Ethics Review Committee approval and informed consent: an analysis of biomedical publications originating from Sri Lanka

Background International guidelines on research have focused on protecting research participants. Ethical Research Committee (ERC) approval and informed consent are the cornerstones. Externally sponsored research requires approval through ethical review in both the host and the sponsoring country. This study aimed to determine to what extent ERC approval and informed consent procedures are documented in locally and internationally published human subject research carried out in Sri Lanka. Methods We obtained ERC approval in Sri Lanka and the United Kingdom. Theses from 1985 to 2005 available at the Postgraduate Institute of Medicine (PGIM) library affiliated to the University of Colombo were scrutinised using checklists agreed in consultation with senior research collaborators. A Medline search was carried out with MeSH major and minor heading 'Sri Lanka' as the search term for international publications originating in Sri Lanka during 1999 to 2004. All research publications from CMJ during 1999 to 2005 were also scrutinized. Results Of 291 theses, 34% documented ERC approvals and 61% documented obtaining consent. From the international journal survey, 250 publications originated from Sri Lanka of which only 79 full text original research publications could be accessed electronically. Of these 38% documented ERC approval and 39% documented obtaining consent. In the Ceylon Medical Journal 36% documented ERC approval and 37% documented obtaining consent. Conclusion Only one third of the publications scrutinized recorded ERC approval and procurement of informed consent. However, there is a positive trend in documenting these ethical requirements in local postgraduate research and in the local medical journal.


Background
Guidelines on ethical conduct of research have focused attention on protecting research participants [1][2][3]. Approval from an ethical review committee (ERC) [1,2,4] and ensuring informed consent from the participants are the cornerstones of ethical human subject research [3,5,6]. Consent is considered 'informed' when given by a person who understands the purpose and the nature of research, what is required from the participant and what may be the potential benefits and risks resulting from the study [2,7]. In order to ensure acceptable ethical standards in externally sponsored research, protocols should undergo ethical review in both the host and the sponsoring country [1,2]. However, research carried out in the developing world which does not comply with accepted ethical standards is a concern for academics, funding agencies, journal editors and governments [3].
Proof of ERC approval and consent of participants can be sought by direct and indirect methods. Inquiring from the relevant ERC or obtaining a copy of the approval is a direct method. Inquiring from the researcher or scrutinising published research for documentation is an indirect method. For informed consent, witnessing the informed consent process is the direct method whereas inquiring from the researcher, research participant, ERC or establishing evidence from documentation in published research are indirect methods.
To our knowledge there are no studies reported from South Asia examining evidence of adherence to these two requirements in published research. However, studies in South Asia have examined the issue of consent from the point of view of participants and researchers. A study carried out in Bangladesh examined participants' understanding of a community-based study of iron supplementation [8]. Although consent had been obtained after detailed explanation of the study, many participants did not understand that they were free to decline to participate, that they could choose to leave the study, and about half believed that participation was part of routine health care [8]. In another questionnaire survey among researchers in the developing world, the majority (62%) stated that they had obtained written informed consent [9]. Several studies have also looked into informed consent practices relevant to patient care and clinical ethics from the perspectives of physicians and patients [10][11][12].
Although Sri Lanka is a developing country with a population of 19 million, it has strengths in health and education [13] and the tradition of biomedicine, both western and eastern, is firmly established. There is archeological evidence of hospitals dating back to the 9 th century AD in the ancient kingdom of Anuradhapura [14]. The civil medical department was established by the British in 1858 and the Colombo Medical School, which opened in 1870, is the second oldest in south Asia [14]. The Ceylon Medical Journal (CMJ) is the oldest surviving English lan-guage medical journal in Australasia, first published as the Journal of the Ceylon Branch of the British Medical Journal in 1887 [14].
Sri Lanka has a long history of scientific research. Aldo Castellani, under British rule, established the Medical Research Institute (MRI) in 1901 [14]. Influential international research collaborations have been reported [15][16][17][18][19][20][21] and leading funding agencies, including the Wellcome Trust (UK), National Institute of Health (USA) and the World Bank have funded research in Sri Lanka. A considerable amount of anthropological and sociological research has also been conducted [22].
In spite of the volume of research, bioethics is at an early stage of development in Sri Lanka [23]. There are no formal training courses for clinical researchers except for some limited teaching in the six medical faculties from which the majority of research emanates [24]. Institute for Research and Development has conducted a basic course in 2003 and a basic and advanced course in 2007.
In this study we draw attention to the issue of informed consent and ERC approval for human subject research in contemporary Sri Lanka. The International Committee of Medical Journal Editors (ICMJE) guidelines of 1981 required researchers to indicate that the research had IRB review [25] and, in 1991, the ICMJE added that when "informed consent has been obtained by authors, this should be clearly stated in the article" [26]. This study aimed to determine the extent to which ERC approval and informed consent procedures were documented in three independent sources of publications arising from human subject research carried out in Sri Lanka.

Methods
The protocol was revised in accordance with the recommendations of reviewers of the funding body. Consensus generation meetings to fine tune the protocol were conducted with the participation of the authors and those who were trained during the ethics course and qualitative workshops conducted by the Institute for Research and Development [27]. 2. A Medline search was carried out with MeSH major and minor heading using 'Sri Lanka' as the search term to find all published research originating from Sri Lanka, between 1 st January 1999 and 1 st September 2004. We downloaded full text papers published in open access journals and those available in Athens, an access management system of academic articles. If the full text papers were not available we downloaded the abstracts. We could not obtain the full text paper copies of those abstracts as the printed journals were not available in Sri Lanka. These publications were initially scrutinized by two authors (ML and MA). Later AS, SS, CS re-scrutinized the papers to select and agree on human subject research publications. Then AS and SS independently scrutinised the papers to identify • any information related to ERC approval in the published work • documentation on the type of consent obtained (written/oral) 3. For CMJ we analyzed the full text papers available as hard copies for the period 1999-2005. This was due to two reasons; it is the only Sri Lankan journal available in the PubMed as abstracts and we had access to hard copies. CMJ publications were initially scrutinized by two authors (ML and MA). Later CS and SH re-scrutinized the papers to select human subject research publications and to identify whether ERC approval and consent was documented. However, only three (1%) information leaflets and 23 (7.5%) consent forms were available from 291 theses.

Medline search results
The We attempted to find out how many of these papers had authors of Sri Lankan origin but it was not feasible from the information available in the papers.   There is no reference to ERC approval or consent in PGIM examination rules and regulations. However each board of study is responsible for ensuring these standards. Board of study in community medicine has made it a requirement that candidates obtain ERC approval (personal communication from the chairman) since 1991. Regulations and guidelines of the board of study in family medicine state that research protocols should include details of approval of the project by a relevant ERC. We were unable to find information about requirements for ERC approval in other disciplines. Instructions to authors in CMJ [28] state that ethical committee approval should be mentioned in the text in 'intervention studies' with a photocopy of the approval letter attached to all submissions.

CMJ search results
Only 38% of full text papers from 1999 to 2004 in international journals had documented ERC approval. This is a cause for concern. In CMJ, ERC approval for the same period was 32.5%. During the same period 39% of papers in international journals and 30.1% in CMJ had documented informed consent.
However, absence of documentation of ERC approval cannot be taken as absence of obtaining approval, as some may have not reported it. Further research should include direct inquiries to authors and journal editors on ERC approval for published research.
Even though the International Committee of Medical Journal Editors requires ERC approval and informed consent to be documented in the manuscripts [25,26], our results shows that this requirement has not been strictly adhered to. Previous studies that had examined research carried out in the developed world have also shown that documentation of IRB review and IC is not consistent, even in journals that state it as a requirement [29][30][31][32][33][34]

Limitations
This study relies on self reported data. Therefore, the content of the ethical review, the process of informed consent, and the level of understanding of research subjects, both of the research, the consent process, and the meaning of informed consent are not explicitly stated. This affects all three surveys. In addition the main limitation in the international journal survey is the inability to review full publications on human subject research as only the abstracts were avaialble. Given space limitations, it is reasonable to anticipate that ethical review would not always be stated in the abstract.
In all three surveys, it is possible that more researchers may have sought ERC approval than reported. Evaluation based on documentation by the researchers as opposed to actual ERC approval and consent is an inherent weakness of the study methodology.

Recommendations
Researchers should be requested to attach the information leaflet and consent forms to the final version of their theses, in the same way they are required to append research instruments. This will enable the examiners to assess adherence to good ethical practices and encourage new researchers to regard ethical standards as an integral part of high quality research. Similarly journal editors should implement ICMJ recommendation more robustly.

Conclusion
Documentation of ERC approval and consent occurs in around 35% of research publications in international journals and CMJ. Even though there is a positive trend towards documentation of ERC approval and informed consent in human subject research in Sri Lanka, the prevailing situation is not satisfactory. Increasing awareness on these two crucial safeguards; ERC approval and informed consent, will be helpful locally to promote basic ethical standards as essential components of scientific research. This may also help adherence to higher ethical standards in international collaborations.