Objectives To assess the potential role of clinical trial (CT) registries and other resources available to research ethics committees (RECs) in the evaluation of complex CT protocols in low-income and middle-income countries.
Methodology Using a case study approach, the authors examined the decision-making process of a REC in Argentina and its efforts to use available resources to decide on a complex protocol. We also analysed the information in the USA and other CT registries and consulted 24 CT experts in seven countries.
Findings Information requested by the Argentinean REC from other national RECs and ethics’ experts was not useful to verify the adequacy of the REC's decision whether or not to approve the CT. The responses from the national regulatory agency and the sponsor were not helpful either. The identification of international resources that could assist was beyond the REC's capability. The information in the USA and other CT registries is limited, and at times misleading; and its accuracy is not verified by register keepers.
Discussion and conclusion RECs have limited access to experts and institutions that could assist them in their deliberations. Sponsors do not always answer RECs’ request for information to properly conduct the ethical and methodological assessment of CT protocols. The usefulness of the CT registries is curtailed by the lack of appropriate codes and by data errors. Information about reasons for rejection, withdrawal or suspension of the trial should be included in the registries. Establishing formal channels of communication among national and foreign RECs and with independent international reference centres could strengthen the ethical review of CT protocols.
- Clinical trials
- Drugs and Drug Industry
- Ethics Committees/Consultation
Statistics from Altmetric.com
The number of clinical trials (CTs) executed and of subjects recruited in Latin America (LA) is expanding, and there is concern that the systems to protect participants remain underdeveloped, facilitating the exploitation of the most vulnerable.1 In most LA countries, the responsibility for the ethical approval of CTs rests with the research ethics committee (REC)i of the institution where the CT is to be implemented or with external or commercial RECs. The performance of these RECs varies widely; while a few are able to adequately evaluate CT protocols, others are poorly staffed and do not have appropriate safeguards to avoid conflicts of interest.2–7
In addition to capacity problems, most LA RECs do not have the resources to monitor the implementation of CTs. The growing presence of commercial RECs, whose existence depends on maintaining good relationships with researchers and industry via the fast approval of protocols, further weakens the ethical review systems.5 ,8–10 As Hyder et al11 have mentioned, ensuring adherence to ethical rules and regulations, especially in countries and research institutions with a history of not protecting the rights of human research participants, is a daunting and almost impossible task. The context in which CTs are implemented in LA facilitates the violation of the principles of autonomy, justice and beneficence.12 ,13 There are also concerns that even in favourable environments, the increased diversity and complexity of CT protocols14–16 challenge the capacity of RECs to fulfil their mission due to limited access to experts that can resolve their queries. Several researchers have advocated the creation and/or strengthening of national/regional RECs5 ,17 and/or the development of networks and formal communication channels to allow RECs to consult and learn from each other.8 ,18 ,19
This case study revolves around two objectives: (1) to establish whether RECs, by consulting regulatory agencies, CT sponsors, other RECs or bioethicists, can improve their assessment of a CT protocol or the appropriateness of their decision to approve or reject the trial and (2) to determine whether the information contained in CT registries can help RECs in LA to evaluate CT protocols.
Argentina lies in the uppermost range of middle-income countries, with a per capita income of US$17 400 (2011). The national drug and food regulatory agency, the Administración Nacional de Medicamentos, Alimentos y Tecnología Médica (ANMAT), will not authorise a CT until its protocol has been approved by an institutional or commercial REC. Some provinces have provincial RECs, which have mostly an administrative role and do not participate in the ethical evaluation of protocols. Therefore, as in many other countries, the ethical approval of CTs rests in institutional or commercial RECs. Some research sites require CTs to be approved by their institutional REC, regardless of whether they have been approved by another commercial or institutional REC.20
Our case study originates in the Private Hospital of the Community of Mar del Plata (Private Hospital) in the city of Mar del Plata, a wealthy community in the Province of Buenos Aires. Private Hospital established its first Ethics Committee in 1995 and a REC to evaluate and monitor research protocols involving humans in 2000. The REC is composed of 14 members, many of whom have dual degrees, are trained in bioethics and have a wide range of areas of expertise: medicine (8), law (3), psychology, nursing, social work, theology and philosophy. The REC meets weekly. In addition to reviewing protocols, it conducts research and trainings in bioethics, publishes in academic journals and provides advice to government entities. It is considered a well-functioning REC that strives for excellence and one of the few that monitors the process of obtaining informed consent from all CT participants, about 75 per year.21
In 2010, Private Hospital was approached by Bayer AG (Bayer) to implement a randomised, double-blind trial to determine the safety and efficacy of moxifloxacin versus comparator in children (<18 years) with complicated intra-abdominal infections. This trial was to be implemented at 104 research sites in 22 countries, including the USA, Canada, Western and Eastern European countries, India and several LA countries.ii
The REC identified several methodological and ethical problems with the trial (see box 1), tried unsuccessfully to clarify them with the sponsor and decided against its approval. Once the decision was made, the REC wanted reassurance that they had reached the correct decision and sought the opinions of other CT experts and bioethicists, including the authors.
Issues identified during the research ethics committee's evaluation of the Bayer protocol22
The proposed statistical analysis was inappropriate to determine the safety and efficacy of moxifloxacin, which were, respectively, the primary and secondary objectives of the trial.
Participation in the study would require exposing children to excessive and unnecessary risks. Other effective treatments exist for these infections, and the use of fluoroquinolones in childrenvi has been discouraged because they can alter cardiac function by lengthening the QT interval25 and negatively affect joint development in young animals and occasionally children.13 ,26–29 The long-term effects of these drugs are unclear,24 and other studies have not been able to confirm the above mentioned effects.30
The patient information sheet downplayed the study's risks by stating that moxifloxacin is safe and effective in adults and adding that it was expected to be equally well tolerated in children.
Bayer was simultaneously conducting a pharmacokinetics study of moxifloxacin in children to determine its appropriate dosage for future clinical trials, further questioning the safety of this particular trial.
If the study results were used to unnecessarily expand the use of fluoroquinolones when other treatment options are available, antimicrobial resistance to this class of antibiotics could grow, negatively affecting the community at large.
The study follows a case study approach and was implemented in two phases. During the first, we examined the Private Hospital REC's path in deciding to reject the trial and their efforts to reassure themselves of having reached the correct decision. The REC contacted Bayer and consulted with Argentine experts, including the principal investigator (PI) at Central Hospital (Central),iii a prestigious private hospital located in Buenos Aires that had reviewed the same protocol; a REC member of another private hospital that was also implementing the study; 10 Argentine experts including bioethicists and members of other RECs; and ANMAT (see table 1).
The second phase emerged in response to the difficulties encountered during the first. The REC wrote the authors (16 August 2011), as coordinators of RELEM, a Latin American network for ethics and pharmaceuticals, help them to contact RECs and experts outside Argentina. Responding to the request we used the US CT registry (ClinicalTrials.gov) to identify sites and CT implementation status. To gather opinions on the ethical and methodological adequacy of the protocol, we contacted the persons and institutions presented in table 2.
We interviewed 24 bioethicists, CT experts and REC members face-to-face (lasting from 30 min to 2 h), by phone and/or by email.
Phase 1: the itinerary of the Argentinean REC
The PI at Central Hospital refused to discuss the trial, and the Hospital's lawyer said that they would share information only at the request of ANMAT.
A REC member of a hospital implementing the trial said that after checking with an infectious diseases expert, he could not find compelling reasons to reject the study on methodological or safety grounds and considered that the results of the study could be useful to mankind. However, he added that Private Hospital REC's arguments were valid, and as a precautionary measure his REC would recommend that the minimum age of eligible children be raised. In his opinion, RECs should not be concerned about the safety of the experimental product since this responsibility was under the purview of ANMAT and the National Health Commission.
Only three of the 10 Argentine experts approached responded and expressed agreement with the issues raised by the REC of Private Hospital. In addition, one mentioned that, in 2002, a CT involving the paediatric use of an antibiotic from the same group had been rejected because of safety concerns. A physician at a large and highly prestigious paediatric hospital indicated that his institution had rejected two research protocols for similar reasons, adding that this incident underscored the need to establish formal communication systems among Argentinean RECs.
When the REC asked ANMAT about the other Argentinean hospitals implementing the trial so that they could confer with their RECs and discuss their questions, ANMAT responded that they should not be concerned with the decisions of other RECs.
Bayer never responded to the inquiries.
The role of CT registries. We found that the US registry only listed the countries, cities and postal codes of the research sites; the names of hospitals, PIs and RECs' chairs were missing. Thus, it was impossible to contact, in a reasonable amount of time, all the sites in the additional 21 countries participating in the study.
The analysis of the US registry led to an unanticipated finding. The code used to indicate the implementation status of Bayer's CT at Private Hospital was ‘terminated’, which signifies that the trial had been approved but halted after enrolling at least one patient. The study could not have been halted because it never started. The US registry does not have a code to identify protocols that are not approved, and from the available codes, the code ‘withdrawn’, indicating that the study had been halted prior to enrolling patients—even if not optimum—would have reflected more closely the Private Hospital's REC decision.
The implementation status in four UK sites was classified as terminated. However, the ethics advisor of the UK's National Research Ethics Service confirmed that his organisation had not reviewed the protocol and the research and development department of one of the UK hospitals had no record of the study. He determined that Bayer had decided not to implement the study in the UK and never requested the national REC's approval.
We discovered that at least two hospitals in Spain that had approved the trial had decided not to continue before recruiting any patients but the code used in the US registry was terminated and not withdrawn. The latter was never used by Bayer (see table 3). We do not know how many of the 41 sites listed as terminated had withdrawn without recruiting any patients, or had rejected the study as did the Private Hospital, or had been excluded from the trial by Bayer as was the case in the UK. We confirmed that 13 sites classified as terminated had never recruited any patients, of which five had not approved the protocol. By December 2012, the registry did not include the name of the health authority responsible for the trial in six (Belgium, Brazil, India, Serbia, Spain and the UK) of the 22 countries originally listed as participants, and the CT implementation status of the sites in these countries was terminated.
The authors wrote Bayer (18 January 2012) requesting the name of the PI of the Spanish leading hospital that had approved the trial.iv Bayer did not respond; following a second request, we were told (6 March 2012) that the information was protected and Bayer could not share the PI's information without his/her written authorisation, and since the implementation of the trial had been subcontracted to a Contract Research Organization, Bayer did not have access to the documents. From other interviewees, we found that the reason for withdrawal was the difficulty of recruiting patients.
The person responsible for CTs at the Spanish regulatory agency mentioned in an email (24 January 2012) that the Bayer CT was included in a paediatric investigation plan (PIP) established by the European Medicines Agency (EMA). The CT's inclusion in the PIP led to its automatic approval in Spain, in February 2010. The authors subsequently asked the Spanish and Belgian regulatory agencies why EMA's decision was binding for Spain and not for Belgium,v but did not receive a response.
The director of the US registry responded (18 July 2011) to our inquiries regarding the incorrect CT implementation status codes found in ClinicalTrials.gov. In the email, she remarked,
Each trial record in ClinicalTrials.gov is “owned” by the trial sponsor (or in some cases the “responsible party”). This entity [National Library of Medicine] has no control over the data. Note that the proper status for a trial (or site) that stops prematurely is “withdrawn” if no subjects were ever enrolled, or “terminated” if one or more subjects had been enrolled. The sponsor of the trial in question [Bayer AG] would be the one to make the necessary changes.
The email added that the US government does not have the resources to verify the quality of the data entered by the sponsor or responsible party.
We consulted the European CT Registry (European Union Drug Regulating Authorities (EUDRA)) and found that (i) it does not include the same fields and codes used by ClinicalTrials.gov, (ii) most of the information is about European participants and (iii) places responsibility for uploading information in the registry with the national entities responsible for the approval of the trials. EUDRA includes a field (N) to record the views of an ethics committee and its reasons for unfavourable opinions. However, not all European agencies had submitted the information and some of the information was also found to be inaccurate; for instance, the Bayer CT appeared in EUDRA as being implemented in Belgium.
Views about ethical aspects of the trial. The Health Research Group (HRG) of Public Citizen responded immediately to our query stating, ‘We agree with the decision by Private Hospital's REC not to approve this study given the adverse safety profile of moxifloxacin and the availability of safer alternative antibiotics for treating children. Furthermore, the ethical problems with this study involving children would not be resolved by having an appropriately worded consent form.’ HRG added that the REC could write a letter of concern to the Food and Drug Administration (FDA) explaining the reasons for its decision. In response to our inquiry about the name of the person or office that should receive the letter, an FDA official responded (18 July 2011): ‘While there is currently no regulatory requirement for an IRB to inform FDA when they do not approve a study, we would not discourage any reporting of this nature to the agency.’
The person responsible for CTs at the Spanish regulatory agency (Agencia Española de Medicamentos y Productos Sanitarios (AEMPS)) clarified (24 January 2012) that better options existed for the treatment of intra-abdominal infections and that the EMA wished to restrict the use of moxifloxacin to severe infections not responding to other treatments. She added that EMA had requested that Bayer (1) perform a preclinical study in young animals, (2) execute a pharmacokinetics study in humans to determine the appropriate moxifloxacin dosage for different age groups and (3) develop pharmaceutical forms appropriate for different age groups. In her view, the trial was justified because it would allow the development of pharmaceutical forms suitable for paediatric patients and would help determine the safety profile of the moxifloxacin in the paediatric population. She affirmed that some children could benefit from this treatment, including immunosuppressed children and children with multidrug-resistant tuberculosis.
Interviews with professionals who work or had worked at the Peruvian CT regulatory agency (Oficina General de Investigación y Transferencia Tecnológica (OGITT)) revealed that a number of ethical issues and doubts had been raised during the discussion of the Bayer protocol. However, OGITT felt pressures to approve it, and since a commercial REC had approved the protocol for the four participant hospitals, it was difficult for them to deny the implementation of the trial.
The bioethics unit of Pan American Health Organization, whose only responsibility is to evaluate the research protocols they finance, responded that it had not had the time to review the risks of the study and remained concerned about the use of the word ‘terminated’ when the trial had not been approved by the REC.
The case presented brings forward a number of ethical issues including the need for some RECs to reach out to other RECs and experts in order to overcome ethical doubts and methodological questions raised during the evaluation of complex protocols. A second issue refers to the sponsor's lack of transparency and misuse of CT registries.
With rare exceptions, LA RECs do not have the ethical and clinical expertise to assess complex protocols. The Private Hospital REC is one exception, with three times the number of members required by national regulations in a country with a relatively solid regulatory agency,31 it has more technical resources and is less susceptible to conflicts of interests than the average REC. As a private hospital in a wealthy community, its decisions are less influenced by economic payments than those of commercial RECs, which in order to survive financially need to satisfy the pharmaceutical industry's requests, even at the expense of compromising ethical principles.32
The Private Hospital's REC felt the need to confirm whether they had reached the correct conclusion when they assessed a complex CT protocol. The lack of a clear and unified response from national and international ethics experts are by themselves indicative of the problem and lead us to emphasise the need for more collaboration and debate among RECs, and for the establishment of national RECs.
National RECs would have more resources and better-trained professionals to determine whether a protocol conforms to internationally accepted ethical principles. Institutional RECs would remain responsible for determining whether the trial is relevant for their population and should be conducted at their site. The Brazilian experience of the national REC (Comissão Nacional de Ética em Pesquisa (CONEP)) that reviews and approves, rejects or requests changes to the decisions made by institutional RECs appears to be positive,2 but needs to be independently evaluated and some of its well-known limitations, such as funding shortages, need to be resolved. National RECs in small-size and middle-size countries may encounter the same problems than institutional RECs. Independent international reference centres to assist RECs would also be a valuable resource.
A second ethical issue relates to the sponsor's lack of transparency, an ethical problem that is increasingly discussed in the scientific and mass media.33 ,34 Bayer denied access to information that could not be considered an industrial secret. Information requested to clarify ethical and methodological issues of the CT was not provided, the name of the PI in Spain was denied without good reason and the sponsor misled users of the US CT registry. Bayer's failure to use appropriate codes denied other RECs this opportunity to learn from others.
Use of appropriate codes could alert other RECs. Thus, if the Peruvian CT regulatory agency had been aware of the ethical concerns of Private Hospital, it might have objected to the implementation of the trial.
The National Library, a prestigious institution, explains that it does not have the resources to control the quality of the data entered into the registry and the deficiencies in the registry are recognised by those who operate it.35 ,36 The industry owns the data entered, but ownership brings with it responsibilities and should not be understood as a license to mislead. The National Library also has an ethical responsibility not to let its resources be used to misinform RECs and researchers.
The National Library could let users know that it is not responsible for the accuracy of the data and warn users that the data have not been verified. To facilitate communication between RECs and researchers, which is especially important in low-income and middle-income countries, the National Library could also demand that CT sponsors willing to register CTs include the information needed to contact the PIs and the chairs of the RECs of the participating sites.
The codes included in the registry to describe the implementation status of trials do not cover all possible scenarios. Given the close relationship between ethical and human rights violations of CT participants, the US registry needs at least an additional code, perhaps ‘rejected by REC’. Transparency requires that users of the registry know the reasons for the rejection even if only in broad terms such as ethical, methodological or other (i.e. administrative or technical capacity) are used. The codes withdrawn and terminated should also be accompanied by further explanations, such as ‘unable to recruit’ or ‘insufficient resources’.
Without additional resources, the National Library of Medicines might not be able to verify errors in the data entered by sponsors, but the regulatory agencies and the owners of registries could impose deterrent penalties in the event that unwarranted or inexcusable errors are found. The response given by the FDA when informed of the problem detected in the US registry and of the ethical concern of the Argentinean ethics committee showed indifference: ‘… there is currently no regulatory requirement… we would not object to be informed’, meaning that the FDA was not going to do much about it. This response is troubling. If CTs are inappropriately approved and conducted and if patients are unaware that they are part of a clinical experiment, they may not report adverse events and might concomitantly use other treatments and, consequently, the regulatory agencies would be approving new products based on faulty efficacy and safety information. In CTs, it is important to be aware of the possible links between ethics, safety and efficacy. If misleading data are entered into the US registry, the FDA could let the sponsor know that there would be a careful scrutiny when the request for market authorisation is made.
If CT registries continue to function in a suboptimal way, they might lead to reduced transparency in international clinical research, which is contrary to the intention behind establishing such registries and other information resources meant to assist RECs in their work.
Finally, one of the major reasons for outsourcing trials to low-income and middle-income countries is the inability to recruit participants in countries where most people can access needed medicines. The poor without access to pharmaceuticals are more willing to participate in CTs than more affluent groups,37 and in low-income and middle-income countries only 35% have access to needed medicines.38 The principle of justice is violated if most participants are recruited from among the poor who do not have to access drugs. If CT registries collected information on the number of patients to be recruited per country and/or some indicator of their educational or socioeconomic status was recorded, it would be easier to study whether the CT respects the principle of justice.
Contributors Both authors contributed equally to the research project and to drafting and finalising the submitted version of this article.
Competing interests None.
Provenance and peer review Not commissioned; externally peer reviewed.
Data sharing statement The information contained in this article has not been published elsewhere.
↵i In the USA, RECs are known as institutional review boards (IRBs).
↵iii Fictitious names have been used to protect the anonymity of institutions and individuals.
↵iv In Spain, the sponsor of a trial selects a leading site responsible for the ethical approval and its decision is binding for all Spanish sites, which can only withdraw for administrative reasons. It is well known that sponsors tend to select the sites with RECs that present fewer objections to protocols.
↵v We had been told that Belgium and India had decided not to participate on the study.
↵vi The only approved paediatric indications of fluoroquinolone are the treatment of pseudomona infection in children with cystic fibrosis, after anthrax exposure and for complicated infections—especially of the urinary tract—that are resistant to other antibiotics.23 The only other indications in which fluoroquinolones might present advantages (including faster recovery and better efficacy) over standard treatment are in cases of paediatric typhoid fever, severe dysentery due to shigella infection and enterococcal meningitis.24
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.