Background

Clinical research forms the basis for evidence-based clinical management of patients and can contribute to effective outbreak control. Although activities have improved collective preparedness to respond to public health emergencies [1, 2], experiences from previous outbreaks have highlighted many ongoing challenges for clinical research responses to epidemics [3, 4]. Some of these stem from the inherently unpredictable nature of emerging infections. Epidemics occur intermittently across geopolitical and cultural boundaries. Some can be forecast, and others emerge unexpectedly and disproportionally affect resource-poor settings with fragile healthcare systems and infrastructure, adding additional challenges to responses [5]. Previous epidemics have generated important information that has helped inform preparedness and response; however, it has also highlighted systemic challenges to our global capability to address important clinical research questions in these environments [1]. Clinical research takes time to plan, conduct and disseminate, a luxury that is rarely available during an outbreak. Ethical and regulatory frameworks designed for non-acute epidemics are not necessarily fit for the purpose of acute epidemic research [6, 7]. Conducting research under emergency conditions requires agility, intense activity, flexibility and adaptability to context [3, 4].

The aim of this scoping review is to identify how challenges to delivering essential clinical research during acute epidemics and pandemics have been approached, in order to inform strategies to strengthen our collective clinical research preparedness to emerging epidemics [8]. This is, to our knowledge, the first systematic scoping review of solutions to political, economic, administrative, regulatory, logistic, ethical and social (PEARLES) challenges to the design, delivery and implementation of clinical research during emerging epidemics and pandemics.

Methods

Drawing on PRISMA extension for scoping review guidelines [9], we developed a scoping review protocol in collaboration with researchers with experience in epidemic outbreak research and systematic evidence review methodologies.

Inclusion criteria

We included published, peer-reviewed quantitative and qualitative studies describing PEARLES challenges and solutions to clinical research responses to epidemics or pandemics identified during previous outbreak responses or through research involving health system stakeholders. We did not exclude reports based on study adesign. We included editorials and other ‘opinion’ articles when these were based on experiences derived from clinical research responses to emerging epidemics or pandemics. Conference abstracts were included as an important source of data not yet published in full [10]. We excluded studies presenting findings only relating to public health responses and not to clinical research. Studies presenting study outcomes without a reflection on challenges and/or solutions were excluded.

Search and retrieval of studies

The search strategy (Additional file 1) and terms were developed collaboratively with an information specialist who systematically searched six databases (Ovid MEDLINE, Ovid Embase, Global Health, Ovid PsycINFO, Scopus, Epistemonikos) for publications in English from 2008 to July 2018, to include up-to-date information relevant to clinical research responses today. The limits were set to capture recent, relevant clinical research responses to Public Health Emergencies of International Concern (PHEIC), where a clinical research response is vital to forward knowledge into risk factors and optimal clinical care to improve patient outcomes and outbreak control. The search terms were piloted by an information specialist and two reviewers. To ensure the search results were relevant and appropriate, after a review of the pilot search, restrictions were implemented using Boolean operators, before the strategy was finalised [9, 11, 12]. The search strategy was adapted for the Ovid databases to include the relevant thesaurus terms, in addition to searching the title or abstract fields (Table 1). Two reviewers independently screened the title and abstracts of the retrieved articles. If either of the reviewers considered a study potentially eligible, the full-text article was assessed independently for inclusion by two reviewers. Disagreements were resolved by a third reviewer. References were checked for additional potentially eligible studies.

Table 1 The search strategy for Scopus and Epistemonikos
Full size table

Data synthesis

One reviewer extracted data from the included studies using standardised forms including information on (1) study characteristics and setting, (2) participants, (3) intervention, (4) type of outcome measures and (5) PEARLES challenges and solutions. A second reviewer checked the extracted data. At the first analysis stage, we coded challenges and solutions according to the PEARLES categories. This showed that although the PEARLES categories were useful for the initial categorisation, there were overlap and interdependencies identified between these categories, especially between political and economic factors and regulatory, logistic and administrative factors and between ethical and social factors. Thus, at the second stage of the analysis, two reviewers identified the sub-themes and actions that emerged under these categories. A risk of bias assessment was not carried out since none of the studies formally evaluated the solutions identified during an epidemic or pandemic. Most of the studies presented challenges encountered while delivering clinical research responses during emerging epidemics and solutions implemented reactively, or solutions identified to address these, without formal evaluation. Lower evidence articles, including opinion pieces, were included to enable capturing the breadth and width of experiences from different settings, to give a voice to research teams delivering clinical research responses in difficult circumstances. Studies covering PEARLES challenges and solutions identified are summarised in the following sections under the interdependent themes that emerged.

Results

Of the 2673 articles identified through database searching, 234 full-text records were screened for inclusion, 71 of these met the inclusion criteria. Five additional articles were identified from references (Fig. 1).

Fig. 1
figure 1

PRISMA diagram

Full size image

Characteristics of included studies

The study designs of the 76 articles included were systematic review (n = 1), narrative reviews (n = 19), randomised controlled trials (n = 7), other randomised trials (n = 8; of which seven were stepped-wedge trials), case-control study (n = 1), cohort studies (n = 3), cross-sectional study (n = 1), time-in-motion study (n = 1), qualitative studies (n = 15) and editorial, comments or other ‘opinion’ pieces (n = 20) (Additional file 2). Most articles presented challenges and solutions identified during the Ebola virus epidemic in 2014 to 2016 and/or during the H1N1 pandemic in 2009. Some articles focused on more than one type of outbreak (Table 2). The studies which used an experimental design (e.g. RCTs) were reporting solutions implemented to deliver the intervention. Most articles related to clinical research responses set in lower-middle-income countries (LMICs) in Africa (n = 37), Latin America and the Caribbean (n = 3) and Asia (n = 1). Thirteen articles related to research responses in higher-income countries (HICs), and 22 articles focused on a global perspective. Most articles addressed more than one PEARLES domains (Fig. 2).

Table 2 Study setting and type of outbreaks. Several of the articles focused on a global perspective covered more than one type of outbreak
Full size table
Fig. 2
figure 2

Type of outbreak and PEARLES domains addressed. ^Articles focused on influenza, severe acute respiratory infections and pandemics; *Non-specified emergency epidemics. VHF, viral haemorrhagic fevers; Arboviruses, arthropod-borne viruses; CNS, central nervous system; ARI, acute respiratory infections

Full size image

Challenges and solutions

There were many solutions identified to address the multiple challenges encountered. These are presented as a narrative summary with illustrative examples from some of the more complex studies. Key actions that emerged are presented in Table 3. Many of these are cross-cutting across domains (Fig. 3).

Table 3 Key actions identified
Full size table
Fig. 3
figure 3

Key cross-cutting actions recommended

Full size image

Political and economic solutions identified (Table 4)

Political impediments to global collaborative networking and a lack of global coordination of funding and efforts were the key challenges encountered [1, 15, 20, 21, 27, 33, 42]. Delays in mobilising funding [15, 27, 36, 37], with approval sometimes taking longer than the outbreak duration, was a challenge during the H1N1 pandemic in HICs [15, 37] and again during the Ebola outbreak in LMICs [36].

Table 4 Solutions to PEARLES challenges encountered
Full size table

Strengthen collaboration and coordination

A cross-cutting theme identified was the need to strengthen collaboration and coordination between organisations involved in outbreak response at all levels [3, 7, 13,14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34,35,36,37, 81]. Effective partnerships between countries and international organisations, such as public health, clinical research organisations, non-governmental organisations (NGOs), pharmaceutical companies and regulatory agencies, were described as instrumental for success [3, 14, 16, 19]. International research collaborations should be tied to capacity building and be genuinely collaborative [22], with local stakeholders engaged from inception [1, 3, 13, 16,17,18,19, 22, 26, 29, 31, 32, 36, 38, 40, 44, 50, 51]. Support from national governments and local communities was a key enabler [16, 17, 23, 32, 36]. Continuous dialogue, led by governments of affected nations [16, 17], was identified as a facilitator for ensuring politically acceptable prioritisation and resource allocation [36]. This was illustrated by Doe-Anderson et al., where the implementation of a vaccine RCT during an Ebola outbreak was attributed to decisive action by the national government and an effective partnership between the USA and Liberia, with strong leadership from both nations [16]. By employing and training local doctors and scientists and renovation of existing sites for use in the trial, the study also strengthened the research capacity for future trials [16].

Establish dedicated funding sources and accelerated funding systems

There were many calls for dedicated funding for emergency research [1, 15, 19, 26, 27, 32, 35, 37], with financial mechanisms for rapid release of funds [1, 15, 20, 21, 26, 27, 33, 42]. Maintaining political awareness of the threat of infectious diseases to global health security (GHS) [35, 63] and an integrated approach to research was recommended to help marshal resources [1, 38]. A coalition of international stakeholders that would provide a global financing facility was suggested, to bring together funds to accelerate and prioritise research and development (R&D) [26, 27] and support R&D for communicable diseases neglected by the commercial market [13]. An example from the UK showed that through an emergency policy activation that allowed expedited funding, approvals and the redeployment of research staff, it was possible to launch a national, multi-site clinical trial within 12 weeks during the 2009 H1N1 pandemic [41, 53].

Invest in health systems and infrastructure in epidemic-prone regions

Limited healthcare systems [3, 29, 32, 39] and supporting infrastructure [1,2,3, 17, 29, 32] and overwhelmed healthcare facilities [31] brought about specific hurdles for delivery of research in LMICs. Investments to strengthen health systems and supporting infrastructure, targeting regions vulnerable to epidemics and pandemics, would facilitate effective responses [1, 2, 13, 29, 36]. Researchers delivering an Ebola vaccine RCT in Sierra Leone illustrated reactive solutions implemented to overcome logistic challenges. To enable recruitment of 8000 healthcare workers, they had to first renovate enrolment sites, laboratories and cold chain facilities and build study facilities and laboratories. Moreover, import freezer equipment and instal satellite-routed internet [29, 42, 43].

Administrative, regulatory and logistic solutions identified (Table 4)

Administrative and regulatory procedures and limited access to staff with research training were persistent challenges in LMICs and HICs [2, 3, 22, 29, 37, 43, 49, 50, 52]. Medical evacuation insurance requirements [42] and delays in recruiting international staff [38] posed additional challenges in LMICs. This can pose a risk of over-reliance on unpaid staff doing research [22, 29] on top of normal duties, with potential risk to routine patient care [72]. Multiple ethics committees, bureaucratic processes and inconsistency between required documentation were additional hurdles in LMICs and HICs [6, 15, 18, 40, 45, 46, 49, 50, 78], together with staff [42] and trial insurance [69] cover. The longest delays were often experienced in gaining site [48] and/or data sharing agreements, as documented in a time-in-motion study by Rishu et al. (Table 5) [15]. The infrastructure, staff time and an agreed standard required for dissemination of data were also often absent during times of crisis [2, 24, 35, 52] and were further compounded by long delays in institutions establishing data sharing agreements [15]. Some attributed a competitive research culture and a fear of losing power [22, 34] to a reluctance to share data [22, 34, 52, 73].

Table 5 Time to initiate an observational study into severe acute respiratory infections in Canada [15]
Full size table

Develop research capacity

The data shows a need to strengthen research capacity [1,2,3, 13, 29, 36] and invest in training for staff across the board [1, 15, 40, 42, 45, 48] particularly in high-risk regions. Primed clinical research networks globally [6, 24, 31, 37, 48] and a pool of researchers and experts that can be redeployed were recommended [1, 6, 41, 53]. The Platform for European Preparedness Against (Re-) emerging Epidemics was cited as an example of a clinical research network set-up to respond at the outset of an epidemic [20].

Frameworks and standards

Internationally agreed frameworks for emergency research to facilitate coordination, focus investments, and to guide implementation of responses are needed. These should identify emerging threats and develop roadmaps to focus R&D investments [19], as well as ethical, regulatory and operational standards for conducting emergency research [1, 7, 13, 19, 27, 35]. The World Health Organization’s (WHO) R&D Blueprint was cited as an example that aims to guide research efforts and set standards for high-priority pathogens [82].

Pre-approval and expedited, emergency protocols and frameworks

Pre-approved, pre-positioned study protocols was a key solution recommended to reduce set-up delays [1, 4, 13, 15, 20, 34, 37, 41, 46, 48, 50, 51, 56, 61,62,63]. These can be strategically developed for a range of syndromes and settings. An international agreement on a financial mechanism to manage clinical trial liability [32] and coverage provided by affected countries [32] was suggested to address delays in gaining insurance cover for all at-risk populations. An article by Lim et al. presented a trial in ‘hibernation’, with full regulatory approvals in place set up in the UK ready to be activated during a future pandemic [46]. Since not all eventualities can be predicted, there is also a need for expedited approval processes for emergencies [1, 2, 6, 7, 13, 18, 32, 34, 41, 45, 53, 54, 56, 64]. There were a couple of examples of expedited approvals (Table 6). A study by Annane et al. concluded that parallel rather than sequential scientific, financial, regulatory and ethics approval, and preparation of study drugs by local pharmacists would have enabled a multi-centre RCT of corticosteroids in ICU patients with H1N1 influenza pneumonia in France to start 1 month earlier, before the peak ‘flu’ wave [6]. Pollard et al. noted that despite expedited processes, the bureaucratic burden was undiminished [53]. Expedited reviews need to be balanced against risk for patients [56]. One article found that double ethical review improved the quality of an Ebola protocol and led to better protection of patients due to the complementarity of the reviews [67]. Ethics committee staff with experience from epidemic research and joint research ethics committees (RECs) with representatives from all affected countries was recommended to facilitate approvals [1, 2, 18, 45] and to protect patient safety [2]. RECs should ensure that protocols are consistent with community values [50], are collaborative and include capacity building [20, 22].

Table 6 Expedited ethical review time frames
Full size table

Dissemination

Keeping stakeholders informed of the study progress was cited as a facilitator for engagement and delivery and to prevent misinformation [5, 21, 29, 33]. However, interim data sharing needs to be managed carefully to reduce the risk of misinterpretations [69]. International agencies, such as WHO, are advised to provide a platform for harmonised data sharing [2, 13, 24, 27, 52] and support capacity for data recording in LMICs. In order to encourage data sharing, study approval and funding may be made on the premise of data sharing, and international agreements include guidance on data sharing between sponsors and host countries [52]. To control data sharing, mechanisms need to be in place to ensure that intellectual property, clinical governance and participant confidentiality are maintained [13, 69]. To overcome issues around traditional publication processes, scientific journals should review policies to improve essential data sharing during emergencies [27, 52, 68]. Journals were advised to pledge that data sharing during an emergency would not prejudice later publication [73]. A shift in perspective to a common goal, rather than publications, was called for [2, 52, 73]. There were no data on the implementation of emergency data sharing initiatives identified.

A publication paradigm must change when lives are at risk. Shaw et al [73]

Ethical and social solutions identified (Table 4)

The temporal and spatial variation in the risk of infection during outbreaks presents not only statistical hurdles but also an ethical challenge for study designs [75]. Previous research responses have highlighted the many challenges in agreeing on simultaneously ethical, scientifically valid and acceptable study designs [18, 36, 42, 51, 59, 64, 66], addressing socially valuable questions [14, 18, 64, 66]. The exclusion of children and pregnant women was an ethical concern during the Ebola outbreak in LMICs [18, 69], compounded by challenges of obtaining complex consent [2, 4, 5, 15, 18, 21, 25, 32, 34, 44,45,46, 50, 69, 70]. Ethical issues also arose around equity in access to health care for participants and non-participants [44, 50, 59].

There were calls for international standards for the conduct of emergency research [21, 51, 60, 62], including refined definitions of what constitutes research under these circumstances [51], acceptable study designs [14] and simplified consent methods [2, 3, 6, 25, 37, 46, 51, 70, 74]. Some studies explored waivered consent [2, 25, 69, 70, 74] and/or proxy consent [69]. However, gaining proxy consent was also found to be challenging during emergencies [15]. A qualitative study set in Europe found public support for consent waiver for publicly funded, low-risk studies and routinely collected anonymised biological samples for research, and for advanced or verbal consent models for pandemics [70, 74]. An interventional Ebola study set in Guinea reported patient preference for verbal consent [69]. One article highlighted the risks of using non-documented verbal consent [5]. Some, but not all countries, allow research without consent under emergency conditions [2]. Cook et al. argued that a priori and surrogate consent models may be contrary to the public good for research involving critically ill patients during an emerging pandemic and presented a contextually dependent consent model [2]. There was a consensus on recommendations for equitable access to the best available standard of care regardless of consent to participate [29, 32, 50, 59].

Community engagement

The need to engage communities is recognised as essential for an effective response. Several articles reported fear and mistrust of international responses as a challenge for research delivery [18, 51, 69, 71, 79]. Integrating all stakeholders into each step of the developing research programme [3, 48, 69], engaging communities as partners [1, 3] and gaining an understanding of power dynamics [34] were facilitators cited to address this. It was also emphasised that interventions need to be culturally sensitive [27, 29, 44, 50, 51, 58] and respond to national as well as global need [36]. Examples of community engagement included regular information sessions [3, 5, 16, 29, 34, 42, 69,70,71], joint communication plans [20, 29], outreach health promotion teams [69] and formation of community advisory boards [69].

Discussion

Our findings highlight the many challenges experienced in the planning, conduct and dissemination of clinical research responses during epidemic and pandemic emergencies and the gap in data on our collective capability to respond globally. Although a range of solutions to address these challenges were identified, we did not identify any studies that formally evaluated these during emergencies. We note the consistency with which key recommendations for change have been made across different outbreak experiences. Recommendations made to mitigate challenges experienced during the response to the H1N1 pandemic in 2009 were not globally implemented, and the same issues were documented again during the Ebola response in 2014 to 2016. Most challenges experienced were similar across outbreaks and settings, with additional challenges due to limited healthcare systems and infrastructure encountered in LMICs, and when initiating multi-site responses. The limited data on our capability to respond to a pandemic in LMICs is another area of concern. These hard-learned lessons need firm commitment and action to build clinical research preparedness for future scenarios [83]. Challenges to making progress are varied but include funding shortfalls in the wake of an epidemic as interest wanes [84], structural problems affecting epidemic responses more generally [85] and the absence of clear accountability for action in a policy area that depends on networked responses from many stakeholders across disciplinary and institutional boundaries [83].

This review shows a need to strengthen global collaborations and investments to strengthen research capacity and capability, targeting regions prone to and vulnerable to epidemics in inter-epidemic times. Progress has been made in some areas. The UK NIHR portfolio of ‘hibernating’ pre-approved protocols [86], the global research collaboration for infectious disease preparedness [8] and the WHO blueprint for R&D into high-threat pathogens [87] are steps in the right direction. However, the latter does not include pandemic influenza, and the WHO public health research agenda for influenza does not constitute a strategic plan of action [88].

The Platform for European Preparedness Against (Re-) emerging Epidemics (PREPARE) is an example of a clinical network primed to respond through a suite of active, syndromic studies across Europe [89]. ALERRT and Pandora-ID-Net are similar networks strengthening research capacity across Africa [90, 91]. These could serve as models for similar networks globally. Importantly, these and similar initiatives need to be evaluated and sustained to capitalise on existing investments. Strategic strengthening of local research capability globally will reduce reliance on external agencies to deliver research responses, while ensuring research responses are locally acceptable, ethical in the context and address local needs and interest. Local and regional empowerment may also reduce the risk of unethical studies on vulnerable populations [5]. Work is also needed to review synergies between different initiatives and disease programmes to optimise effectiveness, sustainability and coordination. Integration of research into outbreak response facilitated the further assessment of a candidate Ebola vaccine during the current outbreak in the Democratic Republic of the Congo (DRC), aiding the containment of the outbreak, illustrating the importance of clinical research for outbreak control [92]. A recent RCT of four Ebola treatments in DRC shows that well-planned scientifically and ethically sound clinical research can be delivered during prolonged outbreaks, through coordinated collaborative efforts between a wide range of stakeholders, including frontline staff and patients [93].

A key strength of this study is that it is the first to consider the wide range of challenges to the design and delivery of clinical research during emergency epidemics from a global perspective. The search strategy was comprehensive, spanning multiple databases and incorporating a range of peer-reviewed literature types. Nevertheless, due to the paucity of empirical evidence in this area, we did not carry out a formal quality assessment of included studies. Although findings are limited by restriction to English language publications, most articles addressed challenges to research responses in LMICs. Most studies were set in West Africa during the Ebola outbreak or in Europe and Northern America during the H1N1 pandemic, highlighting the limited information on research preparedness in LMICs and collective preparedness to deliver timely, coordinated research responses to emergency epidemics globally.

Conclusion

This systematic scoping review shows that potentially effective measures are not being universally implemented despite a good degree of expert consensus on their likely utility. Clinical research communities globally need to evaluate activities, implement solutions identified during previous emergency responses and rehearse and refine outbreak response plans in inter-epidemic times, in collaboration with other organisations involved in outbreak response. Although there may be examples of additional solutions identified in other regions of the world that were not included in this review, we have shown that there is already a substantial body of literature containing valuable experiences and important recommendations. However, without concerted, global action to act and to evaluate those actions in integrated outbreak response plans globally, we may be destined to encounter the same challenges and read about the same suggested solutions in the future. This would mean missed opportunities to forward knowledge into the clinical management of emerging infectious diseases, improve outcomes and strengthen global health security.