As a lawyer and long-time member of a research ethics board (REB) in Vancouver, BC, I read with interest the article by Kotecha et al1 and Dr Upshur’s well-thought-out response.2 It might be helpful to researchers to at least know why proposals are questioned by REBs and why there can be delays before approval. This letter reflects my personal opinions and not necessarily those of my REB, and is based on Kotecha and colleagues’ article rather than the original application and supporting material.
I sympathize with Kotecha et al1 with regard to delays based on infrequent meetings of an REB. Hours are spent by the responsible board members in scrutinizing their assigned research proposals for reporting to the board. Then the board spends more time discussing the research proposals. Research ethics boards are usually made up of volunteers, although sometimes a modest honorarium is paid, resulting in a rate per hour equivalent to minimum wage. As Dr Upshur points out, there is difficulty in getting experienced physicians involved in research to sit on REBs. If the expertise to review the proposed research cannot be found on the board, then the study is usually referred out to an expert, which adds to the delay.
A starting point in research, as in clinical medicine, is that there must be respect for the subject. Mason and Laurie said in talking about consent, “It must be remembered, however, that consent itself is a means to an end and that the real aim is to respect persons and their interests.”3 The second edition of Canada’s Tri-Council Policy Statement (TCPS2) describes “respect for human dignity” as the “underlying value” of the TCPS “since its inception.”4 When the TCPS2 talks about “respect for persons” it includes those subjects whose data are “used in research.”4 This requirement for respect is sometimes forgotten or is only paid lip service by researchers. The primary obligation of the REB in both law and ethics is to the subject. That obligation requires that the proposed research be vetted, from whether or not it is flawed in any way to whether it satisfies both the legal and ethical requirements of respect for the subject; simply satisfying the legal requirements might not satisfy the ethical requirements.
I believe Kotecha et al have their quarrel with the REBs based on what they refer to as “inconsistent interpretations and application of privacy and ethical issues”1 (certainly there will be differences in the privacy legislation between provinces). Let me take a few of Kotecha and colleagues’ points and explain my concerns, which I expect reflect those of some of the other REBs and why researchers will have problems with many REBs if they fail to fairly address these issues before submitting their applications.
Kotecha et al take comfort that the health information of the patient is “de-identified.”1 The TCPS2 says that there are risks of re-identification and both researchers and REBs “should be vigilant in their efforts to recognize and reduce these risks.”4 So, REBs should not simply accept the “de-identified” information as enough to satisfy them but should look at all aspects of the information. In the case of Kotecha et al, the database contains the “first 3 digits of [patients’] postal codes” together with the de-identified health information.1 The TCPS2 identifies risk where a data site “contains information about a population of small geographical area or about individuals with unique characteristics.”4 So my concern on looking at the application for this trial would be whether, given the particular disease, the description of the patient’s particular health information, and the partial postal code, a researcher with that limited information could link it to a particular patient. I would want protection and assurances in place that this could not happen.
I note the assigned patient number “is copied on to a CPCSSN [Canadian Primary Care Sentinel Surveillance Network] key,” which contains all the patient’s personal and identifying information.1 It would appear that CPCSSN at some point has that identifying information, unless the information is copied by the physician who is dealing with the patient (the article does not specify, other than making reference to the key residing at the office of the physician or what is referred to as the “local EMR server”). This would also give me concern as an REB member, and I would want assurances that the database workers and the researchers would not have access to the code and to the local EMR server. The TCPS2 in its definition of coded information recognizes that identifiers are removed from the information and replaced with a code, but depending on “access to the code it may be possible to re-identify specific participants.”4 This is particularly so in regard to linkage to other data sets. I would want to know why the physician looking after the patient could not put the identifying number on the information, removing the patient’s identifiers before it being sent to the database.
When a researcher says, “There is always a trade-off between utility and security, therefore, and a small risk of identification,”1 my antennae go up. Ethically, why should the subject’s confidentiality ever be put at risk when the subject has not consented to the information being released? To me this quote does not show respect for the subject.
What would also concern me is that the physician dealing with the patient displays “information to inform patients about the CPCSSN project and to explain that patients’ information will be included unless they request otherwise.”1 This tells me that the physicians can try to get informed consent from their patients. It is therefore up to the researcher to satisfy the REB that this will not result “in a biased local sample data set not sufficiently representative for the purpose of health surveillance research”1 (as required by Article 3.7[c] of TCPS24). A simple statement by the researcher that this happens is not sufficient; I would want details. Remember, the basic requirement in research is respect for the subject. If the REB is satisfied that it is appropriate to waive the requirement of consent, assurances should be made that the physician talked to each patient about the research and risks (and did not rely solely on the printed material), answered any questions the patient might have had, and made it clear that the patient could opt out without it adversely affecting the relationship between the patient and the physician. I expect that the REB that originally required informed consent based that requirement on the particular circumstances of what is disclosed in the first quote in this paragraph.
I agree with Dr Upshur that researchers should serve on REBs.2 They would better understand the mandate of the REB to protect the subject, and from that understanding would be better able to draw up research proposals that would satisfy an REB. The concerns I have raised are all related to respect for the subject and can be addressed by the researchers in preparing their research proposals. If these concerns are not addressed, there will be delays in the approval of those proposals.
Kotecha et al want a “specialized, national, centralized” REB for “multisite studies related to population health research,”1 despite the varied legal requirements across provinces. Be careful what you wish for. If such a board were created and were truly independent (and met often enough to avoid undue delay), based on the mandate required of the board by TCPS2, it would likely have the same or more requirements and might pose the same questions some of the local REBs did in regard to Kotecha and colleagues’ application.
Footnotes
-
Competing interests
Mr Sullivan is a member of the University of British Columbia’s Clinical Research Ethics Board.
- Copyright© the College of Family Physicians of Canada