Abstract
To end 2017, we have prepared a special double issue covering articles broadly on the work of ethics committees, and on research involving virtual or otherwise vulnerable subjects.
Of particular note is the evaluation of an education and training programme in Canada, not only because I have periodically been involved in training research ethics committees (some readers, I am sure, will remember our often lively discussions), but also because the article’s results probably raise as many questions as they answer. Although some training is now often viewed as mandatory for new members of committees, it is still unclear what should be taught, how it should be relayed, and what markers, if any, demonstrate that it makes a positive difference. Meanwhile, in the background, there remains a debate over the role and remit of such committees, so we may not even have a clear idea of what to aim for. Trivially, ethics committees should approve only ethical research and reject only unethical research. In some quarters, committees are accredited based on some kind of performance, which could include the number of hours their members spend on training. Tempting though it is for regulators to defer to the ethics committees in cases requiring moral judgement, many members are uncomfortable with the lack of clear instruction and with the discretionary powers which come hand in hand with the task. However, the result reported by Stockley and colleagues suggests that the opposite may be true. Before the training programme, most respondents were satisfied with their background knowledge of research ethics, broadly construed, yet were not satisfied with their knowledge of the particular policy document, Tri-Council Policy Statement: Ethical Conduct for Research Involving Humans (TCPS2). When realizing that not even half the participants had read TCPS2, the reported lack of knowledge is not surprising. Yet it is surprising that most were highly satisfied with their background knowledge, and that knowledge seemed the most important dimension for the authors to try to measure. Perhaps the respondents simply needed to read TCPS2? Unfortunately, there is no reported level of satisfaction with knowledge of research ethics generally after the training. (Difficulties with response rates after training – not to be read as a reflection of the training itself – and changes in format of the survey meant that no direct pre- and post- could be made.) It would be plausible, however, to think that knowledge of different cases that were discussed at the training workshops (what moral issues were considered by them and how the committees resolved them), would increase the ability of participants to engage in the analogue reasoning which is impossible to capture in regulation, but which is so valuable in the practice of reviewing research and developing the sought-after expertise. As for reasoning from principles, after training, less than three-quarters of participants could recite the three core values listed in TCPS2, namely Respect for Persons, Concern for Welfare, and Justice, making reliance on deduction from external rules less certain than we might have assumed, however they were to be balanced in different cases. Once established, members of each committee are likely to establish values which are internal to the committee, and may set collective precedents to guide future decision-making. Without external review, quick turnover of members, or regular co-opting procedures, we should expect differences in opinion between committees to become pronounced. To keep such differences within reasonable bounds, members of research ethics committees in the UK are invited to join the Shared Ethics Debate (ShED), which encourages debate of complex cases between members of different committees. First introduced in order to gauge how (dis)similar the opinions of ethics committees really were, it is now used as much to measure ‘performance’ as it is to provide a platform for training.
In many areas of professional practice, pedagogy requires some form of critical self-reflection – a facility which may be largely ignored by ethics committees. They are almost never asked to reconsider the same application on appeal unless prompted by a formal complaint of some sort. A process of reflection need not be confined to confirming that opinions over time are sufficiently similar. If that were so, there may be scope for expert systems or artificial intelligence to check performance. Rather, collective values can be cultivated.
With a view to encouraging education rather than training as such, empirical studies on how committees work are worth reading. For example, Happo and colleagues report the range of queries ethics committees ask of applicants according to the type of study under review. Familiar concerns about the reliability of scientific methods seem to be greater in studies which do not involve physically invasive procedures. As the authors point out, their result could mean that members of ethics committees do not fully appreciate the social value of some study designs, for example qualitative studies. To provide some guidance on considering qualitative methods (in the absence of reviewing the methods first hand or applying widely recognized markers of methodological competence on the basis of which ethics committees can form a judgement of informed trust in the researcher), Mooney-Somers and Olsen outline points to consider. They helpfully provide practical guidance on using formal qualifications and training, explicit claims of competence, and markers of in/competence in the assessment of qualitative research competence. However, the experience of graduate students reported in the article by Petillion and colleagues shows that some work is needed for ethics committees to support students in their learning and form closer relationships with supervisors without adding any extra administrative burden. In the context of how social science is more generally perceived (and some claim it is currently undergoing some form of a methodological crisis following a series of studies lacking reproducible results or predictive power), Gomez and colleagues argue that more quality assurance and control measures in the social sciences should be introduced. Whether the ethics committee can and should have a role in so doing is another question.
The remaining four articles relate to issues of vulnerability, either because of the mental state of the proposed participants (either psychiatrically or developmentally) or because of the situation from which the participants are recruited (namely online social media). As additional risks are often perceived to be associated with such research by virtue of involving vulnerable participants, the following articles and case study address the difficult questions of how these vulnerabilities should be considered and how ethics committees can mitigate against the associated perceived risks.