Urgent need to modernize pharmacovigilance education in healthcare curricula: review of the literature

Objectives Pharmacovigilance education is essential since adverse drug reactions (ADRs) are a serious health problem and contribute to unnecessary patient burden and hospital admissions. Healthcare professionals have little awareness of pharmacovigilance and ADR reporting, and only few educational interventions had durable effects on this awareness. Our future healthcare providers should therefore acquire an adequate set of pharmacovigilance competencies to rationally prescribe, distribute, and monitor drugs. We investigated the pharmacovigilance and ADR-reporting competencies of healthcare students to identify educational interventions that are effective in promoting pharmacovigilance. Methods The PubMed, EMBASE, Cochrane, CINAHL, PsycINFO, and ERIC databases were searched using the terms “pharmacovigilance,” “students,” and “education.”. Results Twenty-five cross-sectional and 14 intervention studies describing mostly medical and pharmacy students were included. Intentions and attitudes on ADR reporting were overall positive, although most students felt inadequately prepared, missed the training on this topic, and lacked basic knowledge. Although nearly all students observed ADRs during clinical rounds, only a few had actually been involved in reporting an ADR. Educational interventions were predominately lectures, sometimes accompanied by small interactive working groups. Most interventions resulted in a direct increase in knowledge with an unknown long-term effect. Real-life learning initiatives have shown that healthcare students are capable of contributing to patient care while increasing their ADR-reporting skills and knowledge. Conclusions There is an urgent need to improve and innovate current pharmacovigilance education for undergraduate healthcare students. By offering real-life pharmacovigilance training, students will increase their knowledge and awareness but can also assist current healthcare professionals to meet their pharmacovigilance obligations. Electronic supplementary material The online version of this article (10.1007/s00228-018-2500-y) contains supplementary material, which is available to authorized users.


Key points
• Undergraduate healthcare students have good intentions and positive attitudes on ADR reporting; however, they feel inadequately prepared and lack basic knowledge on this topic. • Current pharmacovigilance education is predominantly focused on lectures, sometimes accompanied by small interactive working groups although the long-term effects of this type of education are still unknown. • Real-life learning initiatives in pharmacovigilance have proven effective in increasing student knowledge and awareness and also assist current healthcare professionals to meet their pharmacovigilance obligations.
Electronic supplementary material The online version of this article (https://doi.org/10.1007/s00228-018-2500-y) contains supplementary material, which is available to authorized users.

Introduction
Most healthcare students enter clinical practice immediately after graduation and are required to prescribe, distribute, administer, and/or monitor drugs on a daily basis. In order to perform these responsibilities effectively and to ensure the safe use of medications, healthcare students (especially in medicine, pharmacy, dentistry, and nursing curricula) should acquire a minimum set of pharmacovigilance competencies before they graduate and start clinical practice [1,2]. Foreseeing, recognizing, managing, and reporting adverse drug reactions (ADRs) are an important part of rational and safe prescribing and are integrated into multiple steps of the WHO-six-step Guide to Good Prescribing [3]. It is a professional responsibility of all healthcare professionals. Despite this, healthcare curricula often teach little on pharmacovigilance and ADR reporting, with a median of 4-5.5 contact hours [4]. Numerous studies have expressed concern about the lack of healthcare professional competencies in pharmacovigilance [4][5][6]. This lack of undergraduate education and training in pharmacovigilance is consistent with the low level of knowledge, skills, and actions seen not only in physicians but also in practicing pharmacists, dentists, and nurses [7][8][9]. Unfamiliarity with pharmacovigilance, a low level of ADR-reporting skills, a lack of knowledge combined with negative attitudes like ignorance, fear legal liability, and lack of importance are thought to be related to the current inadequate response to many ADRs [10][11][12][13]. Several interventions (implementing protocols, educational workshops, or repeated emailing or telephone calls) have been implemented in an attempt to improve the competence of healthcare professionals [14][15][16][17], but these interventions are costly or fail to produce clinically relevant and long-term effects [8].
Despite the urgency of this problem, each year millions of medication users experience ADRs ranging from minor discomfort to hospital admission, permanent disability, or even death [18]. ADRs are responsible for 3.0-6.5% of all hospital admissions, 0.15% of all deaths, and could have been prevented in 47-72% of cases by good pharmacological and pharmacovigilance skills and knowledge [19][20][21][22].
Pharmacovigilance centers have an important role in the dissemination of current pharmacovigilance knowledge. Their data are mainly based on post-marketing reporting, which is essential for identifying previously undetected, uncommon, or serious ADRs. In most countries, pharmacovigilance center causality assessments of ADRs rely on a mixture of spontaneous reporting by healthcare professionals (physicians, pharmacists, nurses, and dentists) and patients. Since healthcare professionals have a different focus in ADR reporting, it is important to involve all parties [23][24][25][26][27]. With population aging, the increased use of prescription drugs and polypharmacy will probably lead to a drastic rise in the number of ADRs [28]. This together with ADR underreporting [29,30] and the lack of awareness and understanding of ADRs could lead to an even greater burden on patients and healthcare systems in the near future.
By studying the pharmacovigilance and ADR-reporting competencies of healthcare students, we aim to identify effective educational interventions that promote pharmacovigilance early in their education and career. The primary objectives of this review were therefore to analyze the following: (1) what is known about the pharmacovigilance competencies of healthcare students and (2) which educational interventions are effective in pharmacovigilance education.

General methodology
We searched the literature to analyze the current level of competencies and the effects of different undergraduate pharmacovigilance interventions, using the Kirkpatrick model of hierarchy of evaluation, as modified by Freeth [31]. Given the diverse outcome measures, no meta-analysis was performed.

Search strategy
With assistance of a medical information specialist (R.O.), the MEDLINE (PubMed), EMBASE, PsycINFO, Cochrane, CINAHL, and ERIC databases were searched for articles on pharmacovigilance education. MEDLINE was used as the standard medical research database. The Embase, PsycINFO, Cochrane, and CINAHL databases were used for articles published in biomedical and nursing databases. The ERIC database functioned as a supplementary detector for educational articles. All databases were searched until February 1, 2017, with database-specific queries [S4] without additional filters. All queries used Bpharmacovigilance,^Bstudents,^and Beducationô r commonly used abbreviations of similar terms (e.g., adverse drug reporting systems, undergraduate, and teaching, respectively). Articles were retrieved from the local university library or requested from the original authors, institution, or publisher. The references of relevant articles were screened using the snowball method [32].

Study selection
First, two authors (MR and BP) independently screened all articles for eligibility based on their titles and preset inclusion and exclusion criteria [Supplement Table 1]. If there was any discrepancy about the content of the article, the abstract (if available) and/or full article was screened. Disagreements were resolved by mutual consensus. All eligible abstracts and articles were assessed in a similar way. Articles were included if they analyzed pharmacovigilance competencies in undergraduate healthcare students. Articles were not limited to the study setting, country of origin, or publication date. Exclusion criteria were as follows: (1) outcome measure not related to the pharmacovigilance competencies; (2) evaluation of a specialtyspecific ADR; (3) undergraduate healthcare students were not studied (e.g., healthcare professionals or patients); (4) language other than English or Dutch; (5) studying medical or dietary supplements, herbal products, or alternative medicines; and (6) non-original research studies (e.g., reviews, editorials, letters to the editor, and conference abstracts).

Data extraction
Data were extracted by two authors [MR and BP] using a modified coding sheet, based on the Best Evidence Medical Education (BEME) Collaboration coding sheet [33,34]. This modified coding sheet included the study design and aim, instruments used, characteristics of the educational intervention, students' educational level and performance, overall conclusion, and recommendations. The Kirkpatrick model of hierarchy of evaluation, modified by Freeth [31], was added to evaluate the outcome level.

Quality assessment
Study quality was assessed using the Medical Education Research Study Quality Instrument (MERSQI) [35]. This instrument was developed to assess educational studies and consists of six domains: study design, sampling, type of data, validity of the evaluation instrument, data analysis, and outcomes. Scores range from 5 to 18 points. Although there is no defined cutoff for highor low-quality study methods, a previous study considered scores of 5-8.5 to reflect a low-quality study method, 9.0-13.0 to reflect a moderate-quality study method, and 13.5-18 as a high-quality study method [36].

Data analysis
Data were analyzed using SPSS Statistics 22 (Chicago, IL). Descriptive statistics were used to report total mean MERSQI scores, proportion of articles with a different country of origin, type of healthcare student, and study design. The MERSQI scores of the main groups of student outcomes were compared using a one-way ANOVA with an alpha of < 0.05.
Given the differences in study design and outcome measures, only a quantitative analysis was possible. Student motives for reporting ADRs were described using descriptive statistics. Student opinions on educational aspects were recoded into three groups (No: ≤ 33% of students (fully) agreed, Neutral: ≥ 34 ≤ 66% of students (fully) agreed, Yes: ≥ 67% of students (fully) agreed.

Search results
The initial search identified 2468 unique articles. Figure 1 shows the flowchart of the search, selection, and review process. Thirtythree articles were eligible for inclusion. The 727 references of the 33 eligible articles were snowball searched, which yielded 6 new articles. In total, 14 intervention and 25 cross-sectional articles were included in our analysis.

Acquired pharmacovigilance competencies
As shown in Table 1, there is no uniform pharmacovigilance evaluation method. Most articles studied ADR reporting and pharmacovigilance knowledge (Kirkpatrick level 2b) in undergraduate medical and pharmacy students. Two studies [41,42] used identical research and outcome measures and have been compared separately [61].
Sixteen studies analyzed students' opinions of their perceived level of training in pharmacovigilance and ADR reporting (Supplement Fig. 1). One study of pharmacy students concluded that students felt sufficiently trained [37]. Conversely, six studies of pharmacy and medical students showed that students felt inadequately qualified to report ADRs or to perform pharmacovigilance [41-44, 48, 51]. Additionally, three (27%) studies reported that fourth-and fifth-year medical and pharmacy students also felt to have inadequate knowledge to report ADRs [42,49,51]. Healthcare students in almost all ( Kirkpatrick's four levels of training evaluations are as follows: Level 1-participation, covers learners' views on the learning experience, its organization, presentation, content, teaching methods, and aspects of the instructional organization, materials, and quality of instruction; Level 2a-modification of attitudes and perceptions: outcomes relate to changes in the reciprocal attitudes or perceptions between participant groups toward the intervention or simulation; Level 2b-modification of knowledge or skills: for knowledge, this relates to the acquisition of concepts, procedures, and principles; for skills, this relates to the acquisition of thinking problem solving, psychomotor, and social skills; Level 3-behavioral change: documents the transfer of learning to the workplace or willingness of learners to apply new knowledge and skills; Level 4a-change in organizational practice: wider changes in the organization or delivery of care, attributable to an educational program; Level 4b-benefits to patient or clients: this relates to any improvement in the health or well-being of patient clients as a direct result of an educational program pharmacy and medical curricula [38, 40-43, 45, 47, 48, 53, 57, 59, 64-66]. Two studies reported that dentistry [46] and nursing [56] students felt neither positive nor negative about including ADR reporting in their curriculum. Seven studies individually analyzed student reasons for reporting or not reporting ADRs to the competent authority [6,37,39,40,43,47,67] (Supplement Table 3). A lack of encouragement (n = 3), lack of information provided by patients (n = 2), and a lack of knowledge on how to report (n = 2) were the reasons most often given for not reporting ADRs. Educating others (n = 3), improving patient safety (n = 3), and contributing to the safe use of medicines (n = 3) were the reasons most often given for reporting ADRs.

What factors influence pharmacovigilance competencies?
Two comparative studies investigated differences in attitude and knowledge to pharmacovigilance and ADR reporting between medical and pharmacy students [47,57]. Sivadasan et al. [45,57] showed that more medical students than pharmacy students considered ADR reporting to be essential (80.5 vs 75.8%) and considered it their professional responsibility (69 vs 51.6%) [45]. Conversely, Umair Khan et al. showed that significantly more pharmacy students than medical students considered ADR reporting as important as managing patients (79.1 vs 43.5%) [47]. Both studies concluded that final-year pharmacy students had superior pharmacovigilance knowledge compared with medical students: 5.61 ± 1.78 vs 3.23 ± 1.60, 0-10 min/max and 8.4 ± 0.2 vs 3.17 ± 0.06; 0-15 min/max, respectively [47,57].
Additional comparisons between gender, race/ancestry, pharmacology curricula, previous pharmacovigilance or ADR-reporting training, previous ADR-reporting experience, and level of professional year were analyzed to identify factors associated with a higher level of pharmacovigilance competence. Race/ancestry did not influence pharmacovigilance knowledge, although male students knew more about post-marketing surveillance and female students knew more about causality assessments [38]. Overall, PharmD (Master of Pharmacy) students had more positive attitudes and higher knowledge scores than BPharm (Bachelor of Pharmacy) students [37,43], probably because the former had trained for longer. A positive correlation was found between student knowledge and their skills in ADR reporting (r 0.485, p < 0.001) [50]. Previous training in ADR reporting or reporting experience was associated with significantly higher student knowledge scores [40,47,53]. In line with these observations, academically older students had more knowledge, were more aware of ADRs during their internships, and had reported more ADRs.

Which pharmacovigilance interventions are effective?
There is no uniform pharmacovigilance educational intervention (Table 2). Interventions have ranged from short 15-min power point lectures and multiple training workshops to more innovative clinical experiences in ADR reporting or assessment. No replicated intervention studies have been published to our knowledge.
Four articles evaluated student satisfaction regarding pharmacovigilance education [65,67,75,77]. Students found clinical experience more educational than lectures and/or solving fictional casuistry [67]. Students also stressed that pharmacovigilance training should be repeated during the internships [77]. Six articles examined students' intentions and attitudes toward ADR reporting after a pharmacovigilance intervention [65-67, 73, 74, 77]. However, since none of the studies included a baseline assessment and substantial differences were not observed between cross-sectional and intervention studies, it was not possible to draw conclusions.
Two studies by Durrieu et al. focused on students' perception of the risk of ADRs [73,74]. They concluded that after a pharmacology course, students were more aware of potentially serious ADRs. A follow-up study showed that perception of the risks of ADRs was more clinically realistic after clinical training, i.e., students were more aware of potentially serious ADRs associated with anticoagulants and non-steroidal antiinflammatory drugs (NSAIDs) and less conservative about hypercholesterolemia drugs.
Five studies showed a significant increase in pharmacovigilance and ADR-reporting knowledge scores directly after the intervention was completed [66][67][68][69]71]. Since most studies asked different pharmacovigilance questions or used grouped outcome scores [68,69,71], it was not possible to state that one intervention was superior to another. Studies with a longer follow-up time (1-12 months) reported contrasting outcomes. Two studies showed a significant increase in pharmacovigilance knowledge and ADR-reporting skills after 1 and 6 months [71,77]. However, Arici et al. reported a significant increase in pharmacovigilance knowledge directly after an intervention, but this had faded by 12 months [68].
Three studies analyzed pharmacovigilance or ADRreporting skills [60,67,77]. Schutte et al. showed that medical students were significantly more aware of the importance of ADR reporting after assessing a real ADR report themselves [67]. Tripathi et al. and Rosebraugh et al. analyzed the impact of an intervention on the quality of completing a fictional ADR report in undergraduate medical students [60,77]. Both showed that a 15-min lecture significantly increased the quality of an ADR report.
Four articles analyzed pharmacovigilance competences in a real-life clinical setting [67,70,72,76], three of which involved pharmacy students [70,72,76]. Findings suggested that pharmacy students could play an important part in regular  [72,76], and Armando et al. found that second-year pharmacy students were equally capable of recognizing ADRs in a community pharmacy setting as pharmacists [70]. Schutte et al. showed that medical students were also capable of assessing real ADR reports [67].

Discussion
We found that while healthcare students have favorable intentions and positive attitudes toward ADR reporting, most lack the basic skills and knowledge to do so. Overall, academically older students and students with prior pharmacovigilance experience were more competent in recognizing and reporting ADRs. Pharmacy students had slightly more knowledge of pharmacovigilance and ADR reporting than other healthcare students. Students agreed that pharmacists are the most important healthcare professional with regard to pharmacovigilance, although all students felt responsible for pharmacovigilance. Students perceived their knowledge to be moderate at best, felt they did not receive sufficient training, and stated that pharmacovigilance and ADR reporting should be included in their curriculum. It is not surprising that while relatively many students had seen an ADR (63%), few had reported one (10%). This is consistent with previous studies [78] and the current low rate of ADR reporting (medial reporting rate of 6%) [29] among qualified health professionals. Despite this lack of competence in pharmacovigilance and ADR reporting, we identified 14 studies that reported beneficial effects of an intervention. Students valued real and legitimate pharmacovigilance tasks, such as diagnosing, reporting, or assessing ADR reports, more than outdated educational interventions or fictional casuistry. This type of clinical training also leads to a more clinically realistic perception of the risk of ADRs. Although educational pharmacovigilance interventions ultimately aim at a clinically relevant and long-term increase in medication safety, no study has looked at this highest hierarchical level. Most outdated interventions only provide a short-term increase in knowledge, few show clinically relevant results, and none has shown durable clinical outcomes. Repeated clinical training which boosts intrinsic motivation and improves learning outcomes [79,80] should be applied to pharmacovigilance training. Additionally, the interventions that focused on real and legitimate clinical tasks, such as diagnosing and reporting ADRs and assessing ADR reports, also had a positive effect on the healthcare system. Multiple studies have shown the clinical value of student participation in pharmacovigilance tasks.
Although our findings are worrying, the outcome should be interpreted with some caution given the heterogeneity and methodological weaknesses of the included studies. All Significant increase in ADR-reporting skills after 1 and 6 months.
ADRs, adverse drug reactions a Kirkpatrick's four levels of training evaluations are as follows: Level 1-participation: covers learners' views on the learning experience, its organization, presentation, content, teaching methods, and aspects of the instructional organization, materials, and quality of instruction; Level 2a-modification of attitudes and perceptions: outcomes relate to changes in the reciprocal attitudes or perceptions between participant groups toward the intervention or simulation; Level 2b-modification of knowledge or skills: for knowledge, this relates to the acquisition of concepts, procedures, and principles; for skills, this relates to the acquisition of thinking problem solving, psychomotor, and social skills; Level 3-behavioral change: documents the transfer of learning to the workplace or willingness of learners to apply new knowledge and skills; Level 4a-change in organizational practice: wider changes in the organization or delivery of care, attributable to an educational program; Level 4b-benefits to patient or clients: this relates to any improvement in the health or well-being of patients clients as a direct result of an educational program intervention studies were single institution, had variable intervention designs, used different assessment methods of no clear relevance, and were ultimately of moderate study quality (mean MERSQI score 11.1). Since this is the first systematically performed review to investigate the current pharmacovigilance competencies of all types of healthcare students, we cannot compare our findings with those of other studies. A similar review, focusing on only a few competencies in medical students, reported similar outcomes [5]. This review had a number of limitations. Articles may have been missed, although we attempted to reduce the likelihood of this by searching six databases and using a snowball strategy. Overall, the studies were only of moderate quality, with low response rates, and small intervention groups, many of which had not been retested. Despite these weaknesses and the possibility that student capabilities were overestimated, because of publication bias, most competencies are still far from satisfactory. Moreover, the heterogeneity of assessment instruments used, outcome measures, and interventions, in combination with the combined competency scores in some studies, made a full comparison or meta-analysis impossible. However, this heterogeneity could mask some interesting features, since only few frequently reported variables were studied in detail. Lastly, the difference in location of crosssectional studies (66% in Asia) and intervention studies (24% in Asia) may have skewed the analyses.

Conclusion
This review highlights the urgent need to improve and modernize current pharmacovigilance education for undergraduate healthcare students. However, the best way to provide this education still needs to be established, but the content of pharmacovigilance education should at least be as real as possible. We suggest it is given real life context, i.e., with clinical relevance as early responsibility for the student (under supervision). It should be integrated into different healthcare curricula (medicine, pharmacy, dentistry, and nursing) and repeated throughout academic training, starting as early as possible, in the Bachelor phase. By offering real clinical pharmacovigilance training, students can not only increase their knowledge, awareness, and skills, but can also assist current healthcare professionals meet their clinical pharmacovigilance obligations. Future research should therefore focus on valid and reliable methods for assessing pharmacovigilance competencies in clinical practice. To successfully develop and initiate pharmacovigilance educational programs, further work is needed to evaluate educational interventions on Kirkpatrick's highest hierarchical levels, preferably in an inter-professional setting, with a multicenter design and a long follow-up. Internships or student-run clinics may be useful since they offer students early pharmacovigilance experiences with real responsibilities for patient care, with the advantage of assisting current healthcare professionals, limiting the level of underreporting, and ultimately preventing ADRs and increasing patient safety.

Compliance with ethical standards
Conflict of interest The authors declare that they have no conflict of interest.
Originality and agreement statement All of the authors declare that this work has not been and will not be published in whole or part in any other journals, and agree to the contents of the manuscript in its submitted form.
Open Access This article is distributed under the terms of the Creative Comm ons Attribution 4.0 International License (http:// creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.