary of product characteristics will be updated soon and specialists should consider this potential adverse reaction even if it occurs several weeks after MP discontinuation.

References:

• 1. Davidov Y, Har-Noy O, Pappo O, Achiron A, Dolev M, Ben-Ari Z. Methylprednisolone-induced liver injury: case report and literature review. J Dig Dis 2016;17:55-62

Disclosure of Interest: None declared.

120 ISoP18-1169 The Impact of Sex on the Associations Between Ace Inhibitors and Cough and Angioedema: A Systematic Review and Meta-Analysis

120.1 O. Klungel¹, F. Alharbi^*1, A. de Boer¹, A. Kholod¹

120.1.1 ¹Pharmacoepidemiology & Clinical Pharmacology, Universiteit Utrecht, Utrecht, the Netherlands

Background/Introduction: Cough and angioedema are well-known adverse effects of angiotensin-converting enzyme (ACE) inhibitors. Some observational studies in patients using ACE inhibitors have observed that women have a higher incidence of cough and angioedema than men.

Objective/Aim: To evaluate based on randomized controlled trials (RCTs), whether the risks of developing cough and angioedema with ACE inhibitors are modified by sex.

Methods: We searched PubMed and Cochrane databases for all years to August 2016. We included RCTs that contain information about the incidence of cough and angioedema in users of ACE inhibitors and controls (active/placebo) in men and women. We performed meta-analyses using the random effects model. Pooled risk ratios (RRs) for cough and angioedema associated with ACE inhibitors in women and men were estimated and tested for interaction.

Results: We included four RCTs in our analysis (three studies for cough and two studies for angioedema). We found that there was no difference in the RR to develop cough or angioedema for ACE inhibitors versus controls between women and men. For cough in women, the RR was 3.70; 95%CI (2.55–5.35) and for men, 2.61; 95%CI (1.30–5.27) (P value for interaction 0.39). For angioedema, these RRs were 5.56; 95%CI (2.45–12.62) and 6.35; 95%CI (1.81–22.36), respectively (P value for interaction 0.86).

Conclusion: Our meta-analyses show that the risks of developing cough and angioedema associated with ACE inhibitors are not modified by sex. However, these findings should be interpreted cautiously due to limited number of studies involved.

Disclosure of Interest: O. Klungel Grant/Research support from: GSK HTA methodology grant, Other: Educational lecture on unobserved confounding for Roche, F. Alharbi: None declared, A. de Boer: None declared, A. Kholod: None declared.

121 ISoP18-1170 Abciximab-Induced Delayed Thrombocytopenia: Case Analysis in the French Pharmacovigilance Database

121.1 A. Maurier^*1, S. Delepine², A. Lillo-Le Louët³, A. Grandvuillemin⁴, G. Drablier¹, L. Lagarce¹, M. Briet¹, D. Bourneau-Martin¹

121.1.1 ¹Regional Center of Pharmacovigilance of Angers, France; ²Cardiology Department, CHU Angers, Angers, France; ³Regional Center of Pharmacovigilance of Paris, Hopital Européen Georges Pompidou, Paris, France; ⁴Regional Center of Pharmacovigilance of Dijon, CHU Dijon, Dijon, France

Background/Introduction: Abciximab (ABX) is a chimeric monoclonal antibody Fab fragment, blocking the platelet receptor glycoprotein IIb/IIIa, used as antithrombotic therapy during percutaneous coronary intervention. The most common adverse effects are bleeding and thrombocytopenia. If most cases of thrombocytopenia occur within a few hours after ABX, delayed thrombocytopenia (3–6 days after ABX discontinuation) is described in the literature but not in its monography; this may wrongly incriminate others etiologies.

Objective/Aim: Case description of delayed thrombocytopenia with ABX from the French PharmacoVigilance Database (FPVD).

Methods: A query of FPVD was performed with the following criterias: ABX single suspected drug, thrombocytopenia with onset delay over 3 days. Each case was analyzed separately according to the good pharmacovigilance practices.

Results: In the FPVD, among 292 cases of thrombocytopenia with ABX, we extracted 43 cases of delayed thrombocytopenia. Patients studied (32 men and 11 women) were aged 35–87 years (median 58 years). The mean time of onset for thrombocytopenia was 8.2 ± 2.1 days (median 8 days, range 3–15 days) and the mean time of improvement was 5.1 ± 2.7 days (median 5 days, range 1–14 days). Eight cases were not serious and 35 were considered serious (30 hospitalizations, 4 life threatening, 1 medically significant situation). Two patients had hemorrhagic signs (epistaxis and purpura). When the data were known (n = 27), 16 of these thrombocytopenia were grade IV (< 25 G/L), 5 grade III (25–49 G/L), 5 grade II (50–74 G/L) and 1 grade I (75–99 G/L). The median platelet nadir was 19 G/L (range 1–78 G/L) and a platelet transfusion was realized only for 5 patients, (median platelet nadir 5 G/L, range 1–11 G/L), regarding the thrombotic risk context. Among the 43 cases, thrombocytopenia recovered totally or partially in 38 patients. For 8 patients, the diagnosis of delayed thrombocytopenia with ABX was made after detection of anti-complex (ABX-platelet) antibodies.

Conclusion: Several cases of ABX-induced delayed thrombocytopenia are reported in the literature [1–6]. The median onset delay is 7 days (range 3–17 days) and the median time of improvement is 4 days (range 2–7 days) which is concordant with our results. In some cases, treatment was initiated with platelet transfusion, infusion of corticosteroids and/or intravenous immunoglobulin, which does not seem to be the majority of our study’s cases. More than half of the cases manifested hemorrhagic signs, including 1 fatal outcome with cerebral hemorrhage, while in our study 5% of the cases presented bleeding complications (no death). As in the literature, thrombocytopenias reported in the FPVD were mainly severe (grade III/IV). Anti-complex (ABX-platelet) antibodies were identified in FPVD and in literature. The pharmacological mechanism underlying the occurrence of thrombocytopenia with ABX is not well established, but may be immune-mediated. These anti ABX-platelets antibodies may be specific for murine peptide sequences in ABX [5]. In conclusion, physicians administering ABX should be aware of this severe but reversible complication, which may occur at home, and should monitor a platelet control 1 and 2 weeks after hospital discharge, even if it was well tolerated previously [3]. This study underlines the importance of pharmacovigilance investigations with the realization of the medical history.

References:

• 1. Jbara M, Bhogal S, Bajaj K. Abciximab-induced delayed profound thrombocytopaenia. BMJ Case Rep 2017; https://doi.org/10.1136/bcr-2017-219379. (abstract only)

• 2. Giupponi L, Cantoni S, Morici N. Delayed, severe thrombocytemia after abciximab infusion for primary angioplasty in acute coronary syndromes: Moving between systemic bleeding and stent thrombosis. Platelets 2015;26:498–500

• 3. Rasti M, Blostein M. Delayed immune-mediated thrombocytopenia after re-exposure to abciximab therapy. Can J Cardiol 2011;27:869.e13–e14

• 4. McCorry RB, Johnston P. Fatal delayed thrombocytopenia following abciximab therapy. J Invasive Cardiol 2006;18:e173–4.

• 5. Curtis BR, Divgi A, Garritty M and Aster. Delayed thrombocytopenia after treatment with abciximab: a distinct clinical entity associated with the immune response to the drug. J Thromb Haemost 2004;2:985–92

• 6. Jenkins LA, Lau S, Crawford M. Delayed profound thrombocytopenia after c7E3 Fab (abciximab) therapy. Circulation 1998;97:1214-5.

Disclosure of Interest: None declared.

122 ISoP18-1171 Cox Selectivity and Chemical Subgroup of Non-steroidal Anti-inflammatory Drugs and Frequency of Spontaneous Reporting of Hypersensitivity Reactions

122.1 O. Klungel^*1, M. Bakhriansyah¹, R. Meijboom¹, P. Souverein¹, A. de Boer¹

122.1.1 ¹Pharmacoepidemiology & Clinical Pharmacology, Universiteit Utrecht, Utrecht, The Netherlands

Background/Introduction: Use of non-steroidal anti-inflammatory drugs (NSAIDs) has been associated with many adverse events, including hypersensitivity reactions (HSRs), such as angioedema and urticaria. However, no studies have investigated whether cyclooxygenase (COX) enzyme selectivity and/or chemical subgroups are associated with a difference in HSRs.

Objective/Aim: to describe and compare the frequency of HSRs among NSAIDs based on cyclooxygenase selectivity and chemical subgroups.

Methods: A case/non-case study was performed using data from the World Health Organization global database of Individual Case Safety Report (ICSR), VigiBase, containing over 13 million ICSRs submitted by the participating member states enrolled under WHO’s international drug monitoring program by June 2016. This study was nested among ICSRs where NSAIDs were a suspected drug. Cases were ICSRs mentioning HSRs (urticaria, angioedema, anaphylactic shock, anaphylactic reaction, anaphylactoid shock, and anaphylactoid reaction), whereas non-cases were all ICSRs without HSRs. Based on the ratio of inhibitory concentration 80% of COX-1/COX-2, NSAIDs were categorized into coxibs, non-coxib NSAIDs with COX-2 preference, NSAIDs with poor selectivity, and NSAIDs with unknown selectivity. Only ICSRs with complete information on age and sex, and NSAIDs with first market authorization from 1978 onward were included. RORs and 95% confidence intervals (95% CIs) to assess the association between NSAIDs and the reporting of HSRs were calculated using logistic regression analysis.

Results: We identified 16,289 HSR cases and 160,319 non-cases among ICSRs involving NSAIDs. Non-coxib NSAIDs with COX-2 preference, NSAIDs with poor selectivity, and NSAIDs with unknown selectivity were all associated with an increased reporting of HSRs (age- and sex-adjusted ROR 1.70, 95% CI 1.61–1.79, age- and sex-adjusted 2.19, 95% CI 2.11–2.77, and age- and sex-adjusted 1.26, 95% CI: 1.03–1.54, respectively) compared to coxibs.

Conclusion: HSRs were more often reported for NSAIDs with poor selectivity, non-coxib NSAID with COX-2 preference, and NSAIDs with unknown selectivity compared to coxibs.

Disclosure of Interest: O. Klungel Grant/Research support from: GSK HTA methodology research, Other: Educational lecture on unobserved confounding for Roche, M. Bakhriansyah: None declared, R. Meijboom: None declared, P. Souverein: None declared, A. de Boer: None declared.

123 ISoP18-1172 The Impact of Antihypertensive Drugs on Serum Potassium and Sodium Levels in Patients Electively Admitted to a Tertiary Hospital

123.1 O. Klungel^*1, P. Cornu², F. Alharbi¹, M. de Groot³, A. Dupont², J. Weyler⁴

123.1.1 ¹Pharmacoepidemiology & Clinical Pharmacology, Universiteit Utrecht, Utrecht, The Netherlands; ²Research Group Clinical Pharmacology & Clinical Pharmacy, Vrije Universiteit Brussel, Brussels, Belgium; ³Clinical Chemistry and Haematology, University Medical Center Utrecht, Utrecht, The Netherlands; ⁴Epidemiology & Social Medicine, University of Antwerp, Antwerp, Belgium

Background/Introduction: Abnormal serum potassium and sodium levels may lead to serious cardiovascular and neurological conditions.

Objective/Aim: To investigate the association between the use of different antihypertensives and the risk of developing disturbances in potassium and sodium serum levels.

Methods: A cross-sectional study was conducted in antihypertensive users, electively admitted to the University Medical Center Utrecht between January 2013 and September 2016. Data on patient characteristics, antihypertensives, and electrolyte levels where extracted from the Utrecht Patient Oriented Database. The association between the use of different antihypertensives and the electrolyte level was studied using linear and logistic regression.

Results: A total of 6369 elective admissions were included in this study. The most frequent electrolyte disorder was hyponatremia (29.5%), followed by hypokalemia (20.5%). Hyperkalemia (3.4%) and especially hypernatremia (0.1%) were less common. In comparison to the use of monotherapy of beta-blockers, use of monotherapy of calcium antagonists (adj. OR 3.08; 95% CI 2.13, 4.46), thiazide or thiazide-like (adj. OR 2.08; 95% CI 1.14, 3.82) and loop diuretics (adj. OR 1.92; 95% CI 1.13, 3.28) was significantly associated with higher odds of hypokalemia. Most combinations of antihypertensives with thiazide or thiazide-like or loop diuretics were significantly associated with lower potassium serum levels compared to monotherapy of beta-blockers. None of the antihypertensive therapies were significantly associated with hyperkalemia. Monotherapy of potassium sparing diuretics (adj. OR 2.72; 95% CI 1.11, 6.66) and angiotensin receptor blockers (adj. OR 1.63; 95% CI 1.01, 2.63), and some of the combinations with a thiazide or thiazide-like or loop diuretic were significantly associated with higher odds of hyponatremia.

Conclusion: Monitoring of serum potassium and sodium levels should be encouraged in patients with antihypertensive drugs especially antihypertensives inducing hyponatremia or hypokalemia to avoid possible severe consequences of abnormal serum potassium and sodium levels.

Disclosure of Interest: O. Klungel Grant/Research support from: GSK HTA methodology research, Other: Educational lecture on unobserved confounding for Roche, P. Cornu: None declared, F. Alharbi: None Declared, M. de Groot: None declared, A. Dupont: None Declared, J. Weyler: None declared.

124 ISoP18-1173 Vaccine Safety Surveillance in Pregnancy Using Gaia Definitions for Neonatal Conditions: A Feasibility Assessment in Low- and Middle-Income Countries

124.1 A. L. Stuurman^*1, M. Riera ¹, S. Lamprianou ², S. Perez-Vilar ³, S. A. Anderson ³, P. Mangtani ⁴, H. Devlieger ⁵, T. Verstraeten¹, P. Zuber ², C. Guillard Maure ²

124.1.1 ¹P95 Epidemiology and Pharmacovigilance Consulting and Services, Leuven, Belgium; ²Department of Essential Medicines and Health Products, World Health Organization, Geneva, Switzerland; ³Center for Biologics Evaluation and Research, U.S. Food and Drug Administration, Silver Spring, MD, United States; ⁴Department of Infectious Disease Epidemiology, London School of Hygiene and Tropical Medicine, London, United Kingdom; ⁵Department of Development and Regeneration, KU Leuven, Leuven, Belgium

Background/Introduction: Global efforts to adequately monitor the safety of new vaccines targeted for use by pregnant women in low and middle-income countries (LMICs) are needed. The Global Alignment of Immunization Safety Assessment in pregnancy project (GAIA) recently published case definitions based on levels of diagnostic certainty for pregnancy and neonatal outcomes and maternal immunization.

Objective/Aim: As a preliminary step to evaluate the applicability of the GAIA case definitions in LMICs, we aimed to select candidate sites and conduct a feasibility assessment to evaluate their ability to identify selected neonatal outcomes [preterm birth, neonatal death, neonatal invasive bloodstream infection (NI-BSI), stillbirth] and the maternal immunization status, with sufficient information to enable their classification by level of diagnostic certainty. The long-term goal of this work is to assess the feasibility of conducting well-designed collaborative observational studies to monitor the safety of new vaccines targeted for use in pregnant women in LMICs as part of the WHO Global Vaccine Safety Initiative.

Methods: Candidate sites were initially screened using a questionnaire. For each neonatal outcome, screened sites were asked to retrospectively identify and collect information for up to three individuals born in 2016. Sites were selected to participate in the MCCS if at least one subject per outcome could be assessed for diagnostic certainty using the GAIA definitions, demonstrating minimum diagnostic capacity and data access. Feedback on the data collection process was collected.

Results: Fifty sites were screened and 32 participated in the feasibility assessment. Twenty-four sites that identified at least one case per outcome were eventually selected. The majority (80%) of preterm births, neonatal deaths, and NI-BSI subjects, but only 50% of stillbirths, could be assessed for diagnostic certainty. The main reasons for cases not being classified as stillbirths were insufficient information to distinguish between antepartum and intrapartum stillbirth (29%) and that level of diagnostic certainty was not the same across all data elements provided (35%). Vaccination status could be assessed in 56% of mothers, with 49% considered vaccinated and 6% not-vaccinated. Sites indicated that retrospective subject identification restricted ascertainment of maternal immunization status and was resource-intensive due to the need to manually search archives.

Conclusion: GAIA case definitions for four neonatal outcomes and maternal immunization were successfully piloted in 24 sentinel sites across four WHO regions. The modification of the GAIA stillbirth definition could help avoid potential misclassification in LMICs. Vaccine safety monitoring in LMICs will benefit from systematic recording of all immunizations administered during pregnancy.

Disclosure of Interest: None declared.

125 ISoP18-1174 A Comparison of Safety-Related Label Changes for Medicines With and Without Major Objections at Time of Marketing Authorisation

125.1 L. Bloem^*1,2, M. Karomi¹, J. Hoekman³, H. Leufkens¹, O. Klungel^1,4, A. Mantel-Teeuwisse¹

125.1.1 ¹Division of Pharmacoepidemiology and Clinical Pharmacology, Utrecht Institute for Pharmaceutical Sciences, Utrecht University, The Netherlands; ²Dutch Medicines Evaluation Board, Utrecht, The Netherlands; ³Innovation Studies, Copernicus Institute of Sustainable Development, Utrecht University, The Netherlands; ⁴Julius Center for Health Sciences and Primary Care, University Medical Center Utrecht, Utrecht, The Netherlands

Background/Introduction: Previous research identified several product- and procedure-related factors associated with safety-related label changes, but none of these studies focused on concerns expressed during the marketing authorisation (MA) process of medicines. During this process, the European Medicines Agency (EMA) may raise major objections regarding clinical study data. We hypothesise that these are associated with more safety-related label changes post-approval.

Objective/Aim: To associate safety-related label changes with presence of major objections regarding phase III data at MA.

Methods: We performed a retrospective cohort study of innovative medicines that were approved by the EMA in 2009 and 2010, excluding influenza vaccines. For these medicines, we identified changes to all clinical sections of the label during the drug life cycle, i.e., between MA and 31 Oct 2017. We assessed whether changes were efficacy- or safety-related and whether they had a positive or negative impact on the benefit–risk. Major objections regarding phase III data at MA were identified in a previous study [1]. Label changes were extracted from the history of procedural steps available in European Public Assessment Reports. For the sections concerning warnings and precautions (section 4.4) and undesirable effects (section 4.8), we calculated the median number of label changes and interquartile ranges (IQRs) and compared their distribution between medicines with and without major objections at time of MA using the Wilcoxon Rank Sum test.

Results: We identified 40 medicines. Of these, the EMA raised major objections regarding phase III data for 23 medicines and they did not for 17. For the 40 medicines overall, we identified 503 changes to the clinical sections of the label, of which 66 (13%) were efficacy-related and 194 (39%) were safety-related. The remaining 243 changes mostly concerned risk management and study results that did not impact the benefit–risk. Of the 194 safety-related changes, 55 (28%) concerned the warnings and precautions section (n = 52 with negative impact) and 114 (59%) the undesirable effects section (n = 112 with negative impact). Their distribution did not differ between medicines with and without major objections at time of MA (Table 1). The results did not change when we omitted safety-related changes with a positive impact on the benefit–risk from the analysis. Furthermore, of 42 label changes concerning risk management, 28 (67%) concerned the warnings and precautions section, and none the undesirable effects section.

Conclusion: During a 7–9 year follow-up period, more safety-related than efficacy-related label changes occurred for innovative medicines approved by the EMA. Safety-related changes to the warnings and precautions section and the undesirable effects section do not seem to be associated with the presence of major objections regarding phase III data at MA. This may indicate that potential risks associated with these objections were adequately minimised peri or post-approval.

References:

1. Putzeist M, Mantel-Teeuwisse AK, Aronsson B, Rowland M, Gispen-de Wied CC, Vamvakas S, et al. Factors influencing non-approval of new drugs in Europe. Nat Rev Drug Discov 2012;11:903–4

Disclosure of Interest: None declared.

126 ISoP18-1175 Weight Change After Anti-psychotic Drug Treatment: Long-Term Evidence from a Retrospective Study Using Electronic Health Records

126.1 J.C. Bazo-Alvarez^*1,2, I. Petersen¹, T. Morris³, J. Carpenter^3,4

126.1.1 ¹Department of Primary Care and Population Health, University College London, London, United Kingdom; ²Instituto de Investigacion, Universidad Catolica Los Angeles de Chimbote, Chimbote, Peru; ³MRC Clinical Trials Unit, University College London, United Kingdom; ⁴Department of Medical Statistics, London School of Hygiene and Tropical Medicine, London, United Kingdom

Background/Introduction: Immediate weight gain, approximately over the first 6 weeks, is a known problem among patients treated with antipsychotic drugs. However, further evaluation over long term periods is needed, especially from real-life contexts. Electronic health records, such those routinely obtained in general medical practice, are useful sources for obtaining this kind of information.

Objective/Aim: Our main aim was to compare weight change 4 years before and 4 years after initiation of antipsychotic drug treatment, analysing men and women separately.

Methods: The study sample comes from The Health Improvement Network (THIN), a large UK primary care database from which we observed patients with a diagnosed psychotic disorder, between 2005 and 2015. Three retrospective cohorts where constructed, one per drug treatment: Olanzapine, Quetiapine and Risperidone. We modelled weight change over time using linear splines with random effect models, from which three slopes of weight change were estimated for: (i) − 4 years to baseline (pre-treatment), (ii) baseline to + 6 weeks (short-term), (iii) + 6 weeks to + 4 years (long-term).

Results: Results showed that pre-treatment weight change was null for Quetiapine (males and females) and Risperidone (males) cohorts, and slightly negative for the rest of groups. Conversely, all drugs were significantly associated with weight gain after treatment initiation, especially Olanzapine. Using Olanzapine, the short-term weight gain for males was 0.54 kg/week whereas for females was 0.37 kg/week. Later, the long-term weight gain for males was 0.007 kg/week whereas for females was 0.013 kg/week. Women consuming Quetiapine were more affected than men, in both the short and long term. In contrast, men consuming Risperidone were slightly more affected than women.

Conclusion: Although the average weight gain per week for long-term is minor than for short-term, it is continual, so patients do not lose the weight they have gained during the first 6 weeks of treatment. Differences in how women and men weights are affected by these antipsychotics can inform decisions to prescribe. Our next research will be focused on the effect of different doses and exposure periods on weight gain, in similar populations.

Disclosure of Interest: None declared.

127 ISoP18-1176 Antihypertensives Prescribed for Pregnant Women in Japan: Prevalence and Timing Determined from a Database of Health Insurance Claims

127.1 T. Ishikawa^*1, T. Obara^2,3,4, H. Nishigori^3,5, K. Miyakoda⁴, M. Ishikuro^3,4, H. Metoki^4,6, T. Ohkubo⁷, J. Sugawara^4,5, N. Yaegashi^3,4,5, M. Akazawa⁸, S. Kuriyama^3,4,9, N. Mano^1,2

127.1.1 ¹Laboratory of Clinical Pharmacy, Tohoku University Graduate School of Pharmaceutical Sciences, Sendai, Japan; ²Department of Pharmaceutical Sciences, Tohoku University Hospital, Sendai, Japan; ³Environment and Genome Research Center, Tohoku University Graduate School of Medicine, Sendai, Japan; ⁴Tohoku Medical Megabank Organization, Tohoku University, Sendai, Japan; ⁵Department of Gynecology and Obstetrics, Tohoku University Graduate School of Medicine, Sendai, Japan; ⁶Division of Public Health, Hygiene and Epidemiology, Tohoku Medical and Pharmaceutical University Faculty of Medicine, Sendai, Japan; ⁷Department of Hygiene and Public Health, Teikyo University School of Medicine, Tokyo, Japan; ⁸Department of Public Health and Epidemiology, Meiji Pharmaceutical University, Tokyo, Japan; ⁹International Research Institute for Disaster Science, Tohoku University, Sendai, Japan

Background/Introduction: Hypertensive disorder of pregnancy is one of the most common medical complications that arise during pregnancy. Thus, treatment with antihypertensive agents is common [1–4]. Due to potential fetal adverse outcomes of drug exposure, the potential toxicity of antihypertensives to the fetus is important to determine. However, little is known about prescribing antihypertensive agents to pregnant women in Japan.

Objective/Aim: To evaluate the status of prescriptions for antihypertensives primarily during pregnancy, and also before and after pregnancy using a large claims database in Japan.

Methods: We extracted data from a large database of health insurance claims. Dates for the pregnancy onset and delivery were identified from the database using reported algorithms [5]. We also determined birth month and year from the database. The prevalence and timing of prescribed antihypertensives during pregnancy were descriptively evaluated. Time trends of prescriptions for antihypertensives were evaluated using multivariate logistic regression analyses. The prevalence of prescribed antihypertensives was evaluated for 180 days before pregnancy and 180 days postpartum among pregnant women who were members of the health insurance societies during the entire period.

Results: At least one antihypertensive agent was prescribed for 1144 of 41,694 pregnant women (274 per 10,000 deliveries), including 715 (171) and 605 (145) who were prescribed with oral and injectable antihypertensives. At least one antihypertensive was prescribed for 95 (23), 146 (35) and 1055 (253) women during the first, second and third trimesters, respectively. The most frequently prescribed oral antihypertensive during pregnancy was nifedipine (67), followed by methyldopa (66), hydralazine (36) and furosemide (19). Renin inhibitors, angiotensin-converting enzyme inhibitors and angiotensin II receptor blockers were prescribed for 1, 2 and 19 pregnant women, respectively. Nicardipine was the most frequently prescribed injectable antihypertensive during pregnancy (62), followed by furosemide (48), hydralazine (20) and nitroglycerin (19). Yearly prescription trends remained similar except for a significant decrease in prescriptions for oral and injectable furosemide (both P < 0.001) and an annual increase in organic nitrate prescriptions (P = 0.0011). Based on an evaluation of 33,942 pregnant women, the number of prescriptions for antihypertensives increased particularly during the third trimester, and then decreased from 91 to 180 days postpartum.

Conclusion: Various types of antihypertensives are prescribed for Japanese pregnant women. The effects of exposing pregnant Japanese women to these agents should be evaluated.

References:

• 1. Takagi K, Yamasaki M, Nakamoto O, Saito S, Suzuki H, Seki H et al. A review of best practice guide 2015 for care and treatment of hypertension in pregnancy. Hypertens Res Pregnancy 2015;3:65–103

• 2. Shimamoto K, Ando K, Fujita T, Hasebe N, Higaki J, Horiuchi M et al. The Japanese Society of Hypertension guidelines for the management of hypertension (JSH 2014). Hypertens Res 2014;37:253–390

• 3. Cea Soriano L, Bateman BT, García Rodríguez LA, Hernández-Díaz S. Prescription of antihypertensive medications during pregnancy in the UK. Pharmacoepidemiol Drug Saf 2014;23:1051–8

• 4. Bateman BT, Hernandez-Diaz S, Huybrechts KF, Palmsten K, Mogun H, Ecker JL et al. Patterns of outpatient antihypertensive medication use during pregnancy in a Medicaid population. Hypertension 2012;60:913–20

• 5. Ishikawa T, Obara T, Nishigori H, Miyakoda K, Inoue R, Hoshiai T et al. Development of algorithms to determine the onset of pregnancy and delivery date using healthcare administrative data in a university hospital in Japan. Pharmacoepidemiol Drug Saf 2018 (in press). https://doi.org/10.1002/pds.4444. [Epub ahead of print]

Disclosure of Interest: T. Ishikawa Employee of: Pfizer Japan, T. Obara: None declared, H. Nishigori: None declared, K. Miyakoda: None declared, M. Ishikuro: None declared, H. Metoki: None declared, T. Ohkubo: None declared, J. Sugawara: None declared, N. Yaegashi: None Declared, M. Akazawa: None declared, S. Kuriyama: None Declared, N. Mano: None declared.

128 ISoP18-1177 A Review of Benefit–Risk Assessment Over the Product Lifecycle

128.1 M. Miljkovic^*1, C. Bielen², M. Simic-Koumoutsaris³, A. Urakpo⁴

128.1.1 ¹Medical Department, PrimeVigilance, Belgrade, Serbia; ²Medical Depertment, PrimeVigilance, Zagreb, Croatia; ³PrimeVigilance, Guildford, United Kingdom; ⁴Medical Department, PrimeVigilance, Guildford, United Kingdom

Background/Introduction: The assessment of a medicinal product, which evaluates the product’s benefits and risks, is at the heart of drug approval decisions [1]. Benefit–risk assessment (BRA) is a valuable tool that takes place in multiple stages during a medicine’s lifecycle, and this assessment can be conducted in a variety of ways [1–3].

Objective/Aim: The aim was to summarize current BRA methods used during approval decisions and in post-approval settings, and to see possible future directions.

Methods: Relevant reviews, recommendations and guidelines published in medical literature and through regulatory agencies over the past 5 years [3–6] have been examined.

Results: The benefit–risk profile of a drug is dynamic and differs for different indications and patient groups. BRA implies the review of two dimensions: the dimension of benefits (determined mainly by the therapeutic efficacy) and the dimension of risks (comprises the safety profile of a drug). Regulators, industry and academia have developed various approaches, ranging from descriptive textual (qualitative) to decision-analytic (quantitative) models, to facilitate the BRA of medicines during the product lifecycle (from Phase I trials, to authorization procedure, post-marketing surveillance and health technology assessment for inclusion in public formularies). These approaches can be grouped into frameworks which are stepwise structured approaches; metrics which are measures for benefits and risks (usually endpoint specific), estimation techniques such as simulation techniques and meta-analysis, and utility survey techniques to elicit stakeholders’ preferences (utilities). All these approaches share the following two common goals: to assist this analysis and to improve the communication of decisions, but each is subject to its own specific strengths and limitations. Before using any method, its utility, complexity, the extent to which it is established, and the ease of results interpretation should be considered. Despite widespread and long-time use, BRA is subject to debate, suffers from a number of limitations, and currently is still under development.

Conclusion: The use of formal, systematic structured approaches to BRA for regulatory decision-making and quantitative methods to support BRA during the product lifecycle is a standard practice in medicine that is subject to continuous improvement and modernization, not only in methodology but also in cooperation between organizations.

References:

• 1. McAuslane N, Leong J, Liberti L, Walker S. The benefit–risk assessment of medicines: experience of a consortium of medium-sized regulatory authorities. Ther Innov Regul Sci 2017;51:635–44

• 2. Pignatti, F, Ashby, D, Brass, EP. Structured frameworks to increase the transparency of the assessment of benefits and risks of medicines: current status and possible future directions. Clin Pharmacol Ther 2015;98:522–33

• 3. Hallgreen CE, Mt-Isa S, Lieftucht A, et al. (PROTECT Benefit–Risk group). Literature review of visual representation of the results of benefit–risk assessments of medicinal products. Pharmaco-epidemiology and Drug Safety, 2016;25(3):238–250

• 4. Hughes D, Waddingham E, Mt-Isa S, et al. Recommendations for benefit–risk assessment methodologies and visual representations. Pharmacoepidemiol Drug Saf 2016;25:251–62

• 5. Hallgreen CE, van den Ham HA, Mt-Isa S, et al. Benefit–risk assessment in a post-market setting: a case study integrating real-life experience into benefit–risk methodology. Pharmacoepidemiol Drug Saf. 2014; 23:974-83

• 6. James Leong, Stuart Walker, and Sam Salek. A practical approach to communicating benefit–risk decisions of medicines to stakeholders. Front Pharmacol. 2015;6:1–9

Disclosure of Interest: None declared.

129 ISoP18-1178 Cognitive Patterns in Pharmacovigilance Assessments

129.1 D. Sartori¹, B. Grundmark^*1,2, J. Ellenius¹

129.1.1 ¹Uppsala Monitoring Centre, Uppsala, Sweden; ²Department of Surgery, Uppsala University, Uppsala, Sweden

Background/Introduction: Assessments of pharmacovigilance case report series of drug-event combinations (DECs), have historically been carried out by experienced professionals using clinical reasoning, possibly supported by decision algorithms [1]. Analysis of clinical reasoning has previously been performed in medicine, e.g., to characterize and improve decision making in surgery [2], but to our knowledge has not yet been performed in pharmacovigilance assessments.

Objective/Aim: To identify, characterise, and compare the reasoning by experienced pharmacovigilance assessors during assessments of DECs.

Methods: Seven DECs were assessed by 5 medical assessors. Their reasoning was elicited and clarified by semi-structured interviews [3], that were transcribed, coded, and analysed using inductive thematic analysis. Related similar codes were grouped into subtasks indicative of cognitive patterns. Subtasks were iteratively grouped into tasks to match actions taken by participants in assessments and macrocognitive functions [4]. For every assessment subtasks and tasks were qualitatively appraised for patterns and themes.

Results: The analysis revealed 5 tasks comprised by 20 unique subtasks (Table 1) being used by the assessors. Subtasks and tasks were arranged in a non-hierarchical assessment structure. Participants were found to transition from one task to another more than once before reaching a conclusion (indicated by “decide/choose option” subtask in Table 1). Within each task, subtasks were not carried out in any consistent order. Before consulting the case series, participants generally started with a situation assessment, in which they appraised their knowledge of a given DEC and initially formed expectations of the case report series analysis yield. In the planning/re-planning task, goals were established according to initial expectations, which could lead to gathering information, analysing the case reports, or providing a final decision. No decisions were made only based on expectations. Information gathering consisted in consulting drug labelling or publications, leading to hypotheses. Case reports analysis yielded further hypotheses, competing or compatible with initial expectations. Participants addressed only a subset of their hypotheses before reaching a final conclusion. In most cases, assessors reached similar final conclusions on the DECs. Differences were partly explained by participants’ depth of analysis [3], or choice of sources of information. 

Conclusion: These results of inter- and intra-individual variations in cognitive patterns indicate a flexible approach to assessments, and may hold potential implications in informing the design of training programmes for pharmacovigilance assessment skills and decision support tools.

References:

• 1. Agbabiaka TB, Savovic J, Ernst E. Methods for causality assessment of adverse drug reactions: a systematic review. Drug Saf. 2008;31:21–37

• 2. Jalote-Parmar A, Badke-Schaub P, Ali W, Samset E. Cognitive processes as integrative component for developing expert decision-making systems: A workflow centered framework. J Biomed Inform 2010;43:60–74

• 3. Grundmark B, Sartori D, Ellenius J. Do they all agree?: comparing expert signal assessment. Drug Saf. 2017;40:976–7.

• 4. Crandall B, Klein GA, Hoffman RR. Working minds: A practitioner’s guide to cognitive task analysis: Mit Press; 2006.

Disclosure of Interest: None declared.

130 ISoP18-1181 Penicillin-Induced Acute Generalized Exanthematous Pustulosis: The Importance of Patch Testing

130.1 N. Fathallah¹, A. Saii², S. Larif¹, B. Ouni¹, H. Hmouda³, R. Slim⁴, C. Ben Salem^*1

130.1.1 ¹Pharmacovigilance Center of Sousse, Research Laboratory, Sousse University, Tunisia; ²Amen Medical Center; ³Department of Intensive Care Unit, ⁴Pharmacovigilance Center of Sousse, Research Laboratory, Sahloul University hospital, Sousse, Tunisia

Background/Introduction: Acute generalized exanthematous pustulosis (AGEP) is a rare and potentially severe cutaneous reaction. It is usually caused by a drug reaction and its onset is 2–11 days after medication exposure. Antibiotics are the most common trigger [1].

Objective/Aim: To report a penicillin-induced AGEP case confirmed by positive patch test.

Methods: The first case is a previously healthy woman in his 40 s presented with 2 days of fever and a rash after initiation of amoxicillin for dental pain. On examination, a widespread symmetric erythematouseruption with numerous non follicular pinpoint pustules over the trunk and proximal limbs, was present. The patient was febrile (38 °C). Laboratory analysis revealed an elevated white blood cell count of 15 × 10⁹/L with 88% neutrophils. A punch biopsy of the skin lesions showed subcorneal and intraepidermalspongiform pustules. Based on the clinicopathologic features, amoxicillin-induced AGEP was suspected. The second case is a woman with psoriasis history in his 30 s presented with 2 days of fever and a rash after initiation of oxacillin for otitis.On examination, a widespread erythematous eruption, with numerous nonfollicular pinpoint pustules over the trunk and extremities, was observed. The patient was febrile (39 °C). Blood analysis showed a raised white blood cell count of22.7 × 10⁹/L, with a neutrophil count of 17.6 × 10⁹/L and C reactive protein of 28 mg/L (normal range < 7). A biopsy of the involved skin was consistent with AGEP.

Results: Drugs are incriminated in more than 90% of cases of AGEP. Medications from many different pharmacological classes, especially antibiotics such as sulfonamides, aminopenicillins, quinolones, lincosamides and pristinamycin, have been suspected in the development of AGEP. β-Lactams account for 80% of antibiotics implicated in AGEP [2].The major differential diagnosis were infections, generalized pustular psoriasis, and drug reaction with eosinophilia and systemic symptoms (DRESS). Differentiating AGEP from pustular psoriasis is important, as the management of each condition is quite distinct [2].

Conclusion: Drug patch tests have been recently confirmed to be safe, with few reported relapses or severe reactions, and helpful to assess drug imputability in AGEP, with a high proportion of positive results as compared with patients with other drug-related eruptions [3].

References:

• 1. Duong TA, Valeyrie-Allanore L, Wolkenstein P, Chosidow O.Severe cutaneous adverse reactions to drugs. Lancet 2017;390:1996–2011.

• 2. Yek C, Gupta A, Mauskar M. Fever and a pustular rash. JAMA 2017;317:637–8

• 3. Sidoroff A, Dunant A, Viboud C, Halevy S, Bavinck JN, Naldi L, et al.Risk factors for acute generalized exanthematous pustulosis (AGEP)-results of a multinational case–control study (EuroSCAR). Br J Dermatol 2007;157:989–96

Disclosure of Interest: None declared.

131 ISoP18-1182 Anti-epileptic Recurrent Drug-Induced Rash With Eosinophilia and Systemic Symptoms

131.1 N. Fathallah¹, B. Ouni¹, A. Saii², S. Larif¹, H. Hmouda³, R. Slim¹, C. Ben Salem^*1

131.1.1 ¹Pharmacovigilance Center of Sousse. Research Laboratory, Sousse University, Tunisia; ²Amen Medical Center, Tunis, Tunisia; ³Intensive Care Unit, Sahloul University Hospital, Sousse, Tunisia

Background/Introduction: Drug rash with eosinophilia and systemic symptoms (DRESS) is characterized by fever, rash and internal organ involvement after exposure to certain drugs. Most of the aromatic anticonvulsants, such as phenytoin, phenobarbital, and carbamazepine, can induce DRESS [1].

Nonaromatic drugs such as lamotrigine and valproate are known to be more safe than aromatic anticonvulsants and are less responsable of DRESS.

Objective/Aim: To report a case of DRESS following exposure to valproate and recurring after phenobarbital intake.

Methods: A 54-year-old man without previous significant medical history had been diagnosed with brain tumor. He was treated by surgery and received valproate (400 mg daily). No other medications had been taken. Three weeks later, the patient was admitted to the hospital with fever and general eruption. The patient’s face was edematous and erythematous papules were scattered over his entire body. Lymph nodes were palpable. Laboratory findings showed hypereosinophilia and elevated liver enzymes. Viral serology was negative for hepatitis A, B, and C, cytomegalovirus and Epstein–Barr virus. On suspicion of DRESS, valproate was stopped. While clinical symptoms and laboratory findings improved progressively, phenobarbital was started. The patient developed again a generalized rash and fever few days after phenobarbital administartion. The neurosurgeon decided to stop phenobarbital. Symptoms resolved few days later without complications.

Results: Drug rash with eosinophilia and systemic symptoms is a severe adverse reaction with high mortality rates. The aromatic anticonvulsants are the most frequently incriminated drugs [2]. In DRESS, discontinuation of the offending aromatic anticonvulsant is essential for improving the prognosis. In this case, valproate is usually a safe alternative for aromatic anticonvulsants. In fact, Sodium valproate is very rarely responsible for DRESS. As far as we know, DRESS syndrome cases related solely to the use of valproate have not been previously reported. Herin, we report the first case of DRESS primarily induced by sodium valproate and secondarily to phenobarbital. This case illustrates a possible cross-reactivity between valproate and phenobarbital, which are non aromatic and aromatic anticonvulsants, respectively.

Conclusion: Switching anticonvulsants is not usually safe. Clinicians should be more vigilant when adverse events occur first with non aromatic anticonvulsants.

References:

• 1. De A, Rajagopalan M, Sarda A, Das S, Biswas P. Drug reaction with eosinophilia and systemic symptoms: an update and review of recent literature.Indian J Dermatol 2018;63:30–40

• 2. Ben Salem C, Slim R, Denguezli M, Nouira R, Hmouda H, Bouraoui K.A recurrent drug rash with eosinophilia and systemic symptoms. Pediatr Dermatol 2007;24:666–8

Disclosure of Interest: None declared.

132 ISoP18-1183 The Association Between Inappropriate Medication Use and Health-Related Outcomes Among Nursing Home Residents: A Systematic Review and Meta-Analysis

132.1 P. Mongkhon^*1,2, T. Tanasombat¹, R. Jeenapongsa¹, C. Kongkaew¹

132.1.1 ¹Department of Pharmacy Practice, Faculty of Pharmaceutical Sciences, Naresuan University, Mueng, Phitsanulok, Thailand; ²School of Pharmaceutical Sciences, University of Phayao, Mueng, Phayao, Thailand

Background/Introduction: Potentially inappropriate medications (PIMs) are becoming a significant health problem affecting elderly in nursing home (NH) with a prevalence of 43 > 47% [1, 2]. Several studies reported the prevalence of PIM use among people residing in NH settings [3, 4], but only few studies investigated the association between PIM use and health outcomes. Furthermore, the published data are still inconsistent to demonstrate that the PIM use is significantly associated with negative health outcomes.

Objective/Aim: This systematic review and meta-analysis aimed to assess the relationship between PIMs and their consequences for health-related outcomes in NH residents.

Methods: Observational studies that examined the association between health-related outcomes and PIM use among NH residents were identified from the Cochrane library, CINAHL, SCOPUS, and PUBMED until October, 2017. Quality assessment was performed using the National Institute of Health (NIH) Quality Assessment Tool for Observational Cohort and Cross-sectional Studies. The outcome measures were mortality, hospitalization, emergency department (ED) visits and any adverse events association with PIM use among nursing home residents which were reported as odd ratio (OR), risk ratio (RR), or hazard ratio (HR) together with 95% confident interval (95% CI). Pooled estimates were obtained by using the DerSimonian-Laird random-effects model. All analyses were performed using STATA, v14.1.

Results: Eleven studies were included in the systematic review and 8/11 studies in the meta-analysis. The prevalence rate of PIMs among nursing home residents ranged from 11.9 > 77.2% with a median of 46.5% (Interquartile range; IQR = 32.2 > 50.3%). Our meta-analysis demonstrated that PIM users had a significant association with mortality compared to PIM non-users (OR = 1.46, 95% CI 1.24–1.72, I² 51.8%). In the United States, PIM users were more likely to be hospitalized than non-users (OR = 1.27, 95% CI 1.15–1.41, I² 0%). For health-related quality of life (HRQoL), no significant changes were observed in EuroQol-5 dimension (EQ-5D) score (p = 0.07) and EuroQoL visual analog scale (E-VAS) score (p = 0.34). However, when 15D and the psychological well-being scale were employed, quality of life was better for residents who were not PIM users after adjustment for age and gender The cost of pharmaceutical care services was positively correlated with the number of medications that inappropriately prescribed.

Conclusion: The use of PIM was substantial and associated with a higher risk of mortality among nursing home residents. But their impact on HRQoL remained inconclusive. Further studies with prospective long-term follow up are needed to establish its association.

References:

• 1. Morin L, Laroche ML, Texier G, Johnell K. Prevalence of potentially inappropriate medication use in older adults living in nursing homes: a systematic review. J Am Med Dir Assoc 2016;17:862.e1–9

• 2. Storms H, Marquet K, Aertgeerts B, Claes N. Prevalence of inappropriate medication use in residential long-term care facilities for the elderly: A systematic review. Eur J Gen Pract 2017;23:69–77

• 3. Nyborg G, Brekke M, Straand J, Gjelstad S, Romoren M. Potentially inappropriate medication use in nursing homes: an observational study using the NORGEP-NH criteria. BMC Geriatr 2017;17:220

• 4. da Costa FA, Periquito C, Carneiro MC, Oliveira P, Fernandes AI, Cavaco-Silva P. Potentially inappropriate medications in a sample of Portuguese nursing home residents: Does the choice of screening tools matter? Int J Clin Pharm 2016;38:1103–11

Disclosure of Interest: None declared.

133 ISoP18-1185 Process Characteristics and Time to Follow-Up of Adverse Drug Reaction Reports from a Single Center: A Retrospective Analysis

133.1 M. Ganso^*1, S. P. Lenhart², A. Said¹, M. Schulz¹

133.1.1 ¹Drug Commission of German Pharmacists, Department of Medicine, Federal Union of German Associations of Pharmacists, Berlin, Germany; ²Department of Pharmacy, Kbo-Isar-Amper-Klinikum München-Ost, Munich, Germany

Background/Introduction: Spontaneous adverse drug reactions (ADR) reports may be incomplete or may require supplementary information for scientific evaluation. Follow-up methods should be optimized in ways that encourages the reporter to submit missing information of a particular safety concern [1]. This includes the preparation of specific follow-ups without a delay.

Objective/Aim: Evaluation of follow-up characteristics and duration to initiate a follow-up by marketing-authorization holders (MAHs) in comparison to the national pharmacovigilance center Drug Commissions of German Pharmacists (AMK).

Methods: Generated as part of an ADR project, spontaneous reports provided by a clinical pharmacist working on acute psychiatric wards and using the AMK ADR-form were analyzed retrospectively. As usual, the AMK assessed all reports and followed up, if needed. The reports, including the request for additional information, were forwarded to the responsible MAH, who could follow up the reports likewise. The median (range) duration in days (d) was calculated for each process step and compared by paired sample Wilcoxon test (RStudio v 1.1.453). A significance level alpha = 0.05 was chosen.

Results: Between July 2016 and July 2017, the AMK received 42 ADR reports to 70 medicinal products marketed by 25 different MAHs. For 25 reports, the AMK sent a specific follow-up within 6d (1–15) to the reporter. 37 reports were forwarded to the MAH within 5d (1–14), including follow-ups. After receipt of the reports, 10 different MAH sent 22 follow-ups (19 reports) within 17d (5–28) to the reporter. In 8 out of 22 follow-ups, the MAH only sent blank forms. The reporter replied by email to all follow-up requests within 11d (2–31) to the AMK and predominantly by mail within 12d (0–40) to the various MAHs. In total, 12 reports were identified which were followed by both AMK and eight different MAHs. In this subset, the duration to initiate a follow-up by AMK vs. MAH differed statistically significant: median difference 9d [95% confidence interval (− 13.5 to − 4.5); p = 0.002].

Conclusion: The AMK followed up initial ADR reports from a single center significantly faster than the responsible MAH. Regarding the study size and considering a potential impact on MAH follow-ups by AMK’s first handling of reports, results should be interpreted with caution.

References:

• 1. Guideline on Good Pharmacovigilance Practices (GVP)—Module VI (Rev 2), EMA/873138/2011 Rev 2, 28 July 2017

Disclosure of Interest: None declared.

134 ISoP18-1187 Contribution of Patient’s Reports in Signal Detection: Experience of a Pharmacovigilance Center

134.1 C. Le Beller¹, N. AsgariI¹, A. Lillio-Le Louet^*1

134.1.1 ¹Pharmacovigilance Hopital Européen Georges Pompidou, Paris, France

Background/Introduction: Along with statistical data mining in the French Database (FDB), traditional approaches for signal detection are still under way in France. The French Drug Medicine Agency (ANSM) has asked the pharmacovigilance regional centers (CRPVs) to review individual cases in this perspective. Several criteria must be taken into account and formalized in 2016: seriousness, expectedness, vulnerable populations and cluster. Since June 2011 [1], patients can report adverse drug reactions (ADRs) and contribute to signal detection both in case-by-case review and statistical methods [2].

Objective/Aim: To assess the contribution of patient’s reports in manual signal detection.

Methods: Each ADRs report is analysed by the CRPV’s team as a potential signal, and if requirements are met, the cases are transmitted to the ANSM. We retrospectively analysed all cases of potential signal from patients transmitted to the ANSM from June 2011 to December 2017, and more specifically the case descriptions (patient, sex, age, ADRs, drugs involved) and criteria of transmission.

Results: During the study, 9519 spontaneous reports were registered in the FDB, including 1677 patients’ reports (17.6%). We transmitted 125 cases, 16 (12.9%) being directly reported by patients. Half of these cases involved children (1F, 5M, 2 unknown) aged from 7 months to 4 years. In 7 cases out of 8, a vaccine was suspected: 6 cases with meningitis vaccine, in a media context (legal complaints after withdrawal for quality defect) and one narcolepsy with influenza type A H1N1 vaccine. The last case concerned a lack of information about casein, an excipient with risk allergy. Eight patients were adults (5F, 3M), aged from 19 to 66 years. Two cases involved finasteride use in men: an irreversible sexual dysfunction and a misuse in a patient with a history of breast cancer. Other cases transmitted were: hepatic metastasis 24 years after a clear cell carcinoma of the uterine cervix, after in utero exposition to diethylstilbestrol, a malaise with Decontractyl^® (mephenesine) with withdrawal procedure ongoing in France, risk of allergy linked to excipients, early onset of intense headaches with tenofovir alafenamide, mania crisis related to venlafaxine and myelitis after HPV vaccine.

Conclusion: In our CRPV, 17.6% of ADRs collected since 2011 have been patient reports, and a close proportion (12.9%) was transmitted to the ANSM as potential signals, with a high proportion of vaccine-related pediatric cases, in a context of media coverage. Furthermore, some cases meet other criteria: unexpectedness; seriousness or misuse.

References:

• 1. European Directive 2010/84/EU. https://ec.europa.eu/health//sites/health/files/files/eudralex/vol-1/dir_2010_84/dir_2010_84_fr.pdf

• 2. Watson S, Chandler RE, Taavola H, Härmark, Grundmark B, Zekarias A, et al. Safety concerns reported by patients identified in a collaborative signal detection workshop using VigiBase: results and reflections from Lareb and Uppsala monitoring centre. Drug Saf 2018;41:203–12

Disclosure of Interest: None declared.

135 ISoP18-1188 Meningiomas and Progestins: Querying Brest CHRU Data Warehouse

135.1 L. El Aridi¹, C. Le Guillou², H. Jantzem^*1, Y. Audouard-Marzin¹, A. Happe², D. Carlhant-Kowalski¹, J.-M. Cauvin²

135.1.1 ¹Centre Régional de Pharmacovigilance, ²Département d’Information Médicale, CHRU de Brest, Brest, France

Background/Introduction: The hypothesis of an increased risk of meningioma in patients treated with high doses of cyproterone acetate has been validated by the french authority Agence Nationale de sécurité des médicaments et des produits de santé (ANSM) and the European Medicines Agency (EMA) [1, 2]. The role of other progestin treatment, used outside contraception, in the occurrence and/or growth of meningiomas remains to be clarified. The Pharmacovigilance Center of Brest questioned the hospital database in order to evaluate this signal.

Objective/Aim: Bring clinical cases to strengthen the signal with cyproterone and try to quantify this signal with the other progestins.

Methods: The methodology used is that of the HOPIPRAC project. It involves questioning the clinical data warehouse of the University Hospital of Brest since 2003 until June 4th 2017, and to cross two queries, one concerning the drug (progestins used in a hormonal treatment and/or hormone replacement therapy: cyproterone, nomegestrol, chlormadinone…) and the other concerning the clinical reaction: meningioma. This second query combines a text research for the word meningioma in the patients different documents, as well as searching by code D32 (ICD-10 code corresponding to meningiomas).

Results: The drug query counts 8834 patients exposed to progestins, and that of the clinical reaction counts 2585 patients for whom a diagnosis of meningioma was established. Crossing the two requests allows to obtain a sub-warehouse with 150 patients and 1544 documents. These are the patients whose documents contain the drug name and/or the clinical reaction. Based on chronological arguments, reading of these documents allowed us to retain 64 potential cases of meningioma with the different progestins and not notified to the pharmacovigilance center: 37.5% with cyproterone acetate, 25% with nomegestrol, 9.4% with chlormadinone, 9.4% with progesterone, 7.8% with dydrogesterone, 7.8% with progestin contraceptives, and 3.1% with promegestone.

Conclusion: Querying the hospital data warehouse allows to palliate undernotification and to reinforce the signal. Returning to patients medical records and a discussion with clinicians will add additional information regarding the quantification of meningioma risk in terms of dose threshold, duration of treatment and the role of a cumulative exposure to progestins for the same patient in the occurrence of a meningioma.

References:

• 1. Compte-rendu du Comité technique de pharmacovigilance (CTPV) du 18/03/2014: http://ansm.sante.fr/var/ansm_site/storage/original/application/6259701b30dc0177bdccfb40e2c0c874.pdf

• 2. Phamracovigilance Working party (PhVWP) November 2009: http://www.ema.europa.eu/docs/en_GB/document_library/Report/2009/12/WC500016972.pdf

• 3. HOPIPRAC, PRAC signals validation sentinel system, from a hospital data warhouse

Disclosure of Interest: L. El Aridi Shareholder of: none, Grant/Research support from: none, Consultant for: none, Employee of: none, Paid instructor: none, Speaker bureau of: none, Other: none, C. Le Guillou: None declared, H. Jantzem: None Declared, Y. Audouard-Marzin: None declared, A. Happe: None Declared, D. Carlhant-Kowalski: None declared, J.-M. Cauvin: None declared.

136 ISoP18-1189 Example of Using Pharmacogenetics for Drug Safety: Prevalence of the Cyp2c19*2 Polymorphism and Clopidogrel Resistance in Moroccan Population

136.1 A. Lamzouri¹, F. Z. Laarabi², J. Lyahyai³, N. Adadi^2,3, A. El Rherbi¹, A. Tebaa^*1, R. Soulaymani Bencheikh¹, A. Sefiani^2,3

136.1.1 ¹Pharmacology, Moroccan Anti Poison and Pharmacovigilance Center, Rabat, Morocco; ²Medical Genetics, National Institute of Health, ³Human Genomic Center, Faculty of Medicine and Pharmacy, Mohammed V University, Rabat, Morocco

Background/Introduction: Pharmacogenetics is the study of the genetic variation between individuals that affects their response to drugs, both in terms of therapeutic effect as well as adverse effects. Since the beginning of this century, information on pharmacogenetics appears in the summary of product characteristics of drugs. Pharmacogenetic tests particularly concern the enzymes involved in the metabolism of drugs, among which P450 cytochromes. One of the best-known examples is the association of genetic variants of CYP2C19 with Clopidogrel resistance.

Clopidogrel is an antiplatelet agent used as a basic treatment among patients with Acute Coronary Syndromes (ACS) or undergoing percutaneous coronary intervention (PCI). Although its widely described effectiveness, Clopidogrel exhibits a large inter-individual variability of response. This medicine is administrated as an inactive prodrug that requires several biotransformation steps in order to be active. This process takes place via two sequential hepatic reactions of oxidation, catalyzed by the Cytochrome P450 system especially the CYP2C19 located on chromosome 10.

CYP2C19 is a highly polymorphic locus, with almost 27 variants currently reported. The CYP2C19*1 allele corresponds to the normal allele with a complete functional metabolism, and the most described CYP2C19 abnormal variant is CYP2C19*2 (681G > A), its frequencies vary from ~ 15% among Caucasians and Africans, to ~ 29–35% among Asians.

Patients are classified as extensive metabolizers (EMs) if they carry double copy of the wild type allele *1 (*1/*1), intermediate metabolizers (IMs) when carrying one copy of the normal allele and one of the deficient one (*1/*2), and poor metabolizers (PMs) when carrying double copy of a deficient allele (*2/*2).

As the Clopidogrel effectiveness depends essentially on its activation by CYP2C19, therapeutic recommendations based on CYP2C19 genotyping were published, recommending the use of an alternative antiplatelet agent (prasugrel or ticagrelor, if no contraindication) among PMs and IMs patients, as they have reduced platelet inhibition, increased residual platelet aggregation, increased risk for adverse cardiovascular events.

Objective/Aim: The aim of this study was to determine, for the first time, the CYP2C19*2 allele frequency in Moroccan population and explain interest of testing for the CYP2C19*polymorphism in patients being treated with clopidogrel.

Methods: In this study, we used a reliable TaqMan(^®) real-time polymerase chain reaction to detect the CYP2C19*2 allele in 200 unrelated Moroccan newborns. DNA was extracted from umbilical cord blood with maternal consent.

Results: Genotyping of CYP2C19 gene, by real-time polymerase chain reaction using TaqMan probes, revealed a frequency of 18% for CYP2C19*2 allele among 200 Moroccan studied group.

Conclusion: The CYP2C19*2 allele play a vital role in Clopidogrel responsiveness among patients with ACS or after PCI, and Moroccan population has a high prevalence of this allele when compared with other African and Caucasian populations. Hence, testing for the CYP2C19*2 variant is very important for selecting the appropriate anti-platelet agent to treat a cardiac patient, and would be a step towards personalized medicine in morocco.

References:

• 1. Scott SA, et al. Clinical pharmacogenetics implementation consortium guidelines for CYP2C19 genotype and clopidogrel therapy: 2013 update. Clin Pharmacol 2013;94:317–23

• 2. Saydam F, et al. The CYP2C19*2 and CYP2C19*17 polymorphisms play a vital role in clopidogrel responsiveness after percutaneous coronary intervention: a pharmacogenomics study. Basic Clin Pharmacol Toxicol 2017;121:29–36

• 3. Roden DM. Cardiovascular pharmacogenomics: the future of cardiovascular therapeutics? Can J Cardiol 2013;29:58–66

• 4. Simon T, et al. Genetic polymorphisms and the impact of a higher clopidogrel dose regimen on active metabolite exposure and antiplatelet response in healthy subjects. Clin Pharmacol Ther 2011;90:287–95

Disclosure of Interest: None declared.

137 ISoP18-1191 Pregnancies Associated with Etonogestrel Implants: Comparison of Two Five-Year Reporting Periods

137.1 S. Rowlands¹, M. Harrison-Woolrych^*2, E. Cornforth³

137.1.1 ¹Bournemouth University, Bournemouth, United Kingdom; ²ISoP, ³MHRA, London, United Kingdom

Background/Introduction:

Etonogestrel implants are a highly effective contraceptive method licensed to be used for a period of 3 years. Pregnancies are rare but continue to be reported to the British Yellow Card Scheme. Publications in 2005 were the first to identify the issue of the missing implant [1, 2]. There is little information reported so far on failures of Nexplanon, a modified device introduced towards the end of 2010.

Objective/Aim: (1) To identify pregnancies associated with use of the implants Implanon and Nexplanon in the UK during two five-year reporting periods. (2) To classify the reasons for device failure for each implant. (3) To examine any differences between reasons for pregnancies associated with these products.

Methods: Extraction of data from the UK spontaneous reporting system for adverse drug reactions in relation to the etonogestrel implants. All reports indicating pregnancies occurring after insertion of the device were assessed. The years of interest were 2005–2009 (Implanon) and 2012–2016 (Nexplanon). Reasons for failure of the method were placed into one of eight pre-determined categories.

Results: There were 544 cases (280 Implanon and 264 Nexplanon) out of a total of 752 pregnancy reports (363 Implanon and 389 Nexplanon) with sufficient information to make them suitable for analysis. Of the 544 cases, 233 (43%) were considered to be true method failures, 31 (6%) drug–drug interactions, 75 (14%) were already pregnant at the time of insertion, 35 implants were not inserted in accordance with the manufacturers’ instructions, 2 implants were extruded/expelled and 1 placebo device had been inserted in error. There were no reports of self-removal. There were 120 cases (22%) in which the implant was missing: 68 after Implanon insertion and 52 with Nexplanon. In all categories of case there was no clear difference in frequency of pregnancy when the two time periods were compared.

Conclusion: These results are similar to previous Australian and French postmarketing surveillance studies [1–3]. One intended benefit of modification of the device in 2010 was a reduction in the risk of non-insertion but missing implant cases were reported in the second 5-year period in similar proportion to the first. Our assessment suggests that missing implant cases are cases of implant non-insertion: failure to insert the implant and failure to recognise that this has happened. We note that these insertion failures continue despite modification of the device.

Please note that this Abstract does not represent the views of the UK Medicines and Healthcare products Regulatory Agency.

References:

• 1. Harrison-Woolrych M, Hill R. Unintended pregnancies with the etonogestrel implant (Implanon): a case series from postmarketing experience in Australia. Contraception 2005;71:306–8

• 2. Bensouda-Grimaldi L, Jonville-Bera AP, Beau-Salinas F, et al. Insertion problems, removal problems and contraception failures with Implanon^® [in French]. Gynécologie Obstétrique & Fertilité 2005;33:986–90

• 3. Simon C, Agier MS, Béné J, et al. Safety profile of etonogestrel contraceptive implant (Nexplanon and Implanon) reported in France [in French]. Journal de Gynécologie Obstétrique et Biologie de la Reproduction 2016;45:1074–82

Disclosure of Interest: None declared.

138 ISoP18-1192 A Framework for Identifying and Validating Cognitive Services within Pharmacovigilance

138.1 R. Mockute^*1, S. Desai¹, B. Assuncao¹, K. Danysz¹, N. Tetarenko¹, D. Abatemarco¹, M. Widdowson¹, N. Fornaratto¹, S. Beauchamp¹, S. Cicirello¹, E. Mingle¹

138.1.1 ¹Global Drug Safety and Risk Management, Celgene Corporation, Summit, United States

Background/Introduction: The responsibility of pharmacovigilance is to detect, assess, understand and prevent adverse events and any other drug-related problems [1]. To do it effectively it is imperative to collect, collate, evaluate and act upon adverse events. The volume of Individual Case Safety Reports (ICSR’s) describing adverse events has been increasing year on year. Despite this, it is estimated that more than 90% of ADR’s go unreported [2]. To address the evolving landscape, it will be necessary to embrace assistive technologies at scale.

Objective/Aim: Our study outlines a framework for identifying and validating cognitive services for pharmacovigilance, with an aim to set a standard for ensuring appropriate use.

Methods: We developed a framework to validate 40 different cognitive services ranging from information extraction to complex decision making such as ICSR validity, reporter causality, seriousness etc. This framework addresses the following shortcomings: (1) Matching predictions to ground truth needs subject matter expertise (2) Inconsistencies with ground truth (3) Automated validation of prediction can miss context (4) Auto-labeling can lead to inaccurate F1 scores.

The method consists of the following steps: (1) Assessment of cognitive workload (2) Determination of case management decision points where AI can be applied (3) Definition of the training corpus (4) Identification and standardization of PV knowledge elements (5) Cognitive service development (6) Subject Matter Expert (SME) review and validation of cognitive services using the ANSI z1.4 Acceptable Quality Limits standard.

Results: We identified, developed, and validated 40 cognitive services using a corpus of 20,000 ICSR’s. Through the application of this framework, we were able to: (1) Identify decision points that can be candidates for AI (2) Standardize the process to make PV knowledge explicit (3) Preserve PV knowledge and context by embedding SME’s throughout the development process. (4) Standardize the approval process by using established quality inspection principles.

Conclusion: The necessity of AI within PV has been well established, however, its introduction needs the assurance of quality, consistency, and standardization before it can be adopted in a regulated environment. Within PV this is a foundational framework which the industry can use to identify and validate services. More broadly, this framework could be applied to the creation and tuning of cognitive services to be used in a regulated environment.

References:

• 1. Pharmacovigilance [Internet]. World Health Organization. World Health Organization; 2015 [cited 2018Jun1]. Available from: http://www.who.int/medicines/areas/quality_safety/safety_efficacy/pharmvigi/en/

• 2. Nikfarjam A, Sarker A, O’Connor K, Ginn R, Gonzalez G. Pharmacovigilance from social media: mining adverse drug reaction mentions using sequence labeling with word embedding cluster J Am Med Inform Assoc 2015;22:671–81

Disclosure of Interest: R. Mockute Shareholder of: Celgene Corporation, Employee of: Celgene Corporation, S. Desai Shareholder of: Celgene Corporation, Employee of: Celgene Corporation, B. Assuncao Shareholder of: Celgene Corporation, Employee of: Celgene Corporation, K. Danysz Shareholder of: Celgene Corporation, Employee of: Celgene Corporation, N. Tetarenko Shareholder of: Celgene Corporation, Employee of: Celgene Corporation, D. Abatemarco Shareholder of: Celgene Corporation, Employee of: Celgene Corporation, M. Widdowson Shareholder of: Celgene Corporation, Employee of: Celgene Corporation, N. Fornaratto Shareholder of: Celgene Corporation, Employee of: Celgene Corporation, S. Beauchamp Shareholder of: Celgene Corporation, Employee of: Celgene Corporation, S. Cicirello Shareholder of: Celgene Corporation, Employee of: Celgene Corporation, E. Mingle Shareholder of: Celgene Corporation, Employee of: Celgene Corporation.

139 ISoP18-1193 Safety Profile of Immune Checkpoint Inhibitors: An Analysis of Italian Spontaneous Reporting System Database

139.1 P.M. Cutroneo^*1, V. Isgrò², V. Ientile³, M. Santarpia⁴, N. Pellegrino¹, E. Matarangolo⁵, E. Spina^1,2, G. Trifirò⁶

139.1.1 ¹Sicilian Regional Pharmacovigilance Center, University Hospital of Messina, Italy; ²Dept. of Clinical and Experimental medicine, University of Messina, Messina, Italy; ³Unit of Clinical Pharmacology, University Hospital of Messina, Messina; Italy; ⁴Dept. of Adult and Developmental Human Pathology “Gaetano Barresi”, University of Messina, Messina, Italy; ⁵Italian Medicines Agency, Pharmacovigilance Office, Rome, Italy; ⁶Dept. of Biomedical and Dental Sciences and Morpho-functional Imaging, University of Messina, Messina, Italy

Background/Introduction: Immune Checkpoint Inhibitors (ICIs) have significantly changed the treatment of a wide range of malignancies, including melanoma and non-small-cell lung cancer. Several ICIs, most commonly targeting programmed death-1 receptor (PD-1), its ligand (PD-L1), and cytotoxic T-lymphocyte-associated antigen 4 (CTLA-4) have been approved. As a result of their mechanism of action, ICIs may induce immune-related adverse events against normal tissue in multiple organs. These adverse effects have not been though yet fully characterized. Spontaneous ADR reporting system represents the cornerstone for the detection of safety signals and to explore in general safety profile of all marketed drugs including ICIs in clinical practice.

Objective/Aim: To provide an overview of ICI safety data using the Italian Spontaneous ADR Reporting System (SRS).

Methods: We selected all the ADR reports attributed to ICIs in the Italian SRS from January 2011 to December 2017. Study drugs included ipilimumab (anti-CTLA-4), nivolumab, pembrolizumab (anti-PD-1), atezolizumab (anti-PD-L1), all currently marketed in Italy. Descriptive frequency analyses of ADR reports for ICIs as a whole class as well by individual products have been conducted. Reporting odds ratio (ROR) with 95% confidence intervals (CIs) was used as a measure of disproportionality for ADRs associated with the ICI group as compared with all other suspected drugs (reference group), taking into account Standardized MedDRA Queries (SMQs).

Results: Out of a total of 249,457 ADR reports collected in the Italian SRS till end of December 2017, 1461 (0.5%) were related to ICIs. Nivolumab was indicated as a suspected drug in 1066 (73%) reports, followed by ipilimumab (n = 298; 20.4%), pembrolizumab (n = 63; 4.3%) and atezolizumab (n = 28; 1.9%). ICI-related ADR reports mostly involved male patients (63.3%). The mean age (± SD) of ICI users was 63.8 (± 12) years. ICIs were associated with a higher frequency of serious ADR reporting than for reference group (46.8 vs 34.8%; p < 0.001). The most commonly reported serious ADRs were diarrhoea (n = 78), pneumonia (n = 77), lack of efficacy/disease progression (n = 63). The frequencies of general, gastrointestinal, hepatobiliary and respiratory disorders, infections, neoplasms were significantly higher for ICIs than for reference group (p < 0.001). Analysing SMQs, ICIs resulted disproportionally associated with unexpected adverse reactions, such as ischemic heart disease, cardiac failure and ocular motility disorders.

Conclusion: The analysis of ICI-related ADR reports showed that the most frequently reported events concerned general, gastrointestinal and respiratory disorders, of a possible immune-related origin. Potential safety signals, including cardiac and ocular reactions, were detected, requiring further investigation.

Disclosure of Interest: None declared.

140 ISoP18-1195 Cohort Event Monitoring of Newest Antidiabetic Agents Using Patient Reported Outcomes

140.1 A.-M. Van Gorp¹, S. Vorstenbosch^*1, L. Rolfes¹

140.1.1 ¹Netherlands Pharmacovigilance Centre Lareb, Den Bosch, The Netherlands

Background/Introduction: In recent years, several new antidiabetic agents have been marketed, namely the GLP1-agonists, DPP4-antagonists and SGLT2-antagonists. After new drugs enter the market, little is known about characteristics of adverse drug reactions (ADRs), like the time course, management and outcomes.

Objective/Aim: To study the safety profile, including occurrence and characteristics of ADRs, of the new antidiabetic agents.

Methods: We performed a prospective, observational cohort study including 2 GLP1-agonists, 4 DPP4-antagonists and 2 SGLT2-antagonists. New users of these antidiabetic agents were recruited through local pharmacies and asked to participate in this study between 1 February 2008 and 1 January 2016. Participants were invited to complete six web-based questionnaires over a period of 1 year after start.

Results: A total of 818 patient were included, of which 119 users of GLP1-agonist (male/female 38/62%), 624 DPP4-antagonist (57/43%) and 75-SGLT2 (57/43%). The mean age of 60 years was comparable between the groups. 42% (341) of the patients experienced an ADR during the study period. For GLP1-agonist users the percentage of patients that experienced an ADR was the highest with 62% (74), followed by SGLT2-antagonist users with 60% (45), and DPP4-antagonist users with 36% (222). The top 5 reported ADRs (Table 1) and its incidence are comparable with the description in the drugs’ Summary of Product Characteristics (SPC). However, for the SLGT2-antagonists, the incidence seems to be slightly higher in this study compared to the SPC. 26% (47) of ADRs that were reported for GLP1-agonist recovered without withdrawal of the drug. For ADRs reported for DPP4-antagonists and SGLT2-antagonists, this was respectively 22% (90) and 11% (9). Respectively 6.6% (9), 8.8% (55), and 12% (9) of GLP1-agonist, DPP4-antagonist and SGLT2-antagonist users reported withdrawal of the drug due to the occurrence of an ADR.

Conclusion: This study gives insight in the safety profile and characteristics of ADRs of the new antidiabetic agents. This is valuable information that is mostly lacking from premarketing clinical trials.

Disclosure of Interest: None declared.

141 ISoP18-1196 Development and Evaluation of a Digital Recruitment Method of Patients for Lareb Intensive Monitoring

141.1 A.-M. Van Gorp¹, S. Vorstenbosch^*1

141.1.1 ¹Netherlands Pharmacovigilance Centre Lareb, Den Bosch, The Netherlands

Background/Introduction: Before new drugs are marketed, their safety profile is explored. Nevertheless, after market entry of drugs, little is known about characteristics of adverse drug reactions (ADRs), like the time course, management and outcomes. Netherlands Pharmacovigilance Centre Lareb has developed a web-based intensive monitoring system (LIM) which aims to capture data on these topics. Participants for LIM are recruited through pharmacies by an information leaflet. The inclusion rate for LIM studies is low, in 2015 < 1% of the first-time users of non vitamin K anticoagulantia (NOACs) was included for LIM.

Objective/Aim: To increase the number of inclusions for LIM by facilitating pharmacies to send a digital invitation to potential participants.

Methods: During this project a digital method was developed to alert the pharmacy technician about a first-time user and to invite the patient directly for LIM by e-mail. To test this application we performed a 6 months pilot in 30 pharmacies throughout the Netherlands. The number of inclusions was monitored and compared with the number of inclusions of the same pharmacies in the past, when recruitment was through the information leaflet. Furthermore the number of inclusions was compared with the number of inclusions by 30 other pharmacies during the same 6 months period by using the information leaflet method. During and after the pilot the new digital recruitment method was evaluated with the pharmacists of the pilot pharmacies.

Results: The digital application to alert the pharmacy and send an e-mail invitation to first time NOAC users was developed during the first 3 months of the project. Subsequently, this application was introduced by the 30 pilot pharmacies.

During the 6 months pilot the inclusion number for the LIM NOACs study was 58 participants. This is a quadrupling compared to the average number of inclusions over the past 3 years of the same pharmacies. Compared to the 30 pharmacies during the same period, using the leaflet method, the number increased even sevenfold.

Evaluation with local pharmacists provided insight in the practical bottlenecks of the new application. Some technical issues were solved during the pilot period. Other suggestions were mainly about the implementation of the method in the daily practice of the pharmacies.

Conclusion: Using a digital recruitment method to invite patients in local pharmacies led to an increase of the number of inclusions for the LIM NOACs study. The method will be implemented on large scale in the Netherlands.

Disclosure of Interest: None declared.

142 ISoP18-1197 Training an Augmented Intelligent System for Pharmacovigilance: Practical Considerations and Guidance

142.1 D. Abatemarco^*1, S. Desai¹, B. Assuncao¹, N. Tetarenko¹, K. Danysz¹, R. Mockute¹, M. Widdowson¹, N. Fornarotto¹, S. Beauchamp¹, S. Cicirello¹, E. Mingle¹

142.1.1 ¹Global Drug Safety and Risk Management, Celgene Corporation, Summit, United States

Background/Introduction: Regulations are increasing the scope of activities that fall under the remit of drug safety. Currently, individual case safety report (ICSR) collection and collation is done manually, requiring pharmacovigilance professionals to perform many transactional activities before data is available for assessment and aggregated analyses. For a biopharmaceutical company to meet its responsibilities to patients and regulatory bodies regarding the safe use and distribution of its products, improved business processes must be implemented to drive the industry forward in the best interest of patients globally.

Objective/Aim: In this study, we seek to build an augmented intelligent machine-learning system for use across pharmacovigilance. This machine-learning system will handle all pharmacovigilance activities from ICSR intake through signal detection.

Methods: A consortium of 10 cognitive computing services was identified and trained to augment the initial processing of ICSRs for Celgene products. Models were trained using Celgene data and results reviewed by pharmacovigilance subject-matter experts. To be considered adequately trained and functional, each cognitive service was required to reach a threshold of F₁ score ≥ 75%.

Results: All 10 cognitive services have reached an F₁ score ≥ 75% for spontaneous ICSRs.

Conclusion: Cognitive services can augment pharmacovigilance professionals’ decisions and provide efficiency in managing the overwhelming amount of data otherwise manually curated and monitored for ongoing drug surveillance requirements. Cognitive solutions will be key to an organization’s transformation to the real-time and supported decision making needed to meet regulatory requirements and usher in the future of pharmacovigilance as a patient-centered business function.

Disclosure of Interest: D. Abatemarco Shareholder of: Celgene Corporation, Employee of: Celgene Corporation, S. Desai Shareholder of: Celgene Corporation, Employee of: Celgene Corporation, B. Assuncao Shareholder of: Celgene Corporation, Employee of: Celgene Corporation, N. Tetarenko Shareholder of: Celgene Corporation, Employee of: Celgene Corporation, K. Danysz Shareholder of: Celgene Corporation, Employee of: Celgene Corporation, R. Mockute Shareholder of: Celgene Corporation, Employee of: Celgene Corporation, M. Widdowson Shareholder of: Celgene Corporation, Employee of: Celgene Corporation, N. Fornarotto Shareholder of: Celgene Corporation, Employee of: Celgene Corporation, S. Beauchamp: None Declared, S. Cicirello: None Declared, E. Mingle Shareholder of: Celgene Corporation, Employee of: Celgene Corporation.

143 ISoP18-1198 Cognitive Services for Pharmacovigilance Transformation

143.1 S. Bao^*1, S. Perera¹, C. Ramakrishnan¹, R. Routray¹, B. Assuncao², N. Tetarenko², D. Abatemarco², R. Mockute², K. Danysz³, M. Widdowson³

143.1.1 ¹IBM Watson Health, San Jose, CA, United States; ²Pharmacovigilance Innovation, Celgene Corp, Summit, NJ, United States; ³Global Drug Safety, Celgene Corp, Boudry, Switzerland

Background/Introduction: Today, the pharmacovigilance (PV) process is heavily dependent on people throughout each step. With ever-increasing volumes of safety reports, traditional manual approaches of growing in-house teams and outsourcing are becoming unsustainable. Cognitive computing offers capabilities to help handle enormous amounts of information, both structured and unstructured. It improves the ability to understand large volumes of data that humans can’t handle or analyze on their own.

Objective/Aim: Transform PV process with cognitive computing technologies to fully embrace ever-increasing volumes of safety reports. With timeliness and quality of safety operations as critical as ever, cognitive services are designed and integrated to add speed and consistency to the entire pharmacovigilance process from AE intake, triage (event prioritization), evaluation and reporting, to signal detection and assessment.

Methods: Cognitive Services are designed, developed and evaluated based on the following methodologies.

• By design, cognitive services are built on the deep partnership between PV experts and cognitive technology experts, from development to operation. PV experts not only impart expertise via the training data for cognitive service model development but also provide human oversight during production to confirm all evaluations and actions.

• Cognitive services are built on multiple modular components following a micro-services paradigm. Each service is configurable and trainable separately to understand the context of different types of safety reports such as spontaneous reports, clinical trials, and medical literature.

• Cognitive service accuracy is evaluated and validated by following two layers of quality processes. In the first layer, PV experts ensure the high quality of training data and in the second layer, PV experts double check the worst tolerable error cases of each cognitive service by following the Acceptable Quality Limits (AQL) process.

Results: Over 10,000 safe reports have been manually curated and annotated as training data for cognitive model training, and over 40 cognitive services (e.g., adverse event detection, MedDRA coding and ICSR classification) have been identified, trained and approved using AQL process to extract and classify data relevant to each safety report through a comprehensive analysis of the end-to-end PV process focusing on where cognitive technologies can contribute most significantly.

Conclusion: This development accomplishment demonstrates the feasibility and effectiveness of using cognitive computing technologies to transform the pharmacovigilance process with PV expert in the loop.

References:

• 1. Argentinis E, Roberts L, Fraser H. Scaling safety expertise in life sciences: A turning point in pharmacovigilance. IBM Institute for Business Value 2017. https://www-935.ibm.com/services/us/gbs/thoughtleadership/scalingsafety/

• 2. Micro Services: https://en.wikipedia.org/wiki/Microservices

• 3. Acceptable Quality Limit: https://en.wikipedia.org/wiki/Acceptable_quality_limit

Disclosure of Interest: None declared.

144 ISoP18-1199 Augmented Intelligence and the Future of the Drug Safety Professional

144.1 K. Danysz^*1, S. Cicirello², E. Mingle², B. Assuncao², N. Tetarenko², R. Mockute², D. Abatemarco², M. Widdowson², S. Desai²

144.1.1 ¹Global Drug Safety and Risk Management, Celgene, Boudry, Switzerland ; ²Global Drug Safety and Risk Management, Celgene, Summit, United States

Background/Introduction: Automation and augmented intelligence (AI) in the field of pharmacovigilance (PV) will result in the transformation of the drug safety (DS) professional’s work life. A machine-learning (ML) system that will become the new AI-assisted PV database, will impact working practices and processes in the PV industry [1]. The new envisioned work environment promises to address current challenges, such as time pressures, strained resources, and ever-increasing financial restrictions as PV moves from a volume-based business model to a value-based one, better serving the DS professional, the PV industry and ultimately, patients [2]. There is a school of thought that believes use of artificial intelligence will not result in humans losing their jobs with widespread redundancy but instead will create jobs [3].

Objective/Aim: To understand the sentiment of DS professionals towards AI in PV, to envision potential future roles for the DS professional in the new work environment and to ascertain what skills will be required.

Methods: A group of DS employees at Celgene’s Global Drug Safety and Risk Management Department (GDSRM) was interviewed to gain an understanding of their work experience, educational backgrounds, attitudes toward the anticipated changes, hopes for future roles with the ML system, thoughts on support required to adapt their skills to the new environment and the envisioned new roles.

Results: The DS professionals showed interest in development aligned with their own skills, passions, and interests. There was a general need for acquiring new proficiencies in basic computer science, as to understand concepts of natural language processing, ML, and cognitive computing as well as skills in statistics for better data analysis. In their understanding of how the ML system would function, the DS professionals suggested new approaches to daily work and visions for career pathways. A future DS specialist role would focus more on verification of information as suggested by the ML system and may potentially increase their productivity and/or utilize their time accordingly in other work activities. Some other examples of envisioned roles included working in the field of pharmacoepidemiology and focusing on a higher involvement with other functions in the company or with external stakeholders such as regulatory bodies. The DS professionals could potentially bring value in the form of new insights from PV data and be able to assist with company decisions and strategies for patient care and safety. The transcription of non-machine-readable documentation is also required if administrative work is desired.

Conclusion: The DS professionals interviewed were optimistic and enthusiastic about their job roles changing when the ML system for PV was implemented. The vision of all interviewees was one in which PV resources, time, and skills were better used and applied to ultimately achieving better patient outcomes. Managerial implications include transitioning the staff who will engage with the new solution for PV data uptake and analysis. There are opportunities to educate and develop the PV professional through training modules. With the appropriate steps taken, the future of harnessing and utilizing PV talent is promising in the environment where DS professionals are assisted in their daily work by AI.

References:

• 1. Leaf C. Here’s the surprising reason IBM is partnering with Celgene. Time Magazine website. http://time.com/4552874/ibm-watson-health-celgene-partnership/.

• 2. Botsakos G. The future of pharmacovigilance: five imperatives that will drive improved business outcomes. Cognizant Business Consulting website. https://www.cognizant.com/industries-resources/life_sciences/The-Future-of-Pharmacovigilance-Five-Imperatives-that-Will-Drive-Improved-Business-Outcomes.pdf.

• 3. Marr B. Instead of destroying jobs artificial intelligence (AI) is creating new jobs in 4 out of 5 companies. Forbes website. https://www.forbes.com/sites/bernardmarr/2017/10/12/instead-of-destroying-jobs-artificial-intelligence-ai-is-creating-new-jobs-in-4-out-of-5-companies/#58d27da7120d

Disclosure of Interest: K. Danysz Employee of: Celgene, S. Cicirello Employee of: Celgene, E. Mingle Employee of: Celgene, B. Assuncao Employee of: Celgene, N. Tetarenko Employee of: Celgene, R. Mockute Employee of: Celgene, D. Abatemarco Employee of: Celgene, M. Widdowson Employee of: Celgene, S. Desai Employee of: Celgene.

145 ISoP18-1202 Reporting of Qt Interval Prolongation and Torsade De Pointe for Hiv-Antiretrovirals in Vigibase: Focus on Potential Drug–Drug Interactions

145.1 A. Simon^*1,2, C. Marzolini^2,3,4, F. Vanobberghen^1,2, C. Burri^1,2, A. Kuemmerle^1,2

145.1.1 ¹Swiss Tropical and Public Health Institute, Basel, Switzerland; ²University of Basel, Basel, Switzerland; ³Department of Infectious Diseases & Hospital Epidemiology and Department of Clinical Research, University Hospital Basel, Basel, Switzerland; ⁴Department of Molecular and Clinical Pharmacology, University of Liverpool, Liverpool, United Kingdom

Background/Introduction: QT interval prolongation and Torsade de Pointes (QT/TDP) are rare but serious cardiac events that can lead to sudden cardiac death. The prevalence of QT/TDPs in the general population is difficult to determine but people living with HIV have 7–8 times higher risk for QT prolongations compared to the general population, possibly related to administered drugs [1, 2]. Several HIV-antiretrovirals (ARV) and various non-HIV drugs such as methadone, domperidone, haloperidol, and azithromycin have the potential to prolong the QT interval particularly in the context of drug–drug interactions (DDI).

Objective/Aim: Assess the frequency of potential DDIs in individual case safety reports (ICSR) reporting QT/TDPs together with ARVs, and registered in VigiBase, the WHO global database.

Methods: The WHO Collaborating Centre for International Drug Monitoring, Uppsala Monitoring Centre provided an extract of VigiBase including all ICSRs with ARVs considered as suspected or interacting reported by member countries since 1987 until January 2018. The following adverse events (coded as MedDRA Preferred Term) were defined as QT/TDP events: electrocardiogram QT prolonged, long QT syndrome and torsade de pointes. DDIs with ARVs were screened using the Liverpool HIV Drug Interactions database and categorized as deleterious (red), of potential clinical relevance (amber), of weak clinical relevance (yellow) or not interacting (green) [3]. Analyses were descriptive and performed using Stata^® 15 (2017).

Results: Of 180,626 ICSRs, 205 (0.1%) ICSRs mentioned QT/TDPs. Of these, 156 reported at least one co-medication. The five most frequently reported co-medications were methadone, sulfamethoxazole–trimethoprim, fluconazole, lorazepam, and azithromycin. Most frequently observed drug-pairs with red DDI potential involved saquinavir together with methadone, cocaine, or haloperidol. For the amber drug-pairs, atazanavir–methadone, ritonavir–methadone and saquinavir–clindamycin were the most frequently involved. Drug-pairs of the yellow and green categories mainly included co-medication combined mainly with non-nucleoside reverse transcriptase inhibitors (NRTI) with low propensity for metabolic DDI or QT/TDP.

Conclusion: Overall, QT/TDP reporting was low in VigiBase. The number of drug-pairs associated with a higher risk for QT/TDPs was substantial within ICSRs reporting QT/TDP. ARV included boosted atazanavir or saquinavir and co-medications such as methadone, haloperidol, fluconazole, and clindamycin (known risk for QT/TDP), and sulfamethoxazole–trimethoprim (risk for patients with congenital long QT syndrome) [4].

References

• 1. Reinsch N, Arendt M, Geisel MH, Schulze C, Holzendorf V, Warnke A, et al. Prolongation of the QTc interval in HIV-infected individuals compared to the general population. Infection 2017;45:659–67

• 2. Reinsch N, Buhr C, Krings P, Kaelsch H, Neuhaus K, Wieneke H, et al. Prevalence and risk factors of prolonged QTc interval in HIV-infected patients: results of the HIV-HEART study. HIV Clin Trials 2009;10:261–8

• 3. Liverpool Uo. HIV Drug Interactions Tables 2018 [cited 2018 01 March]. Available from: https://www.hiv-druginteractions.org/.

• 4. CredibleMeds. Search for drugs that prolong QT & induce Torsades de Pointes (TdP) [Cited on April 16th, 2018]. Available from: https://crediblemeds.org/index.php/drugsearch.

Disclosure of Interest: None declared.

146 ISoP18-1204 Adverse Effects in Elderly Subjects: Experience of the National Pharmacovigilance Center of Morocco

146.1 S. Yanisse^1,2, A. Tebaa^*1, I. Daoudi^1,2, R. Soulaymani ¹

146.1.1 ¹Antipoison and Pharmacovigilance Center of Morocco, Ministry of Health, Rabat, Morocco; ²Faculty of Medicine and Pharmacy, Rabat, Morocco

Background/Introduction: Adverse drug reactions (ADRs) in the elderly are relatively frequent. This vulnerability to ADR is explained by the physiological and/or pathological aging that leads to pharmacokinetic and pharmacodynamic changes. All these modifications can be responsible for an accumulation of active principles associated with an increased sensitivity to drug treatment, to these elements is added the polypathology responsible for a polymedication which is sometimes necessary for the patient but consequently increases drug interaction risks and therefore the risk of ADR (1, 2).

Objective/Aim: To Describe adverse drug reactions (ADRs) and imputed drugs to subjects aged over 65 in Morocco during the period 2014–2018.

Methods: A retrospective study was conducted on the BDD Database of the CNPV (National Center for Pharmacovigilance). The ADRs were coded according to the Medical Dictionary for Regulatory Activities (MedDRA version 20.1) and grouped into the system organ class (SOC). Each notification was reviewed and analyzed by qualified CAPM medical staff before being recorded in the database. The suspect drugs were coded as International Nonproprietary Names (INN) and classified according to the Therapeutic and Chemical Anatomical Classification (ATC). The analysis is done on vigilyse.

Results: 1830 subjects over the age of 65 were included in the study, of which 973 (53%) are female and 800 (44%) male and 57 (3%) are unknown sex. 24087 ADRs have been detected. The most frequently reported ATC classes were Anti-infectious (24.5%), antineoplastic (22%), and systemic Cardiovascular (21.5%).

In the ADRs, the predominantly imputed drugs were anti-tuberculosis drugs (Ethambutol, Isoniazid, Pyrazinamide, Rifampicin) (18%), beta-lactam antibacterial [Amoxicillin (1.9%), Amoxicillin-clavulanic acid (1.6%)], and others, antineoplastic agents [Rituximab (2.5%), Docetaxel (2%), Cisplatin (1.8%)…], Cardiovascular System [Amlodipine (2.2%), Furosemide (1.8%)…], Acetylsalicylic Acid (4.5%).

The most commonly reported ADRs were skin and subcutaneous tissue disorders (pruritus, rash, urticaria…) (19.9%), gastrointestinal disorders (vomiting, diarrhea, abdominal pain…) (17.9%), general disorders and site administration condition (asthenia) (15.2%)…

Conclusion: The occurrence of ADRs in elderly people patients is a real public health problem. The analysis of the different dimensions such as the quality of the prescription, the transcription, the dispensing, the administration as well as the therapeutic observance of these patients will make it possible to identify and implement corrective measures in order to reduce the avoidable part of occurrence of adverse drug reactions.

References:

• 1. Ankri J. Le risque iatrogène médicamenteux chez le sujet âgé. Gérontologie et société 2002;25:93–106.

• 2. Kanagaratnam L, Taam MA, Heng M, De Boissieu P, Roux M-P, Trenque T. Les effets indésirables médicamenteux graves et leur évitabilité chez des sujets âgés de plus de 65 ans. Thérapie 2015;70:477–84

Disclosure of Interest: None declared.

147 ISoP18-1207 Antibiotics Induced Adverse Drug Reactions in Hospitalized Pediatric Patients: Interim Results from a Prospective Observational Study

147.1 N. B. Bulik¹, A. Farcas², C. Bucsa^*2, I. Cazacu³, A. Slavcovici⁴, O. Oniga¹

147.1.1 ¹Department of Pharmaceutical Chemistry; ²Drug Information Research Center; ³Department of Pharmacology, Physiology and Pathophysiology, Faculty of Pharmacy; ⁴Department of Medical Specialties, Infectious Diseases, University of Medicine and Pharmacy “Iuliu Hatieganu”, Cluj-Napoca, Romania

Background/Introduction: Antibiotics are therapeutic agents widely prescribed in pediatric patients. Therefore, their safety profile is an important issue [1].

Objective/Aim: To assess the ADRs of antibiotics in hospitalized children and to identify the risk factors that lead to ADRs.

Methods: We conducted a prospective observational study in the pediatric department of a Hospital for Infectious Diseases. We enrolled all children with antibiotic therapy prescribed at hospital admission or during hospitalization. Participants were aged up to 18 years and all had bacterial/viral infections. For each child, we collected demographic characteristics, medical treatment, and laboratory findings from their medical chart. The occurrence of new symptoms was assessed through a face-to-face interview with the parent.

Results: Here we present the interim results on the identified adverse events (AEs) for the first 125 participants. The median age of children was 3 years [0–17] and 52% of them were male. The median length of hospitalization was 5 days [2–22]. Penicillin G (31%) was the most frequently prescribed antibiotic, followed by ceftriaxone (27%) and cefuroxime (25%). Dual antibiotic therapy was prescribed to 20 children. A total of 47 (36.7%) children experienced at least one AE. The most frequent AEs were neutropenia (n = 10), vomiting (n = 9) and diarrhea (n = 7). Two cases of rash were reported. The causality, severity and preventability of AEs will be further evaluated and until then these results should be interpreted with caution.

Conclusion: The interim results of our study highlighted the importance of monitoring the safety of antibiotics in the pediatric patients.

References:

• 1. Buccellato E, Melis M, Biagi C, Donati M, Motola D, Vaccheri A. Use of antibiotics in pediatrics: 8-years survey in Italian Hospitals. PloS One 2015;10:e0139097

Disclosure of Interest: None declared.

148 ISoP18-1208 Incidence, Prevalence and Risk Factors of Hypoglycaemia in Diabetes Patients: Systematic Review and Meta-analysis of Observational Studies

148.1 H. Alwafi^*1, A. Alsharif¹, L. Wei¹, A. Naser¹, S. Bell², J. Ilomaki², G. Fang³, M. Al Metwazi⁴, I. Wong^1,5

148.1.1 ¹University college London, London, United Kingdom; ²Monash university, Melbourne, Australia; ³University of North Carolina, Chapel Hill, United States; ⁴King Saud University, Riyadh, Saudi Arabia; ⁵The University of Hong Kong, Hong Kong, Hong Kong

Background/Introduction: Diabetes forms a global health concern and is associated with serious complications and multiple comorbidities [1]. Hypoglycaemia is a major challenge in the treatment of diabetes [2].

Objective/Aim: To review the existing literature for articles that investigated the incidence, prevalence and risk factors of hypoglycaemia in type 1 and type 2 diabetes patients.

Methods: PubMed, Embase and Cochrane library databases were searched up to July 2017. Search strategy was conducted using keywords and MeSH terms. Two reviewers independently screened articles, extracted data and assessed the quality of included studies. Observational studies including patients from all age groups, any study design related to the incidence, prevalence and risk factors of hypoglycaemia in diabetes were included in the review and meta- analysis was conducted where appropriate. Pooled estimate for the incidence and prevalence of hypoglycaemia in type 1 and type 2 diabetes patients were calculated using random effect model with 95% CI.

Results: Our search strategy generated 31,288 articles, of which only 225 studies matched our inclusion criteria and were included in the systematic review. For the meta-analysis, only 88 studies were included. The pooled estimate of the prevalence of hypoglycaemia was 23.17% (95% CI 22.47–23.86), comprising a total of 743,055 patients with diabetes. The pooled episodes rate for hypoglycaemia was 47.47 per 1000 patient-years (95% CI 472.09–477.32), involving 45,083,541 diabetes patients. The pooled episodes rate estimate for hypoglycaemia stratified by type of diabetes was 149.19 episodes per 100 patient-years (95% CI 145.53–152.85) and 136.98 per 100 patient-years (95% CI 132.96–141.01) for T1DM and T2DM, respectively. The pooled episodes estimate stratified by treatment regimen was 184.62 (95% CI 180.60–188.64) for insulin-based therapy and 153.23 episodes per 100 patient-years (95% CI 142.90–163.55) for patients receiving a combination of insulin and oral therapy. The pooled incidence estimate of new cases of hypoglycaemia was 5.99 per 100 patient-years (95% CI 4.17–7.82), involving 148,609 participants. History of previous hypoglycaemia, old age, chronic kidney disease, intensive anti-diabetic therapy, insulin therapy and sulfonylureas were the main risk factors associated with hypoglycaemia.

Conclusion: Hypoglycaemia is very common among patients with diabetes. Further studies are needed to investigate hypoglycaemia-associated risk factors.

References:

• 1. Guariguata L., et al., Global estimates of diabetes prevalence for 2013 and projections for 2035. Diabetes Res Clin Pract 2014. 103:137–49

• 2. Cryer PE, Davis SN, Shamoon H. Hypoglycemia in diabetes. Diabetes Care. 2003;26:1902–12

Disclosure of Interest: None declared.

149 ISoP18-1210 Physician Attitudes Toward Pharmacists Advice on the Appropriateness of Prescribing

149.1 L. Garza Ocañas^*1, M. Galvan-Cantu², M. Rodríguez-Almendariz², J. D. Torres-Garza³, A. Pérez-Garza², E. Pérez-Rodríguez¹

149.1.1 ¹Facultad de Medicina y Hospital Universitario “Dr. José E. González”, Nuevo León, Mexico; ²Hospital Universitario “Dr. José E. González”, Universidad Autónoma de Nuevo León, Nuevo León, Mexico; ^;3Facultad de Medicina, Universidad Autónoma de Nuevo León, Monterrey, Nuevo León, Mexico

Background/Introduction: Medication Errors (ME) is a well-documented problem in pharmacovigilance. The analysis of the 2016 ME in our University Hospital shows that the most common ME are related to the prescription process (66%). The pharmacist advice on the appropriateness of the prescription is an important activity to reduce the number of prescription ME. Acceptance of the physician to pharmacist intervention on drug related problems is an activity that has been not studied previously in our Hospital.

Objective/Aim:

To describe the pharmacist interventions given as the number of interventions in all prescription reviewed during the study period as well as response of physicians toward pharmacists interventions on drug related problems.

Methods:

The study was conducted at Hospital Universitario “Dr. José E. González”, Monterrey, Nuevo León, México, from January to December 2017. The surveillance on the appropriateness of the prescription was made at the Internal Medicine, Pediatrics, Surgery and Ginecology and Obstetrics Departments. Age, sex, type of problem (drug, dose, frequency, route of administration), intervention and outcome (physiciansattitudes) were prospectively recorded.

Results:

Of a total of 719 interventions, the overall acceptance rate was 67%. Most of the interventions were made in prescriptions ordered by surgeons, followed by internal medicine physicians and pediatricians. Antibiotics and analgesic were the most common drug involved. The most common drug related problem were the improper prescription and dosage problem. The main pharmacist interventions were dose adjustment (24%) and optimization of routes and frequency of administration (23%).The type of drug related problem influenced physician attitude, with pharmacist participation being most acceptable for errors involving dose.

Conclusion:

The study indicates that most of the physicians accepted pharmacist counseling with respect to potential adverse drug reactions, related to medication errors due to inappropriate prescription. Training courses are required for further motivation of physicians on importance of pharmacist advice to reduce the number of medication errors due to potentially inappropriate prescription.

Disclosure of Interest: None declared.

150 ISoP18-1211 Analysis of Adverse Drug Reactions Using Patient Safety Network (Psn) Severity Score Method Compared to Naranjo Algorithm

150.1 F. A. Al-Braik^*1, M. M. Tashtoush¹, R. Saad¹, M. Y. Hasan²

150.1.1 ¹Pharmaceutical Department, Abu Dhabi Health Services, Abu Dhabi, United Arab Emirates; ²College of Medicine & Health Silences, UAE University, Al Ain, United Arab Emirates

Background/Introduction: Voluntary reporting is a recognised method of recording adverse drugs reactions (ADRs) in patients. ADRs can be reported utilizing Patient Safety Network (PSN) system available for healthcare professionals at facilities accredited by Joint Commission¹.

Objective/Aim: This study examined incidents of ADRs at Abu Dhabi Health Services (SEHA) a government healthcare network adopted PSN as part of patient records. The study assessed ADRs using PSN Severity score method compared to Naranjo Algorithm.

Methods: 264 ADRs cases were reported during 2017.A Taskforce was assigned for assessing causality and evaluating outcomes through two different methods calculating severity score of PSN and probability score of Naranjo Algorithm.

Results: Highest class involved were: antibacterials (n = 85, 32.2%), anticancer (n = 39, 14.8%), iron (n = 36, 13.6%), NSAIDs (n = 27, 10.2%), opioids (n = 15, 5.7%), anti-nausea (n = 11, 4.2%), vaccines (n = 7, 2.7%) and autoimmune drugs (n = 6, 2.3%). 10 most common medications were: Ferric Carboxymaltose (n = 34, 27.6%), Ceftriaxone (n = 16, 13%), Ciprofloxacin (n = 10, 8.1%), Moxifloxacin (n = 10, 8.1%), Vancomycin (n = 10, 8.1%), Piperacillin + Tazobactam (n = 9, 7.3%), Amoxicillin- Clavulanate (n = 8, 6.5%), Oxaliplatin ((n = 9, 7.3%) and Carboplatin (n = 8, 6.5%).

187 reports were assessed comparing Naranjo Algorithm with PSN Severity Score. According to Naranjo 158(84.5%) cases classified probable, 22(11.8%) possible, 6(3.2%) highly probable and 1(0.5%) doubtful. As per PSN-Severity Score 46(25.1%) cases classified mild, 101(55.2%) moderate, 34(18.6%) sever and 1 case (1.1%) life threating. 5 cases were not identified. Using PSN-harm Score analysis 1(0.5%) case of unsafe condition (harm score = 1), 1(0.5%) case of near miss (Harm Score = 2), 24 (12.8%) cases of no harm evident, physical or otherwise (Harm score = 3), 34 (18.1%) cases of emotional distress or inconvenience (harm score = 4), 123 (65.4%) cases required additional treatment (Harm score = 5) and 5 (2.7%) cases classified as temporary harm (Harm score = 6).Most common outcome was hypersensitivity reported with Antibacterials. Shortness of breath and hypersensitivity reactions such as itching and skin rashes were reported with chemotherapy agents and Ferric medications.

Conclusion: A similar severity assessment result was obtained comparing Naranjo Algorithm with PSN Severity Score [2]. Using PSN added a further dimension of measuring harm score. Majority of ADRs are preventable through improved health professions education and prescribing monitoring [3,4]. Monitoring the trend of ADR and severity and implementing measures for rational use of medicines to improve patient safety is a vital measure

References:

• 1. Madan A, Borckardt D, Borckardt JJ, Herbert J, Cooney H. A new approach to tracking the harmfulness of medical errors in health care systems. Qual Manag Health Care 2010;19:298–303

• 2. Murayama H, Sakuma M, Takahashi Y, Morimoto T. Improving the assessment of adverse drug reactions using the Naranjo Algorithm in daily practice: The Japan adverse drug events study. Pharmacol Res Perspect 2018;6(1) https://doi.org/10.1002/prp2.373

• 3. Angamo MT, Chalmers L, Curtain CM, Bereznicki LR. Adverse-Drug-Reaction-Related hospitalisations in developed and developing countries: a review of prevalence and contributing factors. Drug Saf 2016;39:847–57

• 4. 4-Ganesan S, Sandhiya S, Reddy KC, Subrahmanyam DK, Adithan C. The impact of the educational intervention on knowledge, attitude, and practice of pharmacovigilance toward adverse drug reactions reporting among health-care professionals in a tertiary care hospital in South India. J Nat Sci Biol Med 2017;8:203–9

Disclosure of Interest: None declared.

151 ISoP18-1212 Risk Factors of Major Bleeding in Patients Prescribed Rivaroxaban in Primary Care in England: Based on a Modified-Prescription Event Monitoring Study

151.1 D. Roy^1,2, S. Dhanda^1,2, L. Wise¹, S. Shakir^*1,2

151.1.1 ¹Drug Safety Research Unit, Southampton, United Kingdom; ²University of Portsmouth, Portsmouth, United Kingdom

Background/Introduction: Clinical trials and observational studies have reported bleeding risk in patients taking oral anticoagulants. It is valuable to understand the predictors for major bleeding in patients prescribed rivaroxaban in primary care in England.

Objective/Aim: Multivariable logistic regression (MLR) analyses to explore potential risk factors for major bleeding within gastrointestinal (GI), urogenital (UG) and intracranial (IC) sites.

Methods: A case/non-case design was used to study the association between clinical risk factors and major bleeding in a cohort of rivaroxaban patients (N = 17546) in a single-arm Modified-Prescription Event Monitoring study identifying patients from dispensed prescriptions in England (2012–2016), followed for 12 months. Clinical risk factors for bleeding and bleeding outcomes were collected from prescribing general practitioners via questionnaires sent at ≥ 3, and ≥ 12 months observation. Univariate and multivariable logistic regression analyses were performed to examine the association for each different site. Multivariable analyses models were based on 2 different clinical approaches; Clinical Risk Factors Selection (CS) model and HAS-BLED (HB) model. The CS model included all reported clinical risk factors for bleeding. The HAS-BLED model included the HAS-BLED clinical risk score categories (low, moderate or high risk) and gender. Statistically significant (p < 0.05) associations from the MLR models are presented in the results section.

Results: Risk factors for Major GI bleed (n = 176).

CS Model: Age 65–74 vs < 65 years: OR = 2.4 [95% CI 1.3, 4.6]; Age ≥ 75 years vs < 65 years: OR = 4.2 [95% CI 2.3, 7.5]; Predisposition to or history of bleeding: OR = 4.8 [95% CI 3.1, 7.5].

HB Model: Moderate vs low: OR = 4.0 [95% CI 2.1, 7.6]; High vs low: OR = 8.9 [95% CI 4.0, 19.9].

Risk factors for Major UG bleed (n = 36).

CS Model: Age 65–74 years vs < 65 years: OR = 0.2 [95% CI 0.1, 0.7]; Females vs males: OR = 2.9 [95% CI 1.4, 6.1]; Malignancy: OR = 2.6 [95% CI 1.1, 6.3].

HB Model: Females vs males: OR = 2.7 [95% CI 1.3, 5.6].

Risk factors for Major IC bleed (n = 57).

CS Model: Age ≥ 75 years vs < 65 years: OR = 2.8 [95% CI 1.1, 6.9]; History of cerebrovascular accident/transient ischaemic attack (CVA/TIA) (including haemorrhagic CVA): OR = 2.2 [95% CI 1.3, 3.9]; Predisposition to or history of bleeding: OR 2.6 [95% CI 1.0, 6.7].

HB Model: Moderate vs low: OR = 3.3 [95% CI 1.2, 9.1]; High vs low: OR = 9.0 [95% CI 2.5, 31.9].

Conclusion: Use of the case/non-case design to explore risk factors for bleeding in rivaroxaban patients using the CS model identified age ≥ 65 and history of bleeding or predisposition as statistically significant prognostic factors for major GI bleeds. For major UG bleeds, age group 65–74 years, female gender (which may be related to vaginal bleeding including menorrhagia) and malignancy were statistically significant in the CS model. For major IC bleeds, age ≥ 75 years, history of CVA/TIA and history of bleeding or predisposition to bleeding were identified as statistically significant risk factors in the CS model. The HB model showed that moderate and high risk scores were statistically significant risk factors for major GI and IC bleeds and gender statistically significant for major UG bleeds. Overall, findings from the CS model and the HB model are in keeping with known clinical risk factors for bleeding.

Disclosure of Interest: D. Roy Other: The DSRU is an independent charity (No. 327206) which works in association with the University of Portsmouth. The DSRU has received funding from Bayer, the manufacturer of Xarelto^®. The DSRU makes the final decision on the publication of external communications. S. Dhanda Other: The DSRU is an independent charity (No. 327206) which works in association with the University of Portsmouth. The DSRU has received funding from Bayer, the manufacturer of Xarelto^®. The DSRU makes the final decision on the publication of external communications. L. Wise Other: The DSRU is an independent charity (No. 327206) which works in association with the University of Portsmouth. The DSRU has received funding from Bayer, the manufacturer of Xarelto^®. The DSRU makes the final decision on the publication of external communications. S. Shakir Other: The DSRU is an independent charity (No. 327206) which works in association with the University of Portsmouth. The DSRU has received funding from Bayer, the manufacturer of Xarelto^®. The DSRU makes the final decision on the publication of external communications.

152 ISoP18-1215 Drug Safety: In Vitro Study of Physicochemical Incompatibilities of Infusion Medications Linked to pH Variation

152.1 I. Bennani¹, Z. El Asil², A. Cheikh³, H. Mefetah⁴, M. Draoui¹, M. Bouatia^*1

152.1.1 ¹Laboratory of analytical chemistry Faculty of Medicine and Pharmacy of Rabat, University Mohamed V, Rabat, Morocco; ²Faculty of Medicine and Pharmacy of Rabat, University Mohamed V, Rabat, Morocco; ³Faculty of Pharmacy, Abulcasis University, Rabat, Morocco; ⁴Pediatrics Hospital, CHIS, Rabat, Rabat, Morocco

Background/Introduction: The administration of incompatible products could indeed cause the appearance of a precipitate or the inactivation of one or other of the active ingredients considered. PH changes in infusion medications after contact with other acidic or basic drugs (or solutions) may lead to the formation of a precipitate which can obstruct the pathway and hinder the passage of the drug to the patient; In addition, if the precipitate enters the bloodstream without being solubilized, it can obstruct the blood vessels and cause various problems in the more or less long term.

Objective/Aim: The aim is to detect and investigate the visible physicochemical incompatibilities of injectable drugs commonly used in a resuscitation service with other acid or basic drugs in an infusion, by a simulation model to test the behavior of these drugs. Drugs with respect to the pH change.

Methods: We carried out this in vitro experimental study at the laboratory of analytical chemistry at the Faculty of Medicine and Pharmacy of Rabat, and at the Children’s Hospital of Ibn Sina University Hospital in Rabat.

We selected 31 parenteral drugs most commonly used at the children’s hospital of Ibn Sina University Hospital in Rabat.Powder drugs are prepared using the exact amounts of water for injection required for preparations for injection.Mixtures of these test drugs were then made with a solution of hydrochloric acid (HCl) and sodium hydroxide solution (NaOH) to reveal drug incompatibilities via the formation of precipitates by visual inspection, or by infrared spectrophotometry to identify their nature.

Results: The most important results obtained are summarized in the following table:

Conclusion: This study allowed us to reveal several physicochemical incompatibilities of injectable drugs related to pH, it also allowed us to establish a recommendation that prevents the simultaneous administration to the patient by the same route an acid drug with a basic drug.

Disclosure of Interest: None declared.

153 ISoP18-1216 Pharmacovigilance in Bioequivalence Studies: Is there a Difference in the Prevalence of Adverse Events Between Original and Generic Medicines?

153.1 A. Cheikh¹, M. Bouatia^*2, M. R. Ajaja¹, E. A. Faouzi¹, A. Bouklouze³, A. Benomar¹, Y. Cherrah¹

153.1.1 ¹Bioequivalence Center, Abulcasis University, Rabat, Morocco; ²Chemistry, ³Bioequivalence Center, Mohammed V University, Faculty of Medicine and Pharmacy, Rabat, Morocco

Background/Introduction: Bioequivalence studies are conducted to demonstrate that the pharmacokinetic profile of the original and generic drugs is the same, which could conclude that the pharmacodynamic effect would be the same. These studies can also provide insights into the safety profile of original and generic drugs through the collection of adverse events reported in clinical trials.

Objective/Aim: Our aim trought this study was to assess the prevalence of adverse events in bioequivalence studies conducted in our bioequivalence center. Also, we aimed to compare the prevalence of adverse events notified with generic and original medicines used in bioequivalence studies.

Methods: We calculated the prevalence of adverse events reported in bioequivalence studies conducted in 1 year. We studied all adverse events notified by investigators in bioequivalence studies in our center. The French method was applied to study the imputability between notified adverse events and the administered molecules. Chi² test was used to compare between adverse events notified by using generic and original medicine. Statistical analysis were performed with SPSS 13.0.

Results: 49 adverse events were notified in the 10 bioequivalence studies involving 196 volunteers with the frequency of 4.9 per study and 0.125 per volunteer. The prevalence of adverse events was about 12.5% in our studies population. The prevalence was different depending on the therapeutic class. The prevalence was 58.3% for bisphosphonates, 26.7% for antidepressants, 0% for statins, 0% for betablockers, 4.4% for antihypertensives, 20% for analgesics, 0% for anivirals and 0% for non-steroidal anti-inflammatory medicines. There is no significant differnce in prevalence of adverse events with generic and original medicines (p  = 0.271). All adverse events reported in all studies were reported in the Summary of Product Characteristics for the administered molecules.

Conclusion: This study showed that even with a single dose of each drug, the volunteers saw adverse events whose frequency and severity differed from one molecule to another and from one volunteer to another. Our results were different from those found in other foreign studies [1,2].

Generics should provide pharmaceutical quality similar to the originator drug framed by regulatory guidelines [3,4]. Because generic drugs do not undergo Phase I, II and III clinical trials, bioequivalence testing may be an important opportunity to detect and report adverse events in relation to the administration of generic medicines.

References:

• 1. Gurer C, Pehlivanli AC, Demircigil GC. Pooled bioequivalence study database from Turkey: characterization of adverse events and determination of split points based on Gini Index as a promising method. Springer Plus 2016; 5:709

• 2. Sibille M, Deigat N, Janin A, Kirkesseli S, Durand DV. Adverse events in phase-I studies: a report in 1015 healthy volunteers. Eur J Clin Pharmacol 1998;54:13–20

• 3. United States Food and Drug Administration (2003) Guidance for Industry, Bioavailability and bioequivalence studies for orally administered drug products—General Considerations. http://www.fda.gov/ohrms/dockets/ac/03/ briefing/3995B1_07_GFI-BioAvail-BioEquiv.pdf. Accessed 15 Nov 2015

• 4. European Medicines Agency (2010) Guideline on the investigation of bioequivalence. http://www.ema.europa.eu/docs/ en_GB/document_library/Scientific_guideline/2010/01/WC500070039.pdf. Accessed 15 Oct 2015

Disclosure of Interest: None declared.

154 ISoP18-1217 Material Vigilance in a Hospital

154.1 H. Bechar¹, A. Tebaa^*2, R. Soulaymani²

154.1.1 ¹National Institute of Oncology, ²Ministry of Health’s Poison Control and Pharmacovigilance Center, Rabat, Morocco

Background/Introduction: Materiovigilance is a necessity to monitor medical devices after they are placed on the market and during their life cycle in order to avoid any incident resulting from their use.

Objective/Aim: 1. Describe the frequency of materiovigilance declarations.

2. Evaluate the activity within all the services of the National Institute of Oncology of Ibn Sina hospital in Rabat.

Methods: This is a retrospective study of notifications reported by INO services between January 2017 and December 2017 using the notification forms provided by the national pharmacovigilance center. The results were analyzed according to their severity and their impact on the patient.

Results: During these 12 months, 276 notifications were collected by the pharmacovigilance unit. The prominent cases of these declarations have been classified according to their frequency and severity.

Examples of significant cases:

• 1. Elastomeric diffusers: leakage of cytotoxic products (after contact with the company, it was decided to remove the lot and replace it)

• 2. Venous catheter (intranule): perforation of the vein and hematoma.

After investigation which demonstrated the bad technique of use: user training.

Conclusion: Spontaneous notification has enabled health professionals to become aware of the risks associated with medical devices. The early declaration of any incident can reinforce the safety of the patient hence the interest to sensitize the health professionals to notify all the noticed events.

Pharmacovigilance Register of the National Institute of Oncology.

Disclosure of Interest: None declared.

155 ISoP18-1219 Dress Syndrome Associated with Diclofenac

155.1 O. Charfi^*1, S. Kastalli¹

155.1.1 ¹National Centre of Pharmacovigilance of Tunisia, Tunis, Tunisia

Background/Introduction: Drug rash with eosinophilia and systemic symptoms (DRESS syndrome) is a rare and severe potentially life-threatening condition with a mortality rate of about 10%. Most frequent drug associated with DRESS syndrome include anti epileptic drugs, allopurinol and sulfonamid. Non steroidal anti inflammatory drugs (NSAID) had been rarely associated with this syndrome.

Objective/Aim: Herein, we report the first case of diclofenac induced DRESS syndrome.

Methods: A 53-year-old woman was prescribed for back pain diclofenac 50 mg daily. Ten days after starting the treatment, she presented with fever, cough and multiple lymphadenopathy. She stopped diclofenac and was prescribed amoxicilline, paracetamol, mefenamic acid and cetirizine. One day following the first intake of these drugs, she developed generalized rash involving the trunk and the members. She was admitted to the hospital.

Clinical examination revealed temperature at 39.5 °C, generalized maculopapular pruritic rash, cervical, axillary and inguinal lymph nodes.

Blood tests showed high eosinophils count 1830/mm³. Plasmatic creatinine level was increased at 213 µmol/l. Liver enzymes were normal.

Serology for Epstein Barr virus (EBV), Cytomegalovirus (CMV) and human immunodeficiency virus (HIV) were negative.

Histological findings were compatible with DRESS syndrome.

DRESS syndrome was confirmed and all the medications were stopped except cetirizine.

Skin condition improved slowly and laboratory parameters returned to normal level in about 16 days.

Results: The diagnosis of DRESS syndrome was made in this patient based on the criteria adopted by the European group RegiSCAR. In our case, the RegiScar score was 5 (probable case):

Generalized skin rash (1), eosinophilia (2), kidney involvement (1) enlarged lymph nodes (1).

Diclofenac was suspected to be the responsible drug for the DRESS syndrome in this case based on the French method of imputability. The score for diclofenac was I2 (possible).

In the literature, NSAI drugs have been rarely reported with DRESS syndrome as a causative agent. It was associated with piroxicam, ibuprofen and celecoxib. All the patients presented with liver involvement. In our case the patient had a normal hepatic enzymes but she developed a renal involvement.

Conclusion: To our knowledge, this is the first case of DRESS syndrome induced by diclofenac.

Disclosure of Interest: None declared.

156 ISoP18-1220 Doxycycline Induced Generalized Bullous Fixed Drug Eruption: 2 Case-Reports

156.1 O. Charfi^*1,2,3, G. Lakhoua³, S. Ben Hammamia³, A. Zaiem³, I. Aouinti³, S. El Aidli³, S. Kastalli²

156.1.1 ¹Service de Recueil et d’analyse des Données, ²National Centre of Pharmacovigilance of Tunisia, Tunis, Tunisia; ³Service de Recueil et analyse des Données, National Centre of Pharmacovigilance of Tunisia, Tunis, Tunisia

Background/Introduction: Fixed drug eruption (FDE) is a cutaneous drug reaction that typically recurs in the same site on re-exposure to the culprit drug. Usually it presents as a single plaque on erythematous background. On re-exposure to the culprit dug, old lesions reappear and widespread bullous eruptions rarely occur. Generalized bullous FDE is an extremely uncommon variant, characterized by widespread blisters and erosions mimicking Stevens-Johnson syndrome and toxic epidermal necrolysis.

Objective/Aim: Herein we report 2 exceptional cases of doxycycline induced generalized bullous fixed drug eruption.

Methods: Case report n°1

A 44-year-old woman presented with a 3 days history of multiple itchy erythematous and bullous lesions. These lesions appeared 4 h after the first intake of doxycycline 200 mg. This drug was prescribed for a genital infection.

Physical examination revealed multiple well-defined bullous and erosive lesions over the lower back, the arms and the face. Lesions size varied from 3 to 6 cm. Oral and genital mucosae were not involved.

Medical questioner revealed a history of erythematous lesions involving the same localization few hours following the first intake of doxycycline 2 years ago. These lesions healed without hyperpigmented sequelae.

The patient was prescribed anti histaminic drug. Skin condition improved within 10 days.

Case report n°2

A 62-year-old woman developed less than 1 h after the first intake of docycycline, multiple erythematous and bullous lesions involving left leg, the buttock and upper members. The size of the lesions ranged from 4 to 7 cm.

The patient noticed a history of erythematous lesions involving the same localization few hours following the first intake of doxycycline 1 year ago. These lesions healed with hyperpigmented sequelae.

Skin condition improved in 15 days following drug withdrawal.

Results: The responsibility of doxycycline in inducing bullous fixed drug eruption was evaluated according to the French method of imputability as very likely in both cases because of the suggestive delay and the positive rechallenge.

Generalized bullous FDE is a relatively rare skin condition. Exceptional cases were associated to mefenamic acid, naproxen, cetrizine, and nicotinic acid.

In literature, only 2 cases of doxycycline induced Generalized bullous FDE were reported.

In both cases, lesions appeared few hours after the first drug intake and the patients had history of fixed drug eruption following doxycycline intake.

Conclusion: Present cases highlight an uncommon reaction to a commonly used drug with positive rechallenge. Physicians should be more careful on patient’s medical history to avoid such events.

Disclosure of Interest: None declared.

157 ISoP18-1221 Drug Sefty: In-Vitro Study of Physicochemical Incompatibilities of Injectable Antibiotics Used in Pediatrics with Other Drugs in an Infusion

157.1 I. Bennani¹, Z. El Asil², H. Attjioui¹, A. Cheikh³, H. Mefetah⁴, M. Draoui¹, M. Bouatia^*1

157.1.1 ¹Laboratory of Analytical Chemistry, Faculty of Medicine and Pharmacy of Rabat, Rabat, Morocco; ² Faculty of Medicine and Pharmacy of Rabat, University Mohamed V, Rabat, Morocco; ³Departement of Pharmacy, Abulcasis University, Rabat, Morocco; ⁴Departement of Pharmacy, Pediatrics Hospital, Rabat, Rabat, Morocco

Background/Introduction: Drug incompatibilities are physical and chemical reactions that occur in vitro when two or more drug solutions are combined in one syringe, even infusion fluid, or in the same tube or bottle. During the preparation or administration of the medication.

Objective/Aim: the aim is to detect and investigate visible physicochemical incompatibilities of injectable antibiotics commonly used in pediatrics with other drugs in an infusion.

Methods: We carried out this in vitro experimental study at the laboratory of analytical chemistry at the Faculty of Medicine and Pharmacy of Rabat, and at the Children’s Hospital of Ibn Sina University Hospital in Rabat.We selected 31 parenteral drugs most commonly used at Children’s Hospital of Rabat University Hospital Ibn Sina including antibiotics. form of powder are prepared using the exact amounts of water for injection required for the preparations intended for injection.After mixing of these drugs with the antibiotics to be tested to reveal the drug incompatibilities by visual inspection, and cases of formation of precipitates will be analyzed by infrared spectrophotometry to identify their nature.

Results: The mixture of Amoxicillin/Clavulanic acid with the drugs allowed us to detect 13 drug incompatibilities for example with: Adrenaline and Noradrenaline Cefixime and Cefuroxime, Imipenem/Cilastatin, Trimethoprime, Vancomycin, Furosemide, Ganciclovir, Hydrocortisone, Midazolam, Omeprazole, and with Sugammadex.

The mixture of ceftriaxone with the 30 drugs allowed us to reveal 8 drug incompatibilities, the combination of cefuroxime with the drugs allowed us to detect 13 drug incompatibilities, the mixture of Ganciclovir with the drugs allowed us to detect 17 drug incompatibilities, trimethoprime revealed 6 drug incompatibilities, and Vancomycin revealed 10 drug incompatibilities.

Conclusion: This study allowed us to reveal several physicochemical incompatibilities of injectable drugs, it also allowed us to show the validity of the recommendation that prevents the administration of drugs simultaneously infusion.

Disclosure of Interest: None declared.

158 ISoP18-1222 In Vivo Tests for Investigating Immediate Hypersensitivity to General Anesthetics

158.1 N. Amdouni ¹, M. Bouhlel^*1, S. El Aidli¹, R. Daghfous ¹, S. Kastalli¹, A. Zaiem ¹

158.1.1 ¹Tunisian National Centre of Pharmacovigilance, Tunis, Tunisia

Background/Introduction: Immediate hypersensitivity (IHS) to general anesthetics (GA) can be life threatening for patient ongoing medical procedure. On the other hand, considering a general anesthesia in a patient with history of IHS to general anesthetics must multiply the measures of prevention of recurrence of such an event. In this context anesthesiologist and allergist should collaborate.

Objective/Aim: To study the clinical and epidemiological characteristics of patients suspected of having presented IHS to general anesthetics and who had benefited from in vivo allergic tests as well as the contribution of these tests.

Methods: Retrospective study of the files of patients suspected of having presented IHS to general anesthetics, who were referred to the national pharmacovigilance center of Tunis between January 2009 and December 2017 and in whom in vivo allergic tests were performed.

Results: Twenty-two patients were selected, the sex ratio m/f was 0.29, the age ranged from 1 year to 71 years with a median of 42.5 years. Nine patients were referred by the attending physician. Twelve patients have had at least one prior general anesthesia. Six patients had a history of drug hypersensitivity. In 6 patients the nature of the anesthetic was unknown. The time to perform skin tests ranged from 18 days to 20 years with a median of 4 months. GA used before the event were: morphine 54%, hypnotic 25% and curare 21%. Of the 22 patients tested, only four patients were positive to GA: three to cisatracurium and one to atracurium. The follow-up of 13 patients showed that nine patients were anesthetized without problems. For the other nine patients we had no feedback.

Conclusion: In vivo allergic tests to GA, are interesting for the orientation of the choice to the anesthetic to which they are negative, and for the determination of the responsible agent when they are positive. However, these tests remain limited and would be better interpretable by associating them with the results of the pharmacovigilance survey and biological assays.

Disclosure of Interest: None declared.

159 ISoP18-1224 Comparison of 4 Commonly Prescribed Antipsychotics Patient Information Leaflets of Indian Companies with the Original Smpc from Innovator Companies

159.1 P. Biswas^*1, H. Biswas¹

159.1.1 ¹Symogen Limited, Bourne End, United Kingdom

Background/Introduction: In recent times there have been several safety issues with antipsychotic drugs and cardiovascular ADRs, including withdrawal of thioridazine. In an era where consumer awareness is at an all-time high and rapidly increasing, information available to patients and prescribers for safe and efficient use of medicines is still an area of concern.

Objective/Aim: Compare the PILs of the most commonly prescribed antipsychotics of the Indian manufacturers with that of the SmPCs from the Innovator manufacturers and review if the package inserts adhere to standard guidelines.

Methods: PILs for the Indian manufacturers and SmPCs from the Innovator manufacturers were obtained from the internet. Both the documents were then compared for each section and the missing information specifically in respect to cardiovascular ADRs. The drugs for which the SmPC and PIL were compared were Olanzapine; Quetiapine; Clozapine and Aripiprazole.

Results: Review and analysis of the PILs and SmPCs showed standard labeling guidelines were not adhered and details not updated in the PILs. Discrepancies were found in the undesirable effects section, contraindication, special warnings and precautions for use and interaction with other drugs. Serious adverse reactcions, i.e., QTc prolongation, Ventricular tachycardia/fibrillation, fatal myocardial infarction; cardiac arrest and bradycardia were missing in the PIL for drugs manufactured by the Indian companies when compared with the innovator companies.

Conclusion: Our research has highlighted that Indian companies do not have proper mechanism or process for development of package inserts with important sections on safety of these drugs missing with several discrepancies. Provision of good quality patient information is intended to supplement the prescribers and patients and not replace the advice given to patients by health professionals. As these drugs are used globally for the treatment of patients with schizophrenia and bi-polar disease, it would be beneficial to have globally harmonized information to be incuded in the PIL and SmPCs for consistency and accurate guidance to prescibers for patient safety.

Disclosure of Interest: None declared.

160 ISoP18-1226 Development and Validation of a Model Predictive of Case Inclusion in Pharmacovigilance Reviews

160.1 M. Munoz^*1,2, G. Dal Pan¹, J. Wei², C. Delcher³, H. Xiao², C. Kortepeter¹, A. Winterstein^2,4

160.1.1 ¹Office of Surveillance and Epidemiology, US Food and Drug Administration, Silver Spring, Maryland United States; ²Department of Pharmaceutical Outcomes and Policy; ³Department of Health Outcomes and Biomedical Informatics; ⁴Department of Epidemiology, University of Florida, Gainesville, Florida, United States

Background/Introduction: The rapidly expanding size and complexity of the FDA Adverse Event Reporting System (FAERS) database requires data-driven pharmacovigilance practices. Techniques to systematically identify and distinguish higher utility individual case safety reports (ICSRs) from lower utility ones will support improved management of safety signals.

Objective/Aim: We aimed to develop and validate a model predictive of an ICSR’s pharmacovigilance utility (PVU).

Methods: We used a retrospective cohort of FDA pharmacovigilance reviews completed in 2016 to determine the association between PVU and ICSR features. PVU was operationalized as an ICSR’s inclusion in a case series supporting a recommendation to modify product labeling. Characteristics of ICSRs included in a case series and those excluded were compared using univariate analyses. Next, we developed multivariable logistic regression models to examine likelihood of case inclusion based on: all available variables, completeness variables only, and a parsimonious model using backward elimination. Of these models, we selected the best performing model for bootstrapping validation based on fit, discriminative ability, and calibration. As a sensitivity analysis, we evaluated the validated model’s performance across subgroups of reviewed safety issues.

Results: For model development, we identified 69 pharmacovigilance reviews containing 10,381 ICSRs, of which 2115 ICSRs were included in a case series. The parsimonious model was selected for validation as it had the best discriminative ability with a C-statistic of 0.71. The strongest predictors of ICSR inclusion were reporting of a designated medical event (OR 1.93, 95% CI 1.54–2.43), positive dechallenge (OR 1.67, 95% CI 1.50–1.87), and reason for product use provided (OR 1.60, 95% CI 1.40–1.83). The strongest predictors of ICSR exclusion were death reported as the only outcome (OR 2.72, 95% CI 1.76–4.35), > 3 suspect products (OR 2.69, 95% CI 2.23–3.24), and > 15 events reported (OR 2.69, 95% CI 1.90–3.82). Our sensitivity analysis demonstrated heterogeneity in model performance by review issue. The correct classification of ICSRs was similar to the overall results when the model evaluated hypersensitivity reactions and drug-induced liver injury (C-statistic of 0.70 and 0.74, respectively); however, less discriminative ability was demonstrated with cardiovascular events and events without acute life-threatening outcomes (C-statistic of 0.64 and 0.58, respectively).

Conclusion: Our model demonstrated the feasibility of predicting the PVU of an ICSR. The model’s modest discriminative ability highlights opportunities for further enhancement and suggests algorithms tailored to particular safety issues may be beneficial.

Disclosure of Interest: None declared.

161 ISoP18-1227 Pharmacists’ Role in Risk Communication on Self-Medication: Pilot Study from Bulgaria

161.1 H. V. Lebanova^*1, R. Staynova², V. Getova²

161.1.1 ¹Department of Pharmaceutical Sciences and Social Pharmacy, Faculty of Pharmacy, Medical University-Pleven, Pleven, Bulgaria; ²Department of Pharmaceutical Sciences, Faculty of Pharmacy, Medical University—Plovdiv, Plovdiv, Bulgaria

Background/Introduction: Self-medication is defined as the selection and use of medicines by individuals (or a member of the individuals’ family) to treat self-recognized or self-diagnosed conditions or symptoms [1]. Appropriate self-medication has many benefits but it is also linked to potential risks such as incorrect diagnosis, delays in seeking medical advice when needed, infrequent but severe adverse reactions, dangerous drug interactions, incorrect manner of administration, incorrect dosage, incorrect choice of therapy, masking of a severe disease and risk of dependence and abuse [2].

Objective/Aim: The aim of the study was to assess pharmacists’ attitudes and previous experience with risk communication on self-medication in community pharmacies in Bulgaria.

Methods: An anonymous, questionnaire-based, descriptive study was performed. A prevalidated close-ended questionnaire was distributed among random sample of community pharmacists in Bulgaria. Data was analyzed using SPSS v.19.

Results: A total of 99 questionnaires were collected. 60.6% of the responders claim that patients very often consult them about their self-medication practices but rarely about possible adverse drug reactions and other drug-related problems such as drug interactions. 51.5% of pharmacists consider presenting information about risks as a mandatory part of the pharmaceutical consultation and the majority of them (87.9%) believe that it could affect patients’ decisions. According to 18.2% of the responders up to 80% of self-medicating patients do not use their over-the-counter medicines as recommended in the patient leaflet. 63.6% of the responders believe that over-the-counter medical products have high potential for drug misuse and abuse and 12.1% consider them with a substantial potential for drug interactions.

Conclusion: Data suggests that patients consider over-the-counter medicines safe and rarely seek information about possible risks. The conducted pilot study showed that pharmacists should be the primary and pro-active source of information about risks of self-medication products.

References:

• 1. Hughes CM, McElnay JC, Fleming GF.Benefits and risks of self medication. Drug Saf 2001;24, 1027–37

• 2. Ruiz ME. Risks of self-medication practices. Curr Drug Saf 2010;5:315–23

Disclosure of Interest: None declared.

162 ISoP18-1228 Rebound Effect after Discontinuation of Denosumab: Case Evaluation of the Regional Pharmacovigilance Center Bern in Switzerland

162.1 S. Banholzer^*1,2, M. Haschke^1,2, I. Scholz^1,2

162.1.1 ¹Clinical Pharmacology and Toxicology, Department of General Internal Medicine, Inselspital, Bern University Hospital, Bern, Switzerland; ²Institute of Pharmacology, University of Bern, Bern, Switzerland

Background/Introduction: Denosumab is a human monoclonal antibody that binds with high affinity and specificity to the receptor activator of NF-κB ligand, which is important for the maturation of osteoclasts. As a result denosumab reduces bone resorption and increases bone mass and strength [1]. The effect of the therapy is limited to the duration of treatment and discontinuation of denosumab is associated with a rapid loss of bone mass [2].

Objective/Aim: To describe the frequency and variety of rebound cases after discontinuation of denosumab reported to our regional pharmacovigilance center (RPVC), to the regulatory authorities in Switzerland, in Europe and worldwide.

Methods: Retrospective analysis of reported rebound cases in Switzerland, Europe and using the WHO-Pharmacovigilance database in the time period between January 2013 and May 2018 [3]. Detailed characterization of the cases reported to the RPVC Bern within the same time span. In particular, we are interested in the following points (1) reported fractures (2) osteoporosis treatment after discontinuation of denosumab (3) reported bone-damaging co-medication.

Results: The number of all reported cases in the WHO-Pharmacovigilance database [3] and from our RPVC are shown in the table. From 116 cases reported to the RPVC Bern, 94 cases were found regarding “rebound effect” after discontinuation of denosumab. 35% of these cases were related to bone fractures of any location [vertebral fractures (85%), vertebral and non-vertebral fractures (12%) and non-vertebral fractures (3%)].

In 26% of the cases, no follow-up treatment was performed after discontinuation of denosumab and in 74% of the cases a follow-up treatment with bisphosphonates (64.4% zolendronat, 24.3% ibandronat, 10.0% alendronat, 1.4% risedronat) is documented. In 24.5% (23 of 94) of the cases a bone-damaging comedication was documented (most frequent: 38% proton pump inhibitors, 16% aromatase inhibitors, 16% systemic glucocorticoids).

Conclusion: Our data show that 90% (124 of 138) of all ‘rebound effects’ worldwide have been reported in Switzerland and that 68% (94 of 138) originated from our RPVC in Bern. There seems to be an accumulation in Switzerland or an underreporting outside of Switzerland. To prevent future adverse drug reactions, controlled prospective studies are necessary to determine optimal therapy duration and follow-up treatment.

References:

• 1. Arzneimittelinformation-Publikationssystem (AIPS), (electronic version). Swissmedic, Bern, Switzerland. Available at: http://www.swissmedicinfo.ch/. [Internet]. Available from: www.swissmedicinfo.ch.

• 2. Meier C, Uebelhart B, Aubry-Rozier B, Birkhauser M, Bischoff-Ferrari HA, Frey D, et al. Osteoporosis drug treatment: duration and management after discontinuation. A position statement from the SVGO/ASCO. Swiss Med Wkly 2017;147:w14484

• 3. VigiLyze. Uppsala Monitoring Center—WHO individual case safety report (ICSR) database system. Available from: https://vigilyze.who-umc.org/.

Disclosure of Interest: None declared.

163 ISoP18-1231 The Purchase of Health Products via the Internet: Health Issues and Perspectives

163.1 H. Attjioui¹, Z. Aliat², A. Cheikh³, A. Tebaa⁴, I. Bennani¹, H. Mefetah⁵, M. Bouatia^*1

163.1.1 ¹Mohamed V University—Faculty of Medicine and Pharmacy, ²Mohamed V University- Faculty of Medecine and Pharmacy, CHIS, ³Mohamed V university—Faculty of Medicine and Pharmacy, Abulcasis university, ⁴Antipoison and Pharmacovigilance Center of Rabat, Ministry of Health, Rabat, Morocco; ⁵Paediatric Hospital, Pharmacy, Rabat, Morocco

Background/Introduction: The rise of the Internet and e-commerce is disrupting access to health products, and the offerings of these products on the Internet are often attractive from the point of view of the customer. However, it contributes significantly to the expansion of falsified drug trafficking internationally. According to the WHO, 62% of medicines bought on the internet are counterfeit products and 42% of these products manufactured worldwide are distributed in Africa.

Objective/Aim: To take stock of the significant growth of internet sales of health products with the health challenges encountered and to propose actions and perspectives adapted to public health issues.

Methods: We made a summary of the elements extracted from websites offering health products, data from the drug management and the national pharmacovigilance center, the data collected allowed us to identify the problem and study the courses of action. To consider to secure the purchase of drugs.

Results: Counterfeit medicines have many victims in Morocco, since they are the second largest source of poisoning (25%) after food in the country according to the Anti-poison and Pharmacovigilance Center of Morocco (CAPM).

The most common purchases include so-called “comfort” drugs. Indeed, the products intended to treat the sexual problems, the products against the obesity, the food supplements, are the products most concerned. Generic drugs and medical devices (lenses, breast implantations, etc.) are not spared from this problem, just like cosmetics. In addition, the counterfeiting of anti-cancer products is even more serious and widespread, especially because of their high cost in the conventional circuit. These products may contain compounds of inactive principles, others are over or under dosage, or composed of harmful ingredients.The majority of falsified health products are imported from producing countries (India, China, Turkey and Russia) and then resold at lower cost in other countries.

The online purchase of health products is not the main source of fake drug trafficking in Morocco. Indeed, the illegal importation of contraband drugs also constitutes a serious health risk for the population because these means of acquisition of drugs are doubtful.

Conclusion:

The sanitary dramas related to falsified medicines show that counterfeiting of medical products is not to be taken lightly, as it continues to make victims around the world. For that, it is necessary to set up adapted and responsible actions which begins by creating a regulatory framework governing the purchase of these products on Internet.

In this perspective, prevention campaigns must be implemented to raise the awareness of health professionals and the public by encouraging them to give priority to official distribution channels, to double their vigilance in the face of abnormally low prices and the presence of spelling errors, or grammar on the box and/or on the site, the quality of the packaging, or unusual packaging .

Disclosure of Interest: None declared.

164 ISoP18-1232 The Experience of using Khartoum Medicines Information Center (Khmic) as a Focal Point to Enhance Pharmacovigilance In Sudan

164.1 T. Yousef^*1, A. Abbas², T. Ali¹, A. Abdelghadir¹

164.1.1 ¹Medication Safety Department, Ministry of Health Khartoum State, Khartoum, Sudan; ²Pharmacy Department, Countess of Chester Hospital NHS Foundation Trust, Chester, United Kingdom

Background/Introduction: Introduction: Under reporting of adverse drug reactions (ADRs) and medication errors (MEs) is a major health concern globally [1–2]. Sudan has been a full member of the WHO International Drug Monitoring Program via the Uppsala Monitoring Center since 2008. However, reporting of ADRs via Vigiflow system has been consistently poor over the years for many reasons including time constraints, misconception about spontaneous reporting, bureaucratic reporting procedures and lack of knowledge and awareness [3]. KhMIC is located within Khartoum State Ministry of Health and works closely with its Medication Safety (MS) department. Due to KhMIC ‘s location in the capital of Sudan with access to resources, it answers enquiries from other states regarding all clinical and non-clinical  medication issues including ADRs, management of MEs and poisoning. Furthermore, in November 2017, KhMIC collaborated with UK colleagues to deliver training in pharmacovigilance and medication safety to representative of medicines information centres from other states thereby forming an active network for ADR data collection similar to the Yellow Card Scheme regional centres in the UK.

Objective/Aim: To use nationally collated data processed via KhMIC service for ADRs and ME detection.

Methods: All enquiries received at KhMIC via different methods (telephone, personal and electronically) between December 2017 and May 2018 were screened to select enquiries related to ADRs and MEs. Data was analysed and adverse reaction reports were submitted to the Sudan National Medicines and Poisons Board.

Results: Of a total of 3676 enquiries received between Dec 2017 and May 2018, 13 ADRs and 40 MEs were reported compared to 20 ADRs reports received directly from hospitals by regular pathways during the same period.

Conclusion: Sudan requires financial resources and electronic systems to enhance pharmacovigilance and medication safety activities. However, using KhMIC as a focal reporting centre may be another promising new channel to increase the number of reported ADRs and MEs and ultimately detect signals at national level.

References:

• 1. Pirmohamed M, James S, Meakin S, Green C, Andrew K. Adverse Reactions as a Cause of admissions to hospital: prospective analysis of 18,820 patients. BMJ 2004;329:15

• 2. Mansur JM. “Medication System and the Important Role of Pharmacists. Drugs Aging 2016;33:213–21

• 3. Elnour AA, Ahmed AD, Yousif MA, Shehab A. Awareness and Reporting of Adverse Drug Reactions Among Health Professionals in Sudan. Jt Comm J Qual Patient Saf 2009;35:324–9

Disclosure of Interest: T. Yousef: None declared, A. Abbas: None declared, T. Ali: None declared, A. Abdelghadir Employee of: employees at the Department of Health—Ministry of State Khartoum.

165 ISoP18-1233 Sulfasalazine-Induced Dress Complicated by Hemophagocytic Lymphohistiocytosis in an Adult Ulcerative Colitis Patient

165.1 N. Fathallah¹, B. Ouni¹, S. Larif¹, A. Saii², H. Hmouda³, R. Slim¹, C. Ben Salem^*1

165.1.1 ¹Pharmacovigilance Center of Sousse, Research Labratory, Sousse University, Sousse, Tunisia; ²Amen Medical Center, Tunis, Tunisia; ³Intensive Care Unit, Sahloul University hospital, Sousse, Tunisia

Background/Introduction: Drugs are a rare cause of Drug Reaction with Eosinophilia and Systemic Symptom (DRESS) and also of macrophage activation syndrome (MAS) and extremely rare cause of DRESS complicated with MAS.

Objective/Aim: To report the first case of sulfasalazine-induced DRESS complicated by MAS in an adult ulcerative colitis patient.

Methods: A 45-year-old male was diagnosed with ulcerative colitis, he was started treatment with oral sulfasalazine. Six weeks after drug initiation, the patient developed pruritic skin eruption associated with severe liver failure, so he was admitted to intensive care unit. He was diagnosed with DRESS associated of MAS.

Results: Drug hypersensitivity syndrome is among the most severe drug hypersensitivity reactions and in rare cases it may progress to hemophagocytic lymphohistiocytosis MAS is rare and severe complication of DRESS and often misunderstood [1].

MAS has been observed in immunocompromised patients, regardless of the cause of immunosuppression: pharmacologic, infectious, or neoplastic [2].

To date, the pathogenesis of MAS is not totally known. It is considered as ineffective immune response with activation and uncontrolled proliferation of T lymphocytes, macrophages and natural killer (NK) cells and the ingestion of cellular blood components and their precursors by these macrophages, and secretion of high levels of pro-inflammatory cytokines [2].

To the best of our knowledge, only one case of sulfasalazine induced MAS associated with has been reported [2] and this is the first case of sulfasalazine-induced DRESS complicated by hemophagocytic lymphohistiocytosis .

Conclusion: The most offending drugs incriminated in the developpement of DRESS complicated with MAS are Vancomycine, phenobarbitol, lamotrigine. Clinicians should be aware of this risk.

References:

• 1. Korbi M, Youssef M, Ben Brahim H, et al. DRESS à l’allopurinol compliqué d’un syndrome d’activation macrophagique Ann Dermatol Venereol 2015;142:767–70.

• 2. Klotz U, Maier K, Fischer C, Heinkel K. Therapeutic efficacy of sulfasalazine and its metabolites in patients with ulcerative colitis and Crohn’s disease. N Engl J Med 1980;303:1499–502

Disclosure of Interest: None declared.

166 ISoP18-1234 Immediate Hypersensitivity Reaction to Acetylsalicylic Acid with Positive Skin Tests

166.1 B. Ouni¹, N. Fathallah¹, S. Larif¹, A. Brahim², R. Slim¹, C. Ben Salem^*1

166.1.1 ¹Pharmacovigilance Center of Sousse, Research Laboratory,Tunisia, Sousse University, Sousse, Tunisia; ²Post-Operatory Care Unit, Farhat Hached University Hospital, Sousse, Tunisia

Background/Introduction: Immediate hypersensitivity reactions to acetylsalicylic acid are classified as allergic reaction (immunoallergic) and pseudo-allergic reaction (pharmacological).

Objective/Aim: To report a case of immediate hypersensitivity reaction following oral acetylsalicylic acid administration with positive rechallenge with positive prick skin test.

Methods: A 55-year-old female patient with a medical history of cervical arthritis treated with Sertraline. In 2013, and after taking one tablet of Acetylsalicylic Acid, 2 h later, she developed a generalized urticaria, dyspnea and wheezing. Her hemodynamic state was stable treated in emergency with corticosteroids and oxygen. In 2014 and after taking one dose of acetylsalicylic acid, 1 h later, the patient developed the same symptomatology requiring her hospitalisation in the emergency department and she received corticosteroids with oxygen. The evolution was favorable. Skin prick tests performed (after obtaining informed consent from the patient) returned positive for acetylsalicylic acid and negative for Ketoprofen and Piroxicam. For that reason, we contre-indicated the use acetylsalicylic acid.

Results: Immediate hypersensitivity reactions to NSAIDs arise from 2 pathophysiological mechanisms: immunological or pharmacological. Clinical pattern included urticaria, angioedema and possibly anaphylactic shock. In our case, the association between the chronological and clinical arguments and positive skin prick tests (SPT) results suggests an Ig-E mediated mechanism and allowed us to retain the responsibility for acetylsalicylic acid.  SPT have been successfully applied in the assessment of IgE-mediated immunological drug reactions [1].

Conclusion: In immediate hypersensitivity reactions induced by drugs, it is recommend to perform skin tests in patients with adverse reactions to these drugs before exposing them to an oral challenge.

References:

• 1. Rebelo Gomes E, Geraldes L, Gaspar Â, Malheiro D, Cadinha S, Abreu C, et al. Hypersensitivity reactions to nonsteroidal anti-inflammatory drugs among adults: clinical features and risk factors for diagnosis confirmation. Int Arch Allergy Immunol 2016;171:269–275

Disclosure of Interest: None declared.

167 ISoP18-1237 Allopurinol-Induced Overlapping Dress and Toxic Epidermal Necrolysis

167.1 N. Fathallah¹, S. Mokni², B. Ouni¹, I. Ben Saida³, S. Larif¹, N. Ghariani², R. Slim¹, C. Ben Salem^*1

167.1.1 ¹Pharmacovigilance Center of Sousse, Research Labratory, Sousse University, Sousse, Tunisia; ²Department of Dermatology, ³Intensive care Unit, Farhat Hached University Center, Sousse, Tunisia

Background/Introduction: Drug-induced cutaneous adverse reactions are varying from 2 to 3% of all hospitalized patients. Severe cutaneous adverse reactions (SCARs) represent 2% of the overall adverse cutaneous reactions. Although the diagnosis criteria between the different SCARs seems to be specific, DRESS and toxic epidermal necrolysis (TEN) can share some features, raising the hypothesis of overlap syndromes [1].

Objective/Aim: To report a case of overlapping DRESS and Toxic epidermal necrolysis induced by allopurinol.

Methods: A 35-year-old male, diagnosed in 2002 with chronic renal failure, was initiated with allopurinol for hyperuricemia. Four weeks later, he developed a generalized erythematous drug eruption associated with fever and cervical lymphadenopathy. The mucosae were not involved and no skin detachment was noted. Biological tests showed creatinine levels of 175 µl likely due to his previousely known chronic renal failure and hyperuricaemia. Hemogram showed hypereosinophilia of 1600/mm³. Screening for viral serology was negative. Allopurinol-induced DRESS was probable according to the Regiscar. Allopurinol was stopped and the patient received supportive local treatment. The evolution was marked by a general spread of the eruption to the whole body associated with skin detachment and positive Nikolsky sign, associated with mucosal involvement.

The histopathological findings revealed a necrotic epidermis with spongiosis, necrotic keratinocytes, and rare eosinophils, compatible with SJS. In our patient, the clinical picture fulfilled “probable” score for both DRESS and TEN according to the RegiSCAR criteria. In our patient, our final diagnosis was an overlapping DRESS/TEN.

Results: In some cases, the initial presentation of SCARs can be confusing making the diagnosis more challenging. In fact, although the skin lesions in SJS/TEN are different from those in DRESS, there is some confusion regarding the diagnosis of DRESS syndrome vs TEN. The pathogenesis of SCARs is not fully understood. SCARs are a Type IV reactions mediated by T cells. In the recent classification of Gell and Coombs, distinct T-cell functions led to different clinical phenotypes. T cells recruit and activate different blood cells such as monocytes, eosinophils or neutrophils. For instance, granulysin is the dominant cytokine inducing the destruction of epidermis in SJS-TEN and eotaxine and IL5 in DRESS. Therefore, delayed adverse drug hypersensitivity reaction may be primary based on the preferential activation of drug-specific T cells. However, these complexe immune reactions are not specific and many reactions may be associated. So, an overlap of immune reactions is possible, even if one type of drug-activated specific cells is dominant, explaining clinical ambiguities among SCARs [2].

Conclusion: Clinician should be awer of the risk of overlapping DRESS and Toxic epidermal necrolysis induced by allopurinol.

References:

• 1. Casagranda A, Suppa M, Dehavay F, Del Marmol V. Overlapping DRESS and Stevens-Johnson Syndrome: case report and review of the lLiterature. Case Rep Dermatol 2017;9:1–7

• 2. Horcajada-Reales C, Pulido-Pérez A, Suárez-Fernández R.Severe cutaneous drug reactions: do overlapping forms exist? Actas Dermosifiliogr 2016;107:23–33

Disclosure of Interest: None declared.

168 ISoP18-1239 A Pediatric Case of Severe Rhabdomyolysis in Acute Asthma Crisis

168.1 C. Ladhari¹, N. Fathallah¹, B. Ouni¹, A. Tej², S. Larif¹, J. Bouguila², R. Slim¹, C. Ben Salem^*1

168.1.1 ¹Pharmacovigilance Center of Sousse, Research Labratory, Sousse University, Sousse, Tunisia; ²Department of Pediatry, Farhat Hached University Hospital, Sousse, Tunisia

Background/Introduction: Corticosteroids and β-Agonists are the first line of treatment for severe acute asthma. Adverse drug reactions related to the use of these medications include hyperglycemia, hypertension and nausea. Rhabdomyolysis associated with severe acute asthma is an uncommon complication not well known by clinicians.

Objective/Aim: To present a case of severe rhabdomyolysis occurring after severe acute asthma in an 11 year-old infant.

Methods: An 11-year-old boy with history of asthma controlled with short-acting β2-agonist upon request, had developed coughing and dyspnea after dust’s exposure. Respiratory function had worsened and he was transferred to the pediatric department. The diagnosis of severe acute asthma was retained and the infant was mechanically intubated and ventilated. He received salbutamol 0.25 mg/h, solumedrol 80 mg intravenous then 40 mg intravenous every 6 h. Respiratory signs improved gradually. However, at the biological test, an increased CPK at 1626 UI/ml was noticed. Otherwise, his renal and liver function were in normal ranges. Potassium level was of 5.74 mmol/L and sodium level was of 137 mmol/L. Blood gas measurements showed respiratory acidosis. Rhabdomyolysis secondary to high doses of corticosteroids was suspected. A decrease in doses of solumedrol was recommended. The monitoring of CPK values revealed a decrease in CPK values (1478 then 800). CPK levels returned to normal ranges 1 week later.

Results: This infant had developed a rhabdomyolysis in the course of severe acute asthma. The association of rhabdomyolysis and asthma has been described at the first time by Chud et al. [1] however; the problem consists to determine the cause of rhabdomyolysis occurring in such cases. In the literature, some hypothesis suggest that theophylline intoxication, hypokalemia secondary to excessive use of β2-agonists or infection with legionella pneumophila and Mycolpasma pneumonia are likely to induce rhabdomyolysis.  Yet, the most common cause seems to be hypoxia, acidosis and intensive muscular work of respiratory muscles [2]. In our case, rhabdomyolysis is most likely to be induced by high doses of corticosteroids since decreasing of CPK values when clinicians decreased doses of corticoids given to this infant. Moreover, his potassium blood rate was within normal range. High doses of corticosteroids may induce myopathy which is related to rhabdomyolysis [2].

Conclusion: Close monitoring of CPK in severe acute asthma remains necessary due to the risk of occurrence of rhabdomyolysis. Rapid management of rhabdomyolysis is primordial to aboid severe complications.

References:

• 1. Chugh KS, Singhal PC, Khatri GK. Rhabdomyolysis and renal failure following status asthmaticus. Chest 1978;73:879–80

• 2. Goh AY, Chan PW. Acute myopathy after status asthmaticus: steroids, myorelaxants or carbon dioxide? Respirology 1999;4:97–9

Disclosure of Interest: None declared.

169 ISoP18-1241 Side Effects of Targeted Therapies

169.1 N. Fathallah¹, C. Ladhari¹, B. Ouni¹, S. Larif¹, K. Baccouche², R. Slim¹, C. Ben Salem^*1

169.1.1 ¹Pharmacovigilance Center of Sousse, Research Labratory, Sousse University, Sousse, Tunisia; ²Department of Rheumatology, Farhat Hached University Hospital, Sousse, Tunisia

Background/Introduction: Targeted therapies have nowadays significantly changed the treatment of cancers and inflammatory diseases. Targeted therapies are generally better tolerated than traditional chemotherapy, but they are associated with several adverse effects.

Objective/Aim: To report the adverse effects reported with targeted therapies.

Methods: This retrospective study was carried out on reports of adverse reactions related to targeted therapies. Reports were notified to the Regional Centre of Pharmacovigilance of Sousse from the 1st January 2015 to 31th December 2016. The causality assessment was evaluated by Naranjo’s method of imputability.

Results: This case series included 8 notifications of adverse reaction imputed to targeted therapies. The sex ratio H/F was 1.66 and the median age was 40.6 years (from 17 to 69 years). The culprit drugs were as following: etanercept (N = 1), certolizumab (N = 1), sorafenib (N = 1), natalizumab (N = 1), imatinib (N = 2), Tocilizumab (N =), infliximab (N = 1). Targeted therapies were Indicated to treat cancers in 3 cases [hepatocellular carcinoma (N = 1) and myeloid leukemia chronic (N = 2)]. Inflammatory diseases were as following: rheumatoid arthritis (N = 2), juvenile arthritis (N = 2) and Multiple Sclerosis (N = 1). Side effects noticed were hypersensitivity reactions (anaphylactic reaction), cutaneous lesions, leuconeutropenia, and muscular involvement. The delay of onset was variable from few minutes after administration to 4 months. The outcome was favorable in all cases.

Conclusion: Targeted therapies act at specific sites and are monoclonal antibodies (in our case series: infilixamb, natalizumab, cetrolizumab and tocilizumab) and fusion proteins (etanercept). Receptor tyrosine kinase inhibitors (imatinib and sorafenib) could be included as targeted therapies as they are biological drugs. Concerning adverse reactions, the true incidence is unknown. However, the Spanich Society of Rheumatology Biobadaser database recorded a total of 16361 adverse reactions in 6754 patients treated with monoclonal antibodies until 2012 [1]. Cutaneous adverse reactions are the most reported adverse reactions. Our patients had severe anaphylactic reactions, prurigo and fixed drug eruption. However, in the literature, other cutaneous side effects are reported including: hand-foot syndrome, acneiform eruption, psoriasiform eruption, cutaneous infection, vasculitis, cutanous carcinoma… [2]. Hematologic disorders (leuconeutropenia in our case series) cause recurrent infections and lead to drug withdrawal. Muscular disorders are also reported with target therapies.

Adverse effects to targeted therapies may be severe and may lead to drug cessation. Some measures may be proposed to avoid occurrence of side effect such as hydration and calcium and magnesium supplementation.

References

• 1. Biobadaser. Regitro espanol de acontecimientos adversos de terapia biologica en enfermedades reumaticas. Informe enero 2013

• 2. M. Helloa, S. Barbarotb, J. Connaulta. Skin manifestations of new targeted treatments. Rev Med Interne. 2012;33:273–8

• 3. A. Leroux, N. Clere. Prévention et prise en charge des effets indésirables induits par les thérapies ciblées. Actualités Pharmaceutiques, 2015

Disclosure of Interest: None declared.

170 ISoP18-1242 Skin Testing in Antibiotic Allergic Patients with Immediate and Delayed Hypersensitivity Reactions

170.1 N. Fathallah¹, C. Ladhari¹, S. Mokni², B. Ouni¹, S. Larif¹, K. Baccouche³, N. Ghariani², R. Slim¹, C. Ben Salem^*1

170.1.1 ¹Pharmacovigilance Center of Sousse, Research Labratory, Sousse University, Sousse, Tunisia; ²Dermatology Department, ³Department of Rheumatology, Farhat Hached University Hospital, Sousse, Tunisia

Background/Introduction: Skin tests are performed to investigate cutaneous adverse drug reactions (CADR) in both immediate and delayed hypersensitivity reactions [1]. Sensitivity of skin tests seems to vary according to the tested drugs and the type of CARD.

Objective/Aim: To report the results of skin tests in antibiotic allergic patients with immediate and delayed hypersensitivity reactions.

Methods: A retrospective study was carried out on patients who developed a CARD secondary to antibiotics. These patients were investigated by skin tests in the Regional Pharmacovigilance Centre of Sousse, Tunisia from January 2009 to February 2016. Skin tests including prick tests and intradermoreactions with immediate readings for immediate hypersensitivity reactions and patch tests for delayed hypersensitivity reactions. The results of the tests were performed according to the European Network for Drug Allergy recommendations.

Results: Fifty-eight skin tests to antibiotics were performed in our case series. Previous history of drug reaction was found in 17 patients (29%). CADR were immediate hypersensitivity reactions (urticaria and anaphylactic choc) in 34 cases and were delayed hypersensitivity reactions in 24 patients (including maculopapular eruption, fixed drug eruption and acute generalized exanthematous pustulosis). Tested drugs were amoxicillin (22 cases, 13 Prick tests, 9 patch tests), penicillin G (7 Prick tests), oxacillin (7cases, 4 Prick tests, 3 patch tests), amoxicillin/calvulanate (1 patch test), ampicillin (1 Prick test), cefotaxime (3 cases, 2 Prick tests, 1 patch test), Cefazolin (3 Prick tests), Ceftriaxone (2 patch tests), Cefuroxime (1 patch test), lincomycin (2 patch tests), cotrimoxazole (2 cases, 1 Prick test, 1 patch test), ciprofloxacin (1 Prick test), rovamycine (1 Prick test) and 1 patch test to clarithromycin, vancomycin, pristinamycin, fusidic acid and vibramycin, respectively. Performed skin tests were positive to amoxicillin in 53%, to penicillin G in 43%, to oxacillin in 57.1% and to cefazolin in 33.3% of cases. Skin tests performed toward ciprofloxacin and rovamycin were also positive. Four patients experienced cross reactivity among betalactamins. Indeed, cross reactivity was noticed among amoxicillin and oxacilline in 2 cases, penicillin G and oxacilline in one case, amoxicillin and penicillin G in one case, amoxicillin, oxacillline and ampicillin in one case.

Conclusion: The most frequent tested drugs in our case series were betalactams. Our results confirm those reported in many studies. In fact, hypersensitivity to betalactams is the most reported drug allergy and positivity of skin tests in this class of antibiotics is also most frequent [3]. Drug hypersensitivity investigation in betalactam hypersensitivity is actually well established. Concerning the others drugs, there is no clear strategy to explore CADR [1]. False positive test could be seen toward quinolones due to their non specific histaminoliberation. Skin tests seems to be a safe and rapid method that should be carried out as frequently as possible to establish drug allergy especially among betalactamin antibiotics.

References:

• 1. Barbaud A. Skin testing in delayed reactions to drugs. Immunol Allergy Clin North Am 2009;29:517–35

• 2. Bursztejn AC, Rat AC, Tréchot P, Cuny JF, Schmutz JL, Barbaud A. Results of skin tests to assess drug-induced allergy. Ann Dermatol Venereol 2010;137:688–94

• 3. Matheu V, Pérez E, González R, Poza P, De La Torre F, Sánchez-Machín I, et al. Assessment of a new brand of determinants for skin testing in a large group of patients with suspected beta-lactam allergy. J Investig Allergol Clin Immunol 2007;17:257-60

Disclosure of Interest: None declared.

171 ISoP18-1243 Lichenoid Eruption Induced by Methotrexate

171.1 R. Slim¹, N. Fathallah¹, B. Ouni¹, S. Larif¹, S. Mokni², C. Belajouza², C. Ben Salem^*1

171.1.1 ¹Pharmacovigilance Center of Sousse, Research Labratory, Sousse University, Sousse, Tunisia; ²Dermatology Department, Farhat Hached University Hospital, Sousse, Tunisia

Background/Introduction: Methotrexate, a folate antagonist, has been used in dermatology to treat psoriasis and collagen vascular diseases. Cutaneous adverse reactions attributed to methotrexate therapy are uncommon.

Objective/Aim: Herein, we report a rare case of lichenoid eruption associated with methotrexate.

Methods: A 40-year-old man presented to dermatology department with 2 month history of a widespread pruritic eruption. He had a history of psoriasis treated with methotrexate therapy (10 mg/week) started 8 months before onset of the eruption. Physical examination revealed lichenoid plaques on the abdomen, anterior side of forearms and thighs. The oral and genital mucosae were not involved.

Based on the clinical and histologic findings, a diagnosis of lichenoid drug eruptions (LDE) induced by methotrexate was suspected and the medication was withdrawn. A significant improvement was noted after 2 weeks of treatment with prednisone and potent topical steroid. There was no recurrence of the lichenoid eruption during a 9-month follow-up period.

Results: LDE are uncommon dermatoses that can occur after administration of various medications and chemicals [1]. The diagnosis is often challenging and it can be difficult to distinguish these eruptions from idiopathic lichen planus because both clinical and histological similarities. Furthermore, several findings are more typical of specimens of LDE and include focal parakeratosis, presence of eosinophils and plasma cells, and a deeper perivascular and periadnexal infiltrate [2].

Our patient had most of clinical and histopathological features suggestive of LDE. In addition, the diagnosis of a LDE is favored over idiopathic lichen planus because of the closely linked onset of occurrence of clinical symptoms and the complete resolution after withdrawal of methotrexate. 

Several mucocutaneous adverse reactions have been reported with methotrexate in the treatment of psoriasis, including oral ulceration, erythroderma, and bullous eruption [3].

To the best of our knowledge, this is the first lichenoid drug eruption due to intake of methotrexate following treatment of psoriasis.

The latency period for starting medication and occurrence of LDE ranged from 10 days to several years. LDE normally cleared spontaneously within a few weeks to a few months after withdrawal of the causative drug [1].

Conclusion: As the frequent use of methotrexate in dermatology, clinicians should be aware of the possibility of LDE occurring as a rare adverse effect.

References:

• 1. Ellgehausen P, Elsner P, Burg G. Drug-induced lichen planus. Clin Dermatol 1998;16:325–32

• 2. Van den Haute V, Antoine JL, Lachapelle JM. Histopathological discriminant criteria between lichenoid drug eruption and idiopathic lichen planus: retrospective study on selected samples. Dermatologica 1989;179:10–3

• 3. Sako EY, Famenini S, Wu JJ. Bullous drug eruption in a patient with psoriasis after a test dose of methotrexate. J Am Acad Dermatol 2013;69:e264–5

Disclosure of Interest: None declared.

172 ISoP18-1244 Sulfasalazine Induced Dress Syndrome with Fulminant Hepatitis

172.1 R. Slim¹, N. Fathallah¹, B. Ouni¹, S. Larif¹, H. Hmouda², C. Ben Salem^*1

172.1.1 ¹Pharmacovigilance Center of Sousse, Research Labratory, Sousse University, Sousse, Tunisia; ²Intensive Care Unit, Farhat Hached University Hospital, Sousse, Tunisia

Background/Introduction: Drug reaction with eosinophilia and systemic symptoms (DRESS) is a severe and potentially life-threatening disease that can lead to acute liver failure.

Objective/Aim: We report a rare case of favourable issue of a fulminant hepatitis associated with sulfasalazine DRESS syndrome.

Methods: A 35-year-old man had been taking sulfasalazine for 3 weeks when he experienced symptoms consistent with a viral infection.

He was admitted to hospital on day 6. Physical examination found fever, widespread erythematous macules and cervical lymphadenopathy. Investigations revealed hypereosinophilia and hepatic cytolysis. Abdominal ultrasonography was normal. Viral and autoimmune antibodies were not detected. A skin biopsy was compatible with drug-induced toxicity. Sulfasalazine was stopped, and methylprednisolone was initiated. The immediate evolution was characterized by major hepatic cytolysis (aminotransaminases over 2498 IU/l), and cholestasis (total bilirubin at 9 N). Signs of liver failure occurred, with factor V at 37% and prothrombin at 27%. Consequently, the patient was proposed for emergency liver transplantation. However, aminotransferases rapidly decreased as well as the factor V and prothrombin index were assessed. The decision regarding liver transplantation was therefore postponed. Symptomatic intensive care and corticosteroid therapy were maintained. The patient was discharged on day 25. His liver function tests returned to normal levels in 2 month.

Results: Liver involvement in DRESS is common and may range from a transitory increase in liver enzymes to liver necrosis with fulminant hepatic failure [1].

Sulfasalazine, is a sulfamid currently used for treating inflammatory bowel diseases. Side effects include gastrointestinal symptoms, headaches, and hemolytic anemia. Immunoallergic reactions to sulfasalazine, including DRESS syndromes, have been reported and they describe cases of mild hepatitis, but acute liver failure resulting in death or liver transplantation can also occur [2–3]. Classical management requires immediate withdrawal of the suspected drug. Liver transplantation is usually proposed in emergency situations, but the prognosis remains poor. According the Naranjo probability scale sulfasalazine-induced DRESS syndrome with fulminant hepatitis was probable in our patient [4].

Conclusion: Even though sulfasalazine DRESS is rare, clinicians increase in awareness enables early recognition and treatment so as to reduce morbidity in these patients.

References:

• 1. Husain Z, Reddy BY, Schwartz RA. DRESS syndrome: Part I. Clinical perspectives. J Am Acad Dermatol 2013;68:693.e1–14; quiz 706–8

• 2. Michel F, Navellou JC, Ferraud D, Toussirot E, Wendling D. DRESS syndrome in a patient on sulfasalazine for rheumatoid arthritis. Joint Bone Spine 2005;72:82–5

• 3. Besnard M, Debray D, Durand P, Cezard JP, Navarro J. Sulfasalazine-induced fulminant hepatitis in pediatric Crohn’s disease: report of two cases. J Pedriatr Gastroenterol Nutr 1998;26:119–20

• 4. Naranjo CA, Busto U, Sellers EM, Sandor P, Ruiz I, Roberts EA, et al. A method for estimating the probability of adverse drug reactions. Clin Pharmacol Ther 1981;30:239–45

Disclosure of Interest: None declared.

173 ISoP18-1246 Signals Detection in Pharmacovigilance Based on the Seriousness of Adverse Drug Reactions, Moroccan Pharmacovigilance Database, 2008–2017

173.1 I. Talibi¹, A. Tebaa², A. Khattabi¹, R. Soulaymani-Bencheikh^*2

173.1.1 ¹Ecole Nationale de Santé Publique; ²Centre Anti Poison et de Pharmacovigilance du Maroc, Rabat, Morocco

Background/Introduction: Adverse drug reactions (ADRs) constitute a significant burden for public health. They represent major concerns in terms of individual outcomes and public health expenditures [1–4]. The risk factors for developing serious ADRs are not yet studied in Morocco. We propose a new approach for signal detection in pharmacovigilance based on the measure of the disproportionality for developing the serious ADRs cases.

Objective/Aim: To identify risk factors associated to serious ADRs and to detect pharmacovigilance signals in Moroccan Pharmacovigilance Database (MPD).

Methods: We carried out a retrospective observational study of ADRs in MPD for the period from January 2008 to December 2017. We included in this study only spontaneous cases with the scores of WHO Causality Assessment certain, probable or possible, and excluded the vaccine cases. We used the Anatomical Therapeutic Chemical classification for drugs, WHO Adverse Drug Reaction Terminology and the ICH E2A guidelines to classify the seriousness of cases. Descriptive statistics and logistic regression using Epi Info 7 were undertaken.

Results: Among all the reported ADRs included in our study (n = 9283), 39.3% referred to serious ADRs (n = 3648). The majority of these serious ADRs (73.1%) were caused by drugs belonging to J (32.8%), L (21.3%) and N (19.0%) ATC groups. Cases reported by health professionals were more associated with seriousness than those reported by patients [Odds Ratio = 6.53, 95% CI (5.32–8.02)]. Patient demographics were explored as well as disproportionate reporting signals.

Conclusion: This study has identified factors that associated to developing serious ADRs in Morocco and proposed a new approach to detect signals in pharmacovigilance. We recommend using this data to develop of risk management plans, to generate alert and to implement the risk minimization actions for improving the safety of drug use in Morocco.

References:

• 1. Lazarou J, et al. Incidence of adverse drug reactions in hospitalized patients: a meta-analysis of prospective studies. JAMA 1998; 279:1200–5

• 2. Kongkaew C, et al. Hospital admissions associated with adverse drug reactions: a systematic review of prospective observational studies. Ann Pharmacother 2008;42:1017–25

• 3. van der Hooft CS, et al. Adverse drug reaction-related hospitalisations: a population-based cohort study. Pharmacoepidemiol Drug Saf. 2008;17(4):365–71

• 4. Moore N, et al. Frequency and cost of serious adverse drug reactions in a department of general medicine. Br J Clin Pharmacol 1998;45:301–8

Disclosure of Interest: None declared.

174 ISoP18-1250 Contribution Of Patients To Pharmacovigilance—The Views Of European Patient Organizations

174.1 C. Matos^*1,2, G. Weits³, F. van Hunsel³

174.1.1 ¹Department of Pharmacology, Faculty of Pharmacy, University of Seville, Seville, Spain, ²Department of Pharmacy, Coimbra Health School-IPC, Coimbra, Portugal, ³Netherlands Pharmacovigilance Centre Lareb, ‘s Hertogenbosch, Netherlands

Background/Introduction: Patients’ involvement in actively reporting adverse drug reactions (ADRs) is a key to build better drug safety practices, and growing pharmacovigilance initiatives from consumer organizations are being recognized as potentially valuable sources of information.

Objective/Aim: The objective of this study is to understand the role of European patient organizations as stakeholders to optimize patient involvement in pharmacovigilance.

Methods: A descriptive-correlational study was conducted investigating patient organizations’ opinions and attitudes regarding patient involvement in pharmacovigilance and their initiatives to support drug safety through a web-based questionnaire during March and April 2018. 

Results: A total of 1898 patient organizations were invited to participate in the survey, including 89 pan-European organizations. In total, 297 completed answers (15.65%) were collected from 31 countries. Most organizations stated that would like to increase awareness of patients to specific ADRs related to their medicines (43.19%, n = 130). However, 38.54% declare don’t have any goal in Pharmacovigilance (n = 116). Barriers found to the support in Pharmacovigilance activities include lack of resources to be involved in pharmacovigilance activities (43.77%, n = 130) or don’t receive any support from National Competent Authorities in their countries (33.33%, n = 99).

Table 1. Contribution of the patients to pharmacovigilance

In what way do you think patients can contribute to pharmacovigilance? Total of respondents: 301	Strongly Disagree or Disagree	Neutral	Agree or Strongly Agree
Patients can provide different information from the healthcare professionals	5.64% (17)	14.95% (45)	76.08% (229)
Patients can contribute to the detection of new adverse drug reactions	3.32% (10)	5.32% (16)	89.70% (270)
Patients give more information on the impact of the adverse drug reaction on quality of life	2.66% (8)	6.31% (19)	88.04% (265)
Patients can be useful in describing the severity of reported reactions	4.65% (14)	7.31% (22)	86.38% (260)
Patients can report information that is useful, even without medical confirmation	6.64% (20)	12.29% (37)	79.40% (239)
Patients describe information based on their experience with medicines	3.32% (10)	5.98% (18)	89.37% (269)

1. “Not Answered”/“Not Relevant” answers were not present in the table

Organizations inform patients that they are allowed to report adverse drug reactions (40.40%; n = 120) or communicate them if an important new ADR appears for a specific medicine (40.07%; n = 119), or when a new drug is marketed (30.98%; n = 92); however, more than one-third admitted that never had any involvement in Pharmacovigilance (34.68%; n = 103). Activities related with pharmacovigilance include spread information on how to report (for instance on their websites, newsletter, email, conferences) (28.57%, n = 86), organizing activities (seminars, campaigns, workshops, courses) to help patients to be aware of drug safety (28.57%, n = 86), or have a direct link on patient organization website for patients to report adverse drug reactions (6.64%, n = 20).

Conclusion: Bringing pharmacovigilance stakeholders and patient organizations together could create awareness among patients. However, greater communication by patient organizations on pharmacovigilance topics and facilitation of the reporting is necessary to enable their members to be more active in pharmacovigilance.

Disclosure of Interest: None declared.

175 ISoP18-1252 Present Status and Future Prospects of Pharmacovigilance in Nepal

175.1 R. Karki^*1

175.1.1 ¹Pharmacy, Pokhara University, Pokhara, Nepal

Background/Introduction: Nepal, one of the low income country had started pharmcovigilance program from the year 2004 and become full member of the International Drug Monitoring Programme since 2006. The drug regulatory body of the country, Department of Drug Administration (DDA) is the national pharmacovigilance centre that coordinates 11 regional centres situated in several hospitals across the country.

Though ADR reporting is mandatory in Sweden, France and Italy [1, 4], it is not in Nepal. The current trend suggests high rate of underreporting [2] of ADR in Nepal which might be due to lack of awareness and educational intervention among healthcare professionals [3].

The present study is an attempt to present the current status and future prospects to improve the ADR reporting programme in Nepal.

Objective/Aim: The primary objective was to find out the current status of pharmacovigilance programme since its beginning, to assess the knowledge and perception of final year pharmacy students regarding ADR reporting and pharmacovigilance in Nepal.

Methods: A retrospective analysis of suspected ADR reports from the National Pharmacovigilance Centre, online website was conducted. Similarly, a cross sectional study using pretested questionnaire (cronbach’s alpha:0.77) was performed using census sampling technique.

A total of 27 final year pharmacy students were enrolled in the study. The data collection was done after getting approval from the institution and informed consent from the participants. SPSS version 16 was used for data entry and analysis.

Results: Since 2004, a total of 547 ADRs were reported to the Uppsala Monitoring Center, Sweden from Nepal. The top 10 drugs reported causing ADR were Carbamazepine, Phenytoin, Ciprofloxacin, Amoxicillin, Diclofenac, Isoniazid, Ibuorofen, Paracetamol,Tramadol and Cotrimoxazole. The maximum number of ADR reported in the year 2008 (209) which then reduced gradually and become only 2 suspected ADR reports in the year 2017.Out of 27 participants majority were female (74.9%). The knowledge assessment have showed all participants were known about ADR classification, Post Marketing Surveillance and Hypersensitivity reactions. Only 37% were known about how to report ADR in Nepal. Majority (81.5%) of participants reported pharmacovigilance as not well covered topic in their curriculum.The perception towards pharmacovigilance have showed that 77.8% participants strongly agreed to make ADR reporting compulsory and 81% strongly agreed about teaching on reporting of ADR to pharmacy students. They strongly agreed on counselling about ADR while dispensing (92.6%). However student were not aware about the purpose of spontaneous ADR reporting.

Conclusion: Pharmacovigilance is still in infancy in case of Nepal despite of 14 years of its establishment in Nepal. This study highlights the need of inclusion of pharmacovigilance in the curriculum of pharmacy students as well as pre service training about ADR reporting especially in those places which lack regional pharmacovigilance centre.

References:

• 1. Avong Y. Prospects and Challenges of providing pharmacovigilance services in resource limited countries Pharmaceut Reg Affairs 2015;4:53

• 2. Santosh KC, Tragulpiankit P, Gorsanam P, Edwards IR, Alam K. Strengthening the pharmacovigilance programme in Nepal. Nepal J Epidemiol 2013;3:230–5

• 3. Palaian S, Ibrahim MI, Mishra P. Health professionals knowledge, attitude and practice towards pharmacovigilance in Nepal. Pharma Practa (Granada) 2011;9:228–35

• 4. Olsson S, Pal SN, Dodoo A. Pharmacovigilance in resource limited countries. Expert Rev Clin Pharmacol 2015;8:449–69

Disclosure of Interest: None declared.

176 ISoP18-1254 Adverse Drug Reaction Reports Following a “Brand Switch”. The New Zealand experience

176.1 R. Braund^*1, M. Tatley¹

176.1.1 ¹NZPhvC, University of Otago, Dunedin, New Zealand

Background/Introduction: Internationally the use of generic medicines is increasing. There is growing interest in the “patient experience” of using generic medicines, particularly in relation to risk of adverse drug reactions (ADRs) and therapeutic equivalence. Recently this has focused on ADRs experienced by a patient when switching from the innovator brand of a specific medicine to its generic “equivalent” [1, 2]. Questions have been raised about the impact of generic medicines on the public and health providers and whether this may be increasing the report frequency [3]. Subsidised medicines are available to New Zealanders procured through a tender process which can result in brand changes, typically to generics over time in long-term users. This provides a unique opportunity to evaluate the ADR reporting profile following a “brand switch”.

Objective/Aim: To use a series of “brand switch” examples as case studies to identify the ADR reporting profiles and characteristics.

Methods: Specific “brand switch” examples were identified that provided an opportunity to compare the reporting profiles. These provided an opportunity to determine the potential influence of parameters such as brand-attributed ADRs, user-perceived loss of therapeutic effect, formulation differences or factors such as social media influence of user perceptions. The time course of reports subsequent to the change was determined, the number of reports and type of reactions reported were investigated.

Results: There is a delay following new brand availability before reactions begin to be reported, most likely due to the delay in exhaustion of the previous brand supply. Reports of loss of therapeutic effect tend to predominate. A smaller proportion of reports document mild hypersensitivity reactions, including rash and gastrointestinal disturbances. Some reports of loss of therapeutic effect document return of efficacy on dose re-titration which may reflect individual variations in bioavailability. Reports submitted for generics produced by the same innovator, supports the suggestion that there is an influence of the users perceptions of generic medicines.

Conclusion: When a brand switch occurs there is a predictable spectrum of reactions reported. Some of these may reflect perception of generics rather than the medication, However, there are some reactions reported that do reflect individual responses to either altered bioavailability or formulation factors (i.e. excipients).

References:

• 1. Desari RJ et al. Differences in rates of switchbacks after switching from branded to authorized generic and branded to generic drug product: cohort study. BMJ 2018;361:k1180

• 2. Singh S. The safety of generic prescription drugs in the United States. Drug Saf 2018;41:325–8

• 3. Faasse K et al. The effect of a apparent change to a branded or generic medication on drug effectiveness and side effects. Psychosom Med 2013;75:90–6

Disclosure of Interest: None declared.

177 ISoP18-1256 Acetylsalicylic Acid Adverse Events: A Survey of General Practitioners in Democratic Republic of Congo

177.1 O. M. Mboma^*1, E. S. Tambwe², W. M. Mampuya³, T. N. M. Mpiempie⁴, F. B. Makila⁵, G. K. Mesia⁶

177.1.1 ¹Clinical Pharmacology and Pharmacovigilance Unit and National Pharmacovigilance Centre, University of Kinshasa, The Democratic Republic of the Congo; ²Basis Science, University of Kinshsa, Kinshasa, The Democratic Republic of the Congo; ³University Hospital Centre, University of Sherbrooke, Sherbrooke, Canada ; ⁴National Pharmacovigilance Centre, National Pharmacovigilance, ⁵Centre Nationale de Pharmacovigilance, Centre National de Pharmacovigilance, ⁶Faculty of Pharmaceutical Sciences, Universty of Kinshasa, Kinshasa, The Democratic Republic of the Congo

Background/Introduction: Acetylsalicylic acid (ASA) remains essential in the management of cardiovascular disease both in primary and secondary prevention. It is one of the key measures to reduce embolic strokes incidence [1], the major cause of disability all over the world [2]. According to the World Health Organization (WHO), in the World, cardiovascular diseases (CVD) are responsible for 17.3 million deaths per year, 9% in the Democratic Republic of the Congo (DRC) [3].

Objective/Aim: Identify the various adverse events (AE) of ASA as reported by the prescribing general practitioners (GPs).

Methods: We administered a questionnaire to GPs who had prescribed ASA at least once from October to December 2017, for which patients had complained of AEs.

Results: Questionnaires completed by 280 GPs were analyzed. In total, 66 AEs were reported following the use of ASA, of which 64 were benign and 2 were severe. The severe cases included thrombocytopenia and disseminated intravascular coagulation (DIC). Thirty-three percent of these AEs led to treatment modification. The most commonly reported AEs were edemas (12.1%), gastrointestinal disorders (33.3%), bleeding (42.4%) and allergic reactions (4.6%). Edemas included facial edema (12.5%), lower extremity edema (25%) and unspecified edemas (62.5%). Gastrointestinal disorders included vomiting (4.5%), epigastralgia (9.1%), abdominal pain (4.5%), diarrhea (13.7%), dyspepsia (4.5%), acute gastritis (45.5%) and unspecified gastrointestinal disorders (18.2%). Bleeding (42.4%) included hematemesis (14.3%), gingivorrhagia (10.7%), epistaxis (7.1%), gastrointestinal hemorrhage (7.1%), thrombopenia (3.6%), bleeding at the injection site (3.6%), petechiae (3.6%), hemoptysis (3.6%), metrorrhagia (3.6%) and rectorrhagia (3.6%), unspecified bleeding (32.1%) and unspecified hemorrhages (7.1%). Anemia (4.6%) included unspecified anemia (66.7%) and DIC (33.3%).  Allergic reactions included pruritus (50%) and unspecified allergic reactions (50%).

Conclusion: This study shows that the use of ASA in the DRC is associated with multiple AEs that lead to a change in therapy in 33% of the cases. Bleedings and gastrointestinal disorders are the most prevalent AEs of ASA. Deepened studies on the rational use of ASA are ongoing.

Key words: Acetylsalicylic acid, Aspirin, Adverse events, DR Congo.

References

• 1. Patrono C, et al. Eur Heart J 2011; 32:2922–32

• 2. The Lancet, Volume 385, Issue 9963, 117–171

• 3. WHO 2013

Disclosure of Interest: None declared.

178 ISoP18-1261 Development, Validation and Implementation of an Active Electronic Pharmacovigilance System in Hospitalized Patients

178.1 A. Ceschi¹, P. Hitz¹, L. Müller¹, R. Bertoli^*1, V. Piffaretti²

178.1.1 ¹Institute of Pharmacological Sciences of Southern Switzerland, Division of Clinical Pharmacology and Toxicology, Ente Ospedaliero Cantonale, Lugano, Switzerland; ²Area Information and Communications Technology, Ente Ospedaliero Cantonale, Lugano, Switzerland

Background/Introduction: Spontaneous reporting of adverse drug reactions (ADRs) is the foundation of pharmacovigilance systems worldwide. One of the main limitations is underreporting, which is significant and widespread, and has been estimated to approach a median rate of 94% [1].

Objective/Aim: To develop, validate and subsequently introduce into routine practice a novel, easy to adapt, active electronic pharmacovigilance system in hospitalized patients.

Methods: In close collaboration with the ICT of our hospital network, we developed an electronic system that actively and continuously screens all electronic medical records searching for words and word combinations which we defined as related to ADRs. We subsequently modified the list of words in order to improve the positive predictive value of the system. Chi square and post hoc tests were performed.

Results: With the first word list we obtained 40 and 60% true and false positives, respectively. A subsequent list generated an error message. After adding further words and combinations, we obtained 60 and 40% true and false positives, respectively, and a further revision of the list provided 96.7 and 3.3% true and false positives, respectively. There was a significant improvement of the positive predictive value of the system (p < 0.001) and the third and fourth word lists contributed significantly to this result (p < 0.001 and p = 0.001, respectively). During the subsequent test phase of 51 days the system generated 365 alerts for potential cases. Of these, 283 (77.5%) were true ADRs and 160 (56.5%) were subject to mandatory reporting according to the Swiss law on therapeutic products. Of these reportable ADRs, 140 (87.5%) were serious, 19 (11.9%) medically important and 1 (0.6%) due to medication abuse. Of the 140 serious ADRs, 85 (60.7%) required hospitalization and 55 (39.3%) increased the length of stay. Among the 160 ADRs subject to mandatory reporting, only 14 (8.8%) cases were reported spontaneously to our pharmacovigilance centre. The underreporting rate was therefore 91.2%. There was a significant correlation between seriousness of the ADR and reporting to the centre (p = 0.024).

Conclusion: We developed and tested an easy to adapt active electronic pharmacovigilance system in hospitalized patients, which was able to detect a large number of relevant ADRs which would have gone unnoticed as they were not reported spontaneously to our centre although subject to mandatory reporting. The system was then successfully introduced into routine practice reducing underreporting and therefore contributing to increase medication safety.

References:

• 1. Hazell L, Shakir SA. Under-reporting of adverse drug reactions: a systematic review. Drug Saf 2006;29:385–96

Disclosure of Interest: None declared.

179 ISoP18-1262 Opinion of Costa Rican Pharmacists Regarding a New Adverse Drug Reaction Reporting Platform, 2017

179.1 V. Hall Ramirez^*1

179.1.1 ¹Pharmaceutical Care and Clinical Pharmacy Department, Pharmacy Faculty, University of Costa Rica, San José, Costa Rica

Background/Introduction: The importance of Pharmacovigilance for safe medicines and their safe use has increasingly been recognised during the last few years [1]. As part of the curriculum in the Pharmacy career of the University of Costa Rica, several aspects related to Pharmacovigilance are studied.

Objective/Aim: Describe the results of a survey to community pharmacists in Costa Rica during the second half of 2017, regarding to a new Adverse Drug Reaction reporting platform.

Methods: We performed a descriptive, cross-sectional study.

Application of a data collection instrument to preceptor pharmacists by the Pharmaceutical Care I course students. The data were managed in a group, confidential and anonymous way. The data obtained was analyzed by Microsoft Excel.

Results: Information was collected from 41 community pharmacies. The first question of the instrument was related to the pharmacist’s experience in reporting an Individual Case Safety Report (ICSRS) through the new Regional Online Platform to Report Adverse Drug Reactions (NOTI-FACEDRA) [2], three months after it was release in June, 2017. All the pharmacist (100%) mentioned that they have not reported any ICRS in this period.

Even if they had not used the platform to make a report, they were encouraged to visit the website along with their student, so they can explore NOTI-FACEDRA and after that, give a valuation of the platform.

The pharmacists said that the report is filled quickly (22.53%), the platform is simple and accessible (21.12%) and it is easy to generate statistical information (21.12%), it can bring regional information (12.67%) and allows to make the report in real time (11.27%).

Conclusion: Although the existence of a new website to report online the adverse drug reactions by the health professionals, it is important to make a campaign to ensure not only that the professionals knows how to use it, but also to empower the patient to visit their pharmacist so they can tell all the possible adverse drug reactions they can be experimenting in order to report them and also to manage the reactions by different pharmaceutical interventions.

References:

• 1. Beckmann J, Hagemann U, Bahri P et al. Teaching pharmacovigilance: the WHO-ISoP core elements of a comprehensive modular curriculum. Drug Saf 2014;37:743–59

• 2. Secretaria Ejecutiva de COMISCA. Portal Regional de Notificación en línea de Sospecha de Reacciones Adversas a Medicamentos de uso humano [Internet] [Accessed on June 4th, 2018]. Available in http://www.notificacentroamerica.net/Pages/mapa.aspx#no-back-button

Disclosure of Interest: None declared.

180 ISoP18-1263 Patient Reporting of Adverse Drug Reactions in Romania-Pilot Phase of a Cross-Sectional Survey

180.1 C. Bucsa^*1, I. Cazacu¹, C. Mogosan¹, N. Bulik¹, B. Axente¹, A. Farcas¹

180.1.1 ¹Drug Information Research Center, “Iuliu Hatieganu” University of Medicine and Pharmacy, Cluj-Napoca, Romania

Background/Introduction: Patient reporting of adverse drug reactions (ADRs) in Romania is possible since 2012, once the new European legislation on pharmacovigilance was introduced. However, 4 years later the level of patient reporting is still very low [in 2016 only 26 ADR reports coming from patients were registered to the Romanian National Agency for Medicines and Medical Devices (NAMMD)] [1].

Objective/Aim: To test a questionnaire designed for patients, to investigate their perception and knowledge on ADR reporting.

Methods: A questionnaire was designed in a panel of 5 pharmacists specialized in pharmacovigilance activities. The patients participating to a health campaign on the importance of health monitoring, during 3 days in July 2017 in Cluj-Napoca, were invited to fill-in the questionnaire. Of the 18 total questions, 17 were referring to demography (age, sex, education, occupation), health status (chronic diseases, medicine use, frequency of physician visits, concern for own health), the source of information on medicines, the ability to recognize ADRs and the perception and knowledge on ADR reporting. The last question contained 13 subheadings on the most recent ADR (if the case).

Results: Of the 59 patients who completed the questionnaire, 42 were women and the median age was 57 years [range 22–78].  The great majority of patients (54) knew that every medicine may induce ADRs. For 17 patients the leaflet was the main source of information for identifying ADRs. 20 patients knew about the possibility of reporting ADRs to NAMMD. Nevertheless, of the 24 patients who described an ADR, only 15 reported the ADR, and all to the attending physician. The main reason for reporting ADRs for 10 of the 15 patients was altruism: the importance of knowing about the ADR for other patients/authorities. Four patients filled-in an ADR report form to NAMMD after completing the questionnaire. Following this pilot phase of the study, 7 questions of the questionnaire were reworded to increase clarity. We also added one question to investigate the reason why patients prefer to report ADRs only to the physician and not directly to NAMMD.

Conclusion: Although most of the patients were aware about the risk of ADRs and the possibility to report the ADRs to the NAMMD, they preferred to report to the attending physician only. The study will continue on-line with the questionnaire adapted accordingly.

References:

• 1. Romanian National Agency for Medicines and Medical Devices. Romanian National Agency for Medicines and Medical Devices 2016 activity report [Romanian language only]. Available from https://www.anm.ro/agentie/rapoarte-de-activitate/. Accessed 29 May 2018

Disclosure of Interest: None declared.

181 ISoP18-1264 Adherence to Treatment of Chronic Patients in General Practitioner Office-First Results of a Cross-Sectional Survey

181.1 C. Bucsa^*1, N. Bulik¹, S. Iurian², S. Muresan³, I. Muresan⁴, G. Feher⁵, C. Mogosan¹

181.1.1 ¹Drug Information Research Center, ²Department of Pharmaceutical Technology and Biopharmaceutics, “Iuliu Hatieganu” University of Medicine and Pharmacy, ³Sandra Muresan medical practice, ⁴Ioan Muresan medical practice, ⁵Gabriella Feher medical practice, Cluj-Napoca, Romania

Background/Introduction: Patients’ adherence to chronic treatment plays an important role in preventing disease complications and adverse drug reactions (ADRs). The impact of poor adherence grows as the burden of chronic disease grows worldwide [1].

Objective/Aim: To investigate how patients’ and health system’s characteristics may influence aspects of adherence to treatment in chronic Romanian patients.

Methods: We developed an in-house questionnaire which comprised aspects of patient’s demography and social status (sex, age, marital status, education, occupation, income, medication cost), source of information on medicines, use of medication reminders, choice of pharmacy and discussion with the pharmacist and several questions to investigate the patient’s forgetfulness (4 questions: sometimes, during past month, past 2 weeks and 1 day before the interview) and intentional omission of taking medicines (in case the health status improves or worsens). The questionnaire was applied to chronic patients in 3 general practitioners (GPs) offices by a trained pharmacist before the GP consultation. The questions targeted the month before the interview timeframe.

Results: We present here the results of the first 100 questionnaires. The median age of the patients was 63 years, most of them (66) were females and married (29). The medium income was approximately 257 euro/month and the medium amount spent on medication was 23 euro/month. The medium number of chronic diseases was 2.45 and the medium number of medicines 4.07. The great majority of patients (95) followed only advice of doctor/pharmacist related to their medication.  Most of the patients took their medicines always from the same pharmacy (48), only 6 of them because they preferred a personal pharmacist and 19 because of the convenient location. 66 of the patients did not use a reminder system for taking their medicines. A great part of the patients (45) did not forget to take their medicines during the past month (answered NO to all 4 questions), 34 patients answered YES once, 15-twice and 6-3 times. 20 patients did not take their medication on own initiative when they felt worse and 16 when they felt better. 23 patients had at least one event of forgetfulness and one intentional omission during last month.

Conclusion: First results of our study showed that forgetfulness and intentional omission of taking medicines in chronic patients were present in a high number of patients. Association between the findings will be tested further.

References:

• 1. Sabaté E (editor). Adherence to long-term therapies: evidence for action. 2003 World Health Organization

Disclosure of Interest: None declared.

182 ISoP18-1266 Evaluation of Drug-Related Health Problems by Diagnosis in Children

182.1 A. Akici^*1, N. I. Kirmizi¹, N. Akici², F. İsli³, M. Aksoy³

182.1.1 ¹Department of Pharmacology, Marmara University, Istanbul, Turkey; ²Department of Pediatrics, Haydarpasa Numune Training and Research Hospital, Istanbul, Turkey; ³Turkish Medicines and Medical Devices Agency, Ministry of Health, Ankara, Turkey

Background/Introduction: Drug-related health problems (DRHP) are one of the major reason for hospitalization in worldwide [1]. Children that more vulnerable to these problems represent a highly critical population [2]. As pediatricians are more likely to face DRHPs in children, addressing the extent of these problems in diagnosed by pediatricians can provide more information about DRHP in children.

Objective/Aim: This study aimed to investigate the diagnosis-based distribution of DRHPs in children.

Methods: Diagnoses prescribed to children (< 18-year-old) by pediatricians in Turkey were examined. The diagnoses in the prescriptions registered between 2015 and 2016 in the Prescription Information System of the Turkish Medicines and Medical Devices Agency were analyzed retrospectively. The details of the diagnoses containing DRHPs were examined according to the ICD-10 coding system (F11, F13, F15, F16, F19, T36–T50, Y40–Y59, X40–X44, X60–X64).

Results: A total of 11,642,110 diagnoses were found to be registered during the two-year study, where 838 DRHPs were identified (0.01%). It was detected that 77.3% of these diagnoses belonged to “Y40–Y59” codes (drugs, medicaments, and biological substances causing adverse effects in therapeutic use), which was followed by “X40–X44” coded-diagnoses (accidental poisoning by and exposure to noxious substances-excluding alcohol, gases etc.) with 16.3%. When we examined the details of the 648 diagnoses belonging to the codes of “Y40–Y59”, we observed the most common codes are “Y57” (other and unspecified drugs and medicaments, 29.0%), “Y59” (other and unspecified vaccines and biological substances, 28.7%), and “Y44” (agents primarily affecting blood constituents, 17.0%).

Conclusion: It was the first study to describe various DRHPs experienced by pediatric patients in Turkey. Available findings may shed light on the prevention of DRHPs, especially the most common ones, and on the development of respective pharmacovigilance strategies.

References:

• 1. Shehab N, Lovegrove MC, Geller AI, Rose KO, Weidle NJ, Budnitz DS. US emergency department visits for outpatient adverse drug events, 2013–2014. JAMA 2016;316:2115–25

• 2. Bouvy JC, De Bruin ML, Koopmanschap MA. Epidemiology of adverse drug reactions in Europe: A review of recent observational studies. Drug Saf 2015;38:437–53

Disclosure of Interest: None declared.

183 ISoP18-1267 An Online Collaborative, Open Access Pharmacovigilance Platform for Resource-Limited Countries

183.1 E. Allen^*1, C. Pace²

183.1.1 ¹University of Cape Town, Cape Town, South Africa ; ²Liverpool School of Tropical Medicine, Liverpool, United Kingdom

Background/Introduction: It can be particularly challenging to evaluate drug safety in low-resourced settings for a variety of reasons, including a lack of infrastructure, sometimes weak regulatory environments and poor access to training and education. Although there are good online and on-the-ground options for researchers working in pharmacovigilance in these settings, some may yet be unaffordable. More could be done to pull together existing free or low-cost resources, and develop new such resources, for those with limited acess to funds.

Objective/Aim: To sustain an online pharmacovigilance platform that is accessible to all, interactive, comprehensive, practical and responsive to users, thereby increasing the capacity for high quality drug safety evaluations in low-resourced settings.

Methods: Global Pharmacovigilance (https://globalpharmacovigilance.tghn.org) was launched mid-2016 with seed funding through the ACT Consortion (http://www.actconsortium.org), a global collaboration of researchers answering key questions about artemisinin-based combination therapies for malaria. Since its inception the website, which is not restricted to malaria, has provided links to a range of pharmacovigilance resources, including notices of events, training and education opportunities, special interest groups, and news items and articles on a diverse range of related topics.

Results: Coordinated voluntarily, the platform has grown to be consistently in the top 6 most accessed sites that form The Global Health Network (https://tghn.org), an online science park of 36 specialist health research-related professional member areas. Twitter and RSS feeds have been added and the focus has broadened to include non-drug interventions. As other member areas of TGHN have grown, so this one has drawn in their safety-related resources in a cost-effective and efficient way through its inter-connected infrastruture (and vice versa, raising the profile of pharmacovigilance across the TGHN membership of over 400,000 researchers). The focus going forward is encouraging those working or otherwise engaged in pharmacovigilance to contribute short articles, career advice and potentially mentorship, to ensure content and advice is of a high quality (as it is authored by experts), reflects current thinking, and showcases efforts being made globally to improve the safe use of medicines and other interventions.

Conclusion: The outputs hosted through Global Pharmacovigilance are expected to contribute to improving the practice of safety evaluations in resource-poor settings. We call on those involved in pharmacovigilance to support this platform by sharing their experiences and expertise so that those in under-resourced settings, and their patients, may benefit.

Disclosure of Interest: None declared.

184 ISoP18-1268 Pharmacovigilance Teaching to Undergraduates of Medicine and Surgery School: the Experience of Verona University

184.1 L. Magro^*1, U. Moretti¹, E. Arzenton¹, R. Leone¹

184.1.1 ¹Department of Diagnostics and Public Health, Section of Pharmacology, University of Verona, Verona, Italy

Background/Introduction: Teaching pharmacovigilance (PV) to undergraduates of Medicine and Surgery School is an educational objective of great importance. Their acquisition of theoretical bases and practical skills in detecting and reporting adverse drug reactions (ADRs) is an essential pre-requisite for their future contribution to pharmacovigilance activities. Unfortunately, at national level pharmacovigilance is not compulsory within the Degree Course of Medicine and Surgery (MD) in Italy and its introduction within the course of Pharmacology depends on the individual initiative of some professors.

Objective/Aim: Pharmacovigilance teaching to undergraduates of Medicine and Surgery School at University’s Verona, Italy.

Methods: The teaching methods included 15 h of frontal teaching, problem based learning (PBL) and problem solving (PS). The course aimed to the development of pharmacovigilance knowledge and skills and focused on basis of pharmacovigilance, regulatory aspects and benefit/harm assessment. With regard to PBL and PS, students, divided into groups with the presence of a tutor, were trained on drug–drug interactions, medical diagnosis, recognizing, managing and reporting ADRs through the analysis of a clinical case report. At the end of the course the students’ knowledge was assessed through an oral exam.

Results: At the University of Verona pharmacovigilance has been taught within the Degree Course of MD for at least 20 years, since the professors of Pharmacology are involved in the activities of the Regional PV Centre. Until last year the teaching of pharmacovigilance was only among the training objectives of the Pharmacology course, whereas, starting from the academic year 2017–2018, the course has been divided into two modules: General/Special Pharmacology (8 Educational Credits-ECTS) and Pharmacovigilance (1 ECTS), both carried out during the fourth year of education (120 students). At the conclusion of the PV module students:

> were able to understand the mechanisms of ADRs,

> were able to understand the classification of ADRs,

> knew the epidemiological data on ADRs,

> were able to fill in a reporting form.

Conclusion: Teaching PV is particularly important for MD students, since their knowledge on PV can influence their future ability to prescribe and monitor drug therapies. As a matter of fact, a physician who has the ability to identify ADRs and the habit of reporting them, will certainly pay attention to all aspects related to the patient’s health and consequently to both positive and negative effects of a drug therapy.

Disclosure of Interest: None declared.

185 ISoP18-1269 Exploring the Association Between Monoclonal Antibodies and Depression and Suicidal Ideation and Behavior: A Vigibase Study

185.1 L. Minnema^*1,2, T. Giezen^2,3, P. Souverein¹, T. Egberts^1,4, B. Leufkens¹, H. Gardarsdottir^1,5

185.1.1 ¹Division of Pharmacoepidemiology & Clinical Pharmacology, Utrecht Institute for Pharmaceutical Sciences (UIPS), Utrecht, the Netherlands; ²Medicines Evaluation Board, Utrecht, ³Foundation Pharmacy for Hospitals in Haarlem, Haarlem, the Netherlands; ⁴Department of Clinical Pharmacy, University Medical Centre Utrecht, Utrecht, the Netherlands, ⁵Department of Pharmaceutical Sciences, School of Health Sciences, University of Iceland, Reykjavik, Iceland

Background/Introduction: Recently, exposure to monoclonal antibodies has been associated with neuropsychiatric effects, e.g., depression, and suicidal ideation and behavior. So far, little is known about any differential risk between individual monoclonal antibodies and the potential underlying mechanism.

Objective/Aim: To quantify and characterize spontaneously reported adverse drug reactions (ADRs) of monoclonal antibodies related to depression and suicidal ideation and behavior. Furthermore, to explore the association between the mechanism of action of the monoclonal antibodies and these ADRs.

Methods: ADRs reported until December 2017 in VigiBase, the WHO global database of Individual Case Safety Reports, were included. Reports related to depression and suicidal ideation and behaviour (MedDRA standardized search) for monoclonal antibodies were extracted. Monoclonal antibodies that had been authorised by any global regulatory authority (e.g. FDA, EMA) for at least 3 years were included. Associations were tested by estimating reporting odds ratios (RORs) for the monoclonal antibodies separately (using bevacizumab as a reference) as well as grouped on their influence on the immune system (not influencing the immune system (reference), suppressing the immune system, and stimulating the immune system). Monoclonal antibodies suppressing the immune system were further stratified according to their intended indication (autoimmune diseases, cancer).

Results: A total of 44 monoclonal antibodies were included for which a total of 2924,319 adverse drug reactions were reported. For these 44 monoclonal antibodies, 9455 (0.3%) reports related to depression and 1770 (0.1%) reports related to suicidal ideation and behaviour were analysed. For both depression and suicidal ideation and behaviour, natalizumab and belimumab showed the highest ROR relative to bevacizumab, i.e. for depression 5.7 (95% CI 5.0–6.4) and 5.1 (95% CI 4.2–6.2), for suicidal ideation and behaviour 12.0 (95% CI 7.9–18.3) and 20.2 (95% CI 12.4–33.0) respectively. Monoclonal antibodies suppressing the immune system showed higher ROR relative to monoclonal antibodies not influencing the immune system, i.e. for depression 1.9 (95% CI 1.8–2.0) and for suicidal ideation and behaviour 3.6 (95% CI 3.0–4.4). This difference was only seen in the monoclonal antibodies suppressing the immune system that are used for treating autoimmune diseases.

Conclusion: Adverse neuropsychiatric effects are seen in patients exposed to monoclonal antibodies. Two antibodies peaked relative to bevacizumab (i.e. natalizumab, belimumab) regarding reporting of depression and suicidal ideation and behaviour. Furthermore, monoclonal antibodies used for treating autoimmune diseases showed higher RORs compared to monoclonal antibodies not influencing the immune system. For the interpretation of these data the indications for use and other population characteristics need further consideration.

Disclosure of Interest: None declared.

186 ISoP18-1272 Comparison of Reported Adverse Events of Premature and Term Born Infants Following Childhood Vaccinations in the Netherlands

186.1 L. van Balveren^*1, F. van Hunsel¹, E. van Puijenbroek¹

186.1.1 ¹Netherlands Pharmacovigilance Centre Lareb, ‘s-Hertogenbosch, The Netherlands

Background/Introduction: Most reported common adverse events (AEs) following infant vaccination are well-known. In The Netherlands infants receive two different vaccines; a pneumococcal vaccine and a diphtheria, tetanus, pertussis (acellular, component) hepatitis B, poliomyelitis and Haemophilus influenzae type b vaccine (DTaP/IPV/Hib/HepB) at the gestational age of 2, 3 (only DTaP/IPV/Hib/HepB), 4 and 11 months. Some of the reports on AEs received concerned premature babies. This raised the question whether or not there was a difference between reported AEs of preterm and term born babies.

Objective/Aim: To determine differences between the pattern of reported AEs of spontaneous reports following infant vaccinations between premature and term born babies.

Methods: Retrospective analysis of reported AEs to pharmacovigilance centre Lareb of reports following infant vaccinations between a premature and term born babies. Reports concerning AEs after infant vaccinations between December 2011 and May 2017 were selected. The AEs were listed for premature and term born babies. Odds ratios were calculated to compare the AE pattern for both groups.

Results: The most reported well known AEs in both methods were comparable. Table 1 shows the frequencies of occurrence of these reported AEs in spontaneous reports of term and preterm babies. Statistically significant differences were found for apnoea (OR = 90; 95% CI 36–224), apnoeic attack (OR 109; 95% CI 51–232), decreased oxygen saturation (OR = 86; 95% CI 30–249) and syncope (OR = 2.8; 95% CI 1.5–5.1), which were more frequently reported for premature babies.

Conclusion: This study shows that the pattern of reported AEs is comparable between both groups for well-known common AEs. Apnoeic attacks and comparable symptoms like bradycardia and a decreased oxygen saturation are events also known to occur in premature born babies not being vaccinated. These events are described as rare AE in the SmPC, especially in premature children.

Disclosure of Interest: None declared.

187 ISoP18-1274 Gynecomastia and Galactorrhea: Unlabeled Adverse Drug Reactions of Retinoids used in Dermatology

187.1 M. Atzenhoffer^*1, S. Pierre¹, F. Bellet², S. Pinel³, L. Javot⁴, T. Vial¹, B. Kassai¹, M. Auffret¹

187.1.1 ¹Centre Régional de Pharmacovigilance, Service Hospitalo-Universitaire de Pharmacotoxicologie, Hospices Civils de Lyon, Lyon, France; ²Centre Régional de Pharmacovigilance, Centre-Hospitalo-Universitaire de Saint-Etienne, Saint-Etienne, France; ³Centre Régional de Pharmacovigilance Paris Fernand-Widal, Hôpital Lariboisière-Fernand Widal—APHP, Paris, France; ⁴Centre Régional de Pharmacovigilance, Service Hospitalo-Universitaire de Pharmacologie Clinique et de Toxicologie, CHRU de Nancy, Nancy, France

Background/Introduction: Gynecomastia and galactorrhea are rare adverse drug reactions (ADRs) mainly associated with exogenous estrogens, antiandrogens, 5 alpha-reductase inhibitors, spironolactone and cimetidine. Several case reports of gynecomastia and galactorrhea have been reported in the literature with retinoids [1, 2]. However, these ADRs are not listed in the Summary of Product Characteristics of retinoids.

Objective/Aim: The aim of this study was to describe cases of gynecomastia and galactorrhea concomitant with the use of retinoids in the French Pharmacovigilance database (FPVD).

Methods: All cases corresponding to the MedDRA term “breast disorder” (High Level Group Term) and associated with oral or topical retinoid use for a dermatological purpose were extracted from the FPVD between 1985 and May 2018. Cases were excluded if another drug- or non-drug etiology was finally retained based on the French method for causality assessment.

Results: Thirty-one cases were included. Among them, there were 22 cases associated with oral isotretinoin use (14 cases of gynecomastia, 6 of galactorrhea and 2 of both gynecomastia and galactorrhea), 8 cases of isolated gynecomastia with oral acitretin use and one case of gynecomastia with topical tretinoin use. The indications were mostly acne and psoriasis. Patients’ median age was 22 years [Interquartile range (IQR) 19–35]. Gynecomastia and/or galactorrhea were unilateral for almost half of the cases with known clinical description (7/15). The median time of onset was 90 days (IQR 39–347, n = 26). The outcome was known for 27 patients and a total recovery after withdrawal of retinoid was observed for 63% of them. There were two cases of positive rechallenge. Only two cases were considered as serious: one hospitalization and one case that needed surgery.

Conclusion: Gynecomastia and/or galactorrhea could be associated with the retinoid treatment for cutaneous disease. Physicians should be aware of these potential ADRs and inform their patients. For most of the cases, gynecomastia and galactorrhea resolved after withdrawal of the treatment.

References:

• 1. Larsen GK. Iatrogenic breast discharge with isotretinoin. Arch Dermatol 1985;121:450-1

• 2. Carmichael AJ, Paul CJ. Reversible gynaecomastia associated with etritinate. Br J Dermatol. 1989;120:317

Disclosure of Interest: None declared.

188 ISoP18-1275 Reactogenicity and Safety of the Measles–Mumps–Rubella Vaccine: A Prospective Observational Real World Study

188.1 C. Bucsa^*1, N. Bulik², A. Farcas¹, S. Muresan³, I. Muresan⁴, O. Oniga²

188.1.1 ¹Drug Information Research Center, ²Pharmaceutical Chemistry Department, “Iuliu Hatieganu” University of Medicine and Pharmacy, ³Sandra Muresan medical practice, ⁴Loan Muresan Medical Practice, Cluj-Napoca, Romania

Background/Introduction: Measles outbreaks are spreading in many European countries due to parents’ refusal to vaccinate their children. The highest number of measles cases in Europe during 2017 was reported in Romania [1].

Objective/Aim: To assess the reactogenicity and safety of MMR vaccine as reported by parents of vaccinated children, in real-world general practice.

Methods: Children vaccinated with MMR according to the national immunization schedule (1st dose at 1 year old and 2nd dose at 5 years old) in 2 general practitioners’ (GPs) offices from Romania during March 2016–May 2017 were followed-up for 6 months. Demographic characteristics, medical history, history of adverse events (AEs) after previous vaccination and prior use of medication were recorded. We collected the solicited symptoms (Table) within 4 days after vaccination via an online questionnaire, and any AEs urging medical visit within 6 months after vaccination via follow-up phone calls.

Results: A total of 216 children, 123 aged 9–17 months (Group 1) and 93 aged 4–7 years (Group 2) received 219 MMR vaccine doses alone or together with other pediatric vaccines. 3 children in Group 1 received 2 MMR vaccine doses (at 9 and 12 months) due to change in the national vaccination schedule. The incidence of solicited symptoms is presented in the Table 211 AEs urging medical visit were recorded during the 6 months after vaccination in Group 1 and 49 AEs in Group 2. The most frequent unsolicited AEs urging medical visit were upper respiratory tract infections, tonsillitis, bronchiolitis, diarrhea and otitis media. No death was reported.

Conclusion: Incidences of both solicited and unsolicited AEs were generally higher in 9–17 months age group, as expected. Likewise, incidence of solicited general symptoms was higher when MMR vaccine was co-administered with other vaccines.

References:

• 1. European Centre for Disease Prevention and Control. Measles in the EU/EEA: current outbreaks, latest data and trends—January 2018. Available from https://ecdc.europa.eu/en/news-events/measles-eueea-current-outbreaks-latest-data-and-trends-January-2018 [AccessedJuly 2018]

Disclosure of Interest: None declared.

189 ISoP18-1276 What is Risky about Risk Communications: a Case Study on Dengue Vaccine

189.1 K. Y. Hartigan-Go^*1, H. Larson²

189.1.1 ¹Zuellig School of Development Management, Asian Institute of Management, Makati, Philippines; ²Vaccine Confidence Project, London School of Hygiene and Tropical Medicine, London, United Kingdom

Background/Introduction: Dengue is a tropical disease that affects large populations and creates a heavy burden of disease. While there are public health strategies like vector control, intensive surveillance, clinical monitoring and intervention, these approaches are often inadequate in hyper-endemic areas. In late 2015, the first Dengue Vaccine was made available and became part of a critical intervention for poputlations at risk. The Philippines licensed this vaccine in December 2015 and adopted it for a pilot public health intervention in three regions of the country with heavy burden of disease. The cases of dengue were particularly high and rising since 2012. In some localities, a state of calamity was declared. There were brewing objections about this immunization program by some quarters. On the 29th November 2017, Sanofi Pasteur informed the Philippine government of their retrospective analysis that those with seronegative status and vaccinated woudl have a trend risk for sever dengue and increased hospitalization after 3 years of protection.

Objective/Aim: Lessons and recommendations will be discussed in the spirit of improving risk management and communications in a setting of public health, politics, culrture, media manipulation, poor understanding of science and benefit–risk even by health professionals, and anti-vaccine lobby.

Methods: Review of the government records, interviews and analysis of the socio-political-health events will describe the evolution of damage caused by false narratives in social media, news and investigations.

Results: A perfect storm of events followed leading to product withdrawal and stoppage of immunization. Social media was weaponized and false narratives were made including allegations that vaccine killed a number of children. The vaccine was labelled as dangerous. Fear was created among parents and children as media continually reported a sensational and bias news, leaving both public health insitutions and workers’ reputation damaged.

Conclusion: The Dengue Vaccine scare led to an unintended consequence. It created a loss of vaccine confidence. As a result, cases of measles went up due to poor adoption of vaccines more broadly in both the public programs and private clinics.

Disclosure of Interest: K. Hartigan-Go Other: The issue is a subject of investigation for which the author is involved. H. Larson: None declared.

190 ISoP18-1277 Anaphylactic Shock Secondary to Blue Patent Administration with Positive Skin Tests

190.1 B. Ouni¹, N. Fathallah¹, O. Kaabia², A. Brahem³, S. Larif¹, R. Slim¹, S. Haydar², C. Ben Salem^*1

190.1.1 ¹Pharmacovigilance Center of Sousse, Labratory of Research, Sousse University, Sousse, Tunisia; ²Department of Gynecology, ³Reanimation Care Unit, Farhat Hached University Hospital, Sousse, Tunisia

Background/Introduction: Patent Blue is commonly used for selective localisation of the lymphatic system before sentinel lymph node biopsy, but is known to induce severe anaphylaxis. The prevalence of anaphylactic reactions to patent blue is reported to be 1.8% (0.6–2.8%) [1].

Objective/Aim: To report a case of anaphylactic shock secondary to patent blue administration with positive skin tests.

Methods: A 50-year-old female patient without history of allergic diseases or any drug, food or latex allergies, was diagnosed with breast cancer (T2, N0, Mx). She underwent lymph node dissection with tumorectomy, panectomye. About 5 min after patent blue injection, the patient developed dyspnea, agitation and generalised skin eruption. The patient’s oxygen saturation was 71%, his blood pressure was 70/50 mmHg with tachycardia. The diagnosis of anaphylactic shock was retained. The patient was recieved adrenaline and fluids. Thereafter, the patient symptoms improved and she fully recovered without apparent complications. Skin prick tests performed (after obtaining informed consent from the patient) returned very positive for patent blue. For that reason, we contre-indicated the use of patent blue.

Results: Patent Blue is commonly used to identify the lymph node before sentinel lymph node biopsy, but is known to induce severe anaphylaxis mediated by IgE. The prevalence of anaphylactic reactions to patent blue is reported to be 1.8% (0.6–2.8%) [1]. The skin test (skin prick or intradermal) is used seems to be the ideal test with high sensitivity and specificity to identify patients with hypersensitivity to patent blue. Our patient had a hypersensitivity reaction grade II which evolved satisfactorily without sequelae with positive skin prick test [2].

Conclusion: Both anaesthetists and surgeons should be aware of severe anaphylactic reactions when administering Patent Blue.

References:

• 1. Parvaiz MA, Isgar B. Anaphylaxis and blue urticaria associated with Patent Blue V injection. Anaesthesia 2012;67:1275–6

• 2. Maranhão MV, da Nóbrega DK, Anunciação CE, Maia Bde A, Mariano PV. Allergic reaction to patent blue dye in breast surgery—case report. Braz J Anesthesiol 2016;66:433–6

Disclosure of Interest: None declared.

191 ISoP18-1279 Serious Adverse Drug Reaction Reports in the Nigerian Vigiflow Database from September 2004 to December 2016

191.1 A. Ibrahim^*1, C. K. Ogar¹, A. S. Abiodun¹, M. C. Adeyeye², H. Nosiri¹, E. Amadi¹

191.1.1 ¹PV/PMS Directorate, Wuse-Abuja, Nigeria; ²Director General, NAFDAC, FCT, Nigeria ³Director General, NAFDAC, FCT, Nigeria

Background/Introduction: Adverse drug reactions (ADRs) are a source of concern in healthcare delivery as they negatively affect patients. Serious adverse drug reactions (SADRs), which may occur during clinical care of patients, have greater impact on patients and the health system; both in terms of morbidity and financial burden. Documenting and sharing reports of serious ADRs help to generate critical data that provide evidence of how the healthcare system can be improved.

The Nigerian National Pharmacovigilance Centre collects spontaneous reports on ADRs experienced by patients using the Adverse Drug Reaction Reporting Forms. A review of the reports submitted between September 2004 to December 2016 was carried out to identify all ICSRs classified as serious in the VigiFlow database.

Obje