Double ethical review

Trials undertaken in a host country but sponsored, financed, and conducted by an external organisation, are called externally sponsored.16 The Council for International Organizations of Medical Sciences (CIOMS) Guidelines and the Nuffield Council on Bioethics recommend that they are subject to the ‘double ethical review’, that is, ethical review carried out in the study country and in the sponsor's country.16 ,25 This mechanism is also suggested by independent research groups,26 but it is generally not legally enforced, and the international GCP codes are silent on it.

We evaluated the double ethical review in a multicountry malaria trial,27 and in an emergency Ebola trial.28 Given the high number of involved countries or committees, it practically resulted in a ‘multiple’ ethical review. Procedurally, it was complex, due to a multiplicity of requirements. Its impact on the trials' timelines was variable. In the ‘routine’ malaria trial, it took up to 9 months to secure all approvals, which caused delays in the trial's milestones. In the Ebola trial, conducted under the tight schedule of a public health emergency, the ethical reviews were fast, with no impact on the study's milestones. Content wise, the double ethical review was beneficial: the complementarity of perspectives improved the quality of the protocols and the protection of patients and communities. But the efficiency of the process would have been enhanced by joint reviews, or at least by direct dialogue between the different committees.29 ,30 Direct dialogue would also have allowed exchanges across the committees, mutual learning and managing conflicting opinions or inconsistent recommendations (for instance, in the Ebola-Tx trial, it was requested at the same time to shorten the informed consent form, and to add more elements to it).

We suggest that, in absence of a pre-existing harmonised regulatory framework between the host and the sponsor country, the international GCP code incorporates the notion of double ethical review for externally sponsored trials, and recommends measures to make it more efficient, for example, by promoting direct dialogue between the involved committees.

Minors' consent

The international GCP code requires that when minors are enrolled in clinical trials, informed consent is given by the parents or a ‘legally acceptable representative’, acting in the minor's best interest. But the concept of ‘legal representative’ is not consistent with the practice in many LMICs, where guardians are not formally nominated by a tribunal. For instance, Vischer et al31 mention the difficulty of finding legally acceptable representatives for children as a major challenge when implementing GCP in African settings. In countries like the Democratic Republic of Congo,2 Burkina Faso and Benin, children are informally entrusted to relatives or community members. The management of the child's routine clinical care is delegated to these caregivers, but one cannot assume that the terms of delegation encompass decisions concerning participation in trials, particularly when the ‘informal caregiver’ is appointed on a temporary basis.

To skip this challenge, researchers might decide not to enrol children cared by an informal caregiver, but this would unfairly exclude from research the most vulnerable ones, in addition to creating a selection bias. Alternatively, they could ignore the question, by accepting the self-declared guardian without further investigation. This is likely to occur in practice, and brings the risk of obtaining the consent from a person not entitled to give it or not acting in the child's best interest. Local ethics committees, having access to comprehensive information on research carried out in their region, are in the best position to develop locally-tailored good practices. For instance, in Zimbabwe, the caregivers of HIV orphans were not legally entitled to decide on the participation in the AntiRetroviral Research fOr Watoto (ARROW) trial. The ethics committee thus waived the legal guardianship requirement, allowing caregivers to provide consent, provided that they signed an ad hoc affidavit.32

These challenges cannot be ignored. We suggest that the international GCP codes accept and regulate the notion of ‘ethically and socioculturally acceptable representative’, beside the one of ‘legal representative’. The way the representative is designated and the conditions under which he/she is entitled to act should be described in the protocol, be compliant with ethical principles and local laws and regulations, and should get an explicit ethical approval. This mechanism should not lower children's protection. On the contrary, it could strengthen it, by structuring and controlling informal mechanisms that are already in place.

Illiteracy

Illiteracy may exacerbate vulnerability, so adequate measures are needed to ensure protection of illiterate people in medical research. The international GCP codes require that an ‘impartial witness' confirms that their consent is given in full autonomy, with no undue influence or coercion. However, finding impartial witnesses for illiterates may be a major challenge in some African settings.31 In addition, local interpersonal dynamics may impair the function of the witness, especially in contexts with high illiteracy rates, where patients without a literate relative/friend are assigned a witness previously unknown to them. This may be perceived as a breach of privacy, or as an imposition of the views of the witness, rather than an opportunity to foster dialogue and improve understanding.2

These observations are limited to Central Africa and should be confirmed in comparable settings. However, they strongly suggest that the mere presence of an independent witness is not per se sufficient to ensure participants protection. The international GCP codes could acknowledge these complexities and underline the need of context-tailored measures to ensure that the ‘witnessed consent’ process achieves its objective of protecting, rather than being the mere implementation of a formal requirement.

Post-trial access

Research ethics guidelines concur that research may only be conducted in a given population if it is ‘pertinent’, and if there is a reasonable likelihood that the research results will be available to that population.16 In practice, this means that innovative medicines should be made available to those in need in the countries where the trials were conducted. If this does not happen, there are legitimate concerns that trials were delocalised for convenience reasons.12 But few legislations have binding clauses for post-trial access to research results to the study communities.

Available evidence shows that trials' globalisation is not accompanied by strategies for making new medicines available and affordable in the trial countries. A limited analysis of commercial trials for haematological malignancies showed that about 30% of phase 3 trials involved sites in middle income countries (MICs) that are usually unable to sustain the high costs of the marketed medicines.33 Others showed that clinical trials conducted in India, South Africa and Latin America, often did not result in market authorisations in the study countries.34 ,35

According to Dauda and Dierickx,36 benefit-sharing should be formulated into a legal framework for international research, to ensure a better protection of poor trials' communities. We would complete this call by advocating that the principles of ‘benefit-sharing’ and ‘post-trial access to research results to host communities’ are incorporated in the international GCP codes.

The community

There are major differences across countries, and across communities within a country, in terms of culture, population, economic indicators, education, infrastructure, quality of healthcare, etc.30 In addition, the perspectives on research held by the communities often differ from those held by investigators, sponsors and ethics committees. For instance, social vulnerability may create a ‘push’ to trial participation,4 while concerns surrounding the meaning and reasons for collecting biological samples may create a reluctance to enrol.37 ,38 Our experience strongly suggests that researchers should gain an upfront, in-depth and multifaceted understanding of the perspectives of each community, for example, of how research is perceived and how local features and beliefs may influence comprehension and voluntariness,39 and that interdisciplinary qualitative research may help to build a genuine partnership between clinical researchers and communities.

Community engagement is increasingly promoted in health research, particularly in the field of HIV-AIDS.40 However, there is no universally accepted definition of ‘community engagement’: the concept itself, and the way in which it is best implemented in practice, remains not sufficiently studied.41 In addition, there is no national or international regulation on community engagement,41 and the international GCP codes do not list the community among the actors in clinical research.

For pushing to consistently achieve true engagement with trials' communities across different groups and disciplines, and not only in participatory research, we suggest that international GCP guidelines formally include the community (and/or patients associations) among the key-actors in clinical research.

Non-commercial sponsors

The GCP definition of a sponsor, that is, ‘an individual, company, institution or organisation which takes responsibility for the initiation, management and/or financing of a clinical trial’, does not capture the situation of most non-commercial trials. These are ‘externally-funded’, thus the legal responsibility (sponsor) and the funding source (external funding agency) are de facto split. The ambiguity of the GCP definition exposes inexperienced sponsors to the risk of underestimating the scope and extent of their own responsibility.42 Furthermore, the dependence on external funding often results in budget constraints and/or poor flexibility,43 and the competition for external funds is so great that most non-commercial sponsors are not in the position to question the conditions set by funders.42 The sponsor's definition of new European Regulation on Clinical Trials,44 which uses the wording ‘setting up the financing of’ a trial instead of ‘financing’ it, better represents the reality of externally funded research.

The international GCP codes would better reflect today's reality of international research, if the distinction between commercial sponsors (in charge of ‘initiation, management and financing’) and non-commercial, externally-funded sponsors (in charge of ‘initiation, management and securing the financing from third parties’) was acknowledged.

Contractual agreements

Contractual negotiations may be a major challenge for non-commercial sponsors who often have limited contract management capacity.42 ,43 This seems particularly true for South-based sponsors. According to the Council in Health Research for Development (COHRED), ‘recent snapshot surveys of research institutions in the African and Asian regions have revealed some significant gaps in the contracting and contract management capacity in these regions’.45 The major challenges identified by COHRED include the negotiation on the overheads, the supply of investigational medicinal products (IMPs), the transfer and sharing of trials' data and samples, and the policy insurance contract.

Public guidance in these domains is either absent or quite vague, both in GCP and from national regulators. For instance, we lack guidelines on how to meaningfully negotiate data sharing and material transfer agreements. Similarly, there are few or no concrete guidelines for negotiating research policy insurance contracts,46 which may result in poor insurance arrangements, leading to insufficient protection for participants and to increased liability risks for the sponsors.

The international GCP codes could give helpful guidance for non-commercial research, by providing as annexes some guidance for key-research contracts. This could then be adapted at national level, especially in countries that lack detailed regulatory guidance.

Laboratory quality management systems

Laboratories processing specimens from clinical trials require an appropriate set of good practices. The upgrade of local laboratories and the harmonisation of laboratory quality systems represent a major challenge in LMICs, especially in rural or remote areas.21 ,46 Unfortunately, the international GCP codes put little focus on laboratory requirements, and they do not refer to the Good Clinical Laboratory Practices (GCLP) code, first issued in 2003 in the UK and retaken in 2009 by the United Nations Development Program (UNDP)-World Bank-WHO Special Programme for Research and Training in Tropical Diseases (TDR).47 The GCLP code provides a useful GCP-compliant framework for the analysis of biological samples from clinical trials.

The poor attention of GCP for trials' laboratories may be explained by the fact that commercial sponsors in HICs rely on accredited laboratories, already run under a structured quality system. Unfortunately, the neglect of GCP for trial's laboratories triggers poor awareness of their importance among non-commercial sponsors and funding agencies (which often ignore the GCLP code).

We propose that international GCP should include a dedicated section describing the qualifications, responsibilities and monitoring/supervision of trial laboratories, including a cross-reference to GCLP.47 This provision would put GCP in line with the priority needs of globalised clinical research, and would help to ensure the quality, integrity and traceability of laboratory-generated data in international clinical trials.

Investigational medicinal products

To ensure the quality of IMPs, the international GCP codes require compliance with locally applicable Good Manufacturing Practices (GMP). However, this is insufficient today, since a combination of globalisation of production and insufficient regulatory overview results in variable drug quality on the international market.48 ,49

Poor pharmaceutical quality may influence trials' results, and there is evidence that degraded, underdosed and non-bioequivalent medicines have or could have been used in clinical trials.50 ,51 These challenges were ignored by GCP in 1995–1996, but it is quite surprising that a more stringent reference to drug quality is not considered in the ongoing revision of the ICH GCP, since the lack of regulatory supervision on global pharmaceutical supply allowed poor-quality medicines to reach ICH countries,52 and at least in one case a falsified medicine could have been used as a comparator in a US trial.53

If poor-quality medicines were used in clinical trials, they would harm patients, lead to erroneous results, and inappropriately inform public health policy. We support the call of Newton et al50 that the international GCP codes should include more detailed and documented statements on the checking and reporting of the quality of drugs and medical devices in clinical research, and should require independent chemical analysis results for IMPs that are not approved by a stringent regulatory authority.