Evidence-Based Research Ethics and Determinations of “Engagement in Research”

Developing and implementing human subject protections should be based on data when possible. This is the rationale for the imperative to conduct empirical research in bioethics.^1,2 For example, knowledge of research subjects’ experiences, opinions, and perceptions can inform the way research studies and their informed consent processes are designed in order to optimize the protection of human subjects. This kind of research involves studying actual subjects in research studies. Such ethics research is itself human subjects research and raises the question of how IRBs implement the federal regulations and guidance provided by the Office of Human Research Protection (OHRP) in overseeing such research. This topic is particularly relevant to IRBs, bioethicists, and others who are committed to improving the practice of research ethics oversight.

In this brief article, we focus on how IRBs deal with the issue of what constitutes “engagement in research” in cases where staff from a clinical research study (which we will call “parent studies,” to distinguish them from the ethics research that focuses on the participants of those studies) take steps to help an ethics study team recruit subjects. Our experience indicates that IRBs’ interpretations of the guidance from the OHRP could have significant implications for the quality of ethics research.

IRB interpretations of ‘engagement in research’

The subjects of our ethics interview studies are actual participants in (and decliners of) neurosurgical and pharmacological randomized clinical trials. Our interviews assess subjects’ motivations for participating in research and their understanding and attitudes regarding various aspects of the clinical trials. Recruiting such subjects involves considerable time and effort. Most sponsors and principal investigators appreciate the importance of the ethical questions we propose to study and agree to allow us access to their subjects. However, even when we obtain such agreements from study sponsors and principal investigators, the individual study sites are often wary of potential additional burdens, given the demands of their own recruitment and other study efforts. Thus, recruitment of our subjects requires negotiations and sensitivity in working with the parent study sites. This issue is crucial because effective recruitment is key to gathering generalizable findings. There can be dozens of sites for a clinical trial, and given the potential for significant burdens on sites due to varied and extensive lengths of time needed to coordinate IRB reviews,^3,4 there is a need for a regulation-compliant yet efficient method for working with numerous sites.

Over the years, we have developed regulation-compliant procedures to maximize recruitment without unnecessarily burdening the parent study sites. Research subjects can directly provide most of the relevant information for our purposes, so there is little reason to request any subject specific data from the parent studies. Therefore, there is no need to involve the parent study sites beyond the initial access to the subjects for recruitment. For such situations, the OHRP has provided clear guidance on how such access to the subjects can be conducted without creating the need for additional IRB reviews at the parent study sites. According to the OHRP, the following type of involvement “would make an institution not engaged in human subjects research”:

Institutions whose employees or agents:

1. inform prospective subjects about the availability of the research;

2. provide prospective subjects with information about the research (which may include a copy of the relevant informed consent document and other IRB approved materials) but do not obtain subjects’ consent for the research or act as representatives of the investigators;

3. provide prospective subjects with information about contacting investigators for information or enrollment; and/or

4. seek or obtain the prospective subjects’ permission for investigators to contact them.

An example of this would be a clinician who provides patients with literature about a research study at another institution, including a copy of the informed consent document, and obtains permission from the patient to provide the patient’s name and telephone number to investigators. [Guidance section III.B.4, italics added]⁵

In our studies, staff at the parent study sites approach their subjects and briefly inform them of our interview study. If the subject is interested in speaking with us, the site obtains permission from the subject to forward the subject’s contact information to us. This process is documented by the site on a form and forwarded to our research coordinator. No other information (e.g., study data, clinical data, etc.) is shared with our team. Our own institutions’ IRBs accepted the OHRP guidance and agreed that the sites providing access to subjects using the above procedures were not engaged in our ethics research study. We also subsequently contacted OHRP directly to confirm that our procedures conformed to the conditions of the guidance.

We have found, however, that the parent study sites are generally not familiar with the OHRP guidance and are wary of unilaterally determining that their involvement in our research (e.g. providing access to subjects) does not require IRB review. We ask the parent sites to discuss the OHRP guidance and its application to our study with their IRBs and obtain their opinion about whether an IRB review at their sites is necessary; and we provide each parent study site with a written explanation of our procedures as well as a copy of the OHRP guidance to aid their interactions with their IRBs. Since the purpose of contacting the IRBs is to determine whether a formal IRB review is necessary, these interactions with the IRBs are less formal than the usual IRB review. For example, we spoke informally with the staff of IRB offices at our own institutions to confirm that they agreed with the OHRP guidance.

The Table summarizes 26 interactions of our parent study sites and their IRBs regarding the level of IRB review, if any, required for our ethics study.

Table.

Summary of 26 interactions between parent study sites and their IRBs about ‘engagement in research.’

	IRB agreed that parent study not engaged in research	IRB determined review required	IRB allowed opt-in letters only
Neurosurgical RCT 1	1	4	2
Neurosurgical RCT 2	4	0	1
Pharmacological RCT	9	3	2
Total	14	7	5

In 14 instances, the IRBs agreed that the parent study sites are not engaged in research, as defined by the OHRP’s guidance. But in 12 cases, the IRBs told the parent study staff that review by the site’s IRB was needed. However, in 5 of these 12 instances the IRBs offered an alternative to an IRB review: the parent study staff could send an opt-in letter to the subjects with our study’s contact information, rather than releasing to us the subjects’ names and contact information with the subjects’ permission. All of these 5 parent study sites stated to us that they were not staffed to bear the burden of shepherding a new application through their IRBs and they were willing to help our ethics study only if we accepted this “opt in” letter process.

In 7 cases, the IRBs required IRB review and did not offer an alternative. 2 of these parent study sites declined to help us recruit. In both cases, their IRBs required amendments to their study protocols for review and the sites declined to amend their protocol. In 5 of the 7 cases, the parent study sites submitted and received IRB approval for our ethics study (1 of these was submitted as an amendment to their clinical trial).

Of note, we attempted to recruit individuals who declined participation in the parent studies. Some of the IRBs raised concerns about this. These IRBs were concerned about the procedure of the parent study staff approaching decliners of the parent study and informing them of our ethics study and, if the decliners agreed, passing their contact information along to us. One IRB determined that only an opt-in process was acceptable for decliners. Another IRB allowed the site to identify and approach decliners in a prospective fashion only (i.e., as they declined in real time), but not individuals who had already declined. Another IRB prohibited recruitment of any decliners.

Discussion

This report raises a seldom discussed issue in research ethics: How well is the current oversight of empirical studies in research ethics functioning? Our experience raises concerns that the current institutional practices could in the long run unnecessarily compromise the quality of research studies that are aimed at improving the human subjects’ protections system.

Although the majority of the IRBs agreed with the OHRP guidance, in 12 of 26 instances the IRB responses were more restrictive. It seems reasonable that an agency’s specific guidance should, barring special considerations, be followed. There may have been local, administrative reasons for not following the guidance, just as some IRBs decline to exempt any studies as a matter of local policy “to avoid confusion.”⁶ But from an ethical point of view, there are good reasons to doubt that the more restrictive practice is necessary.

First, following the OHRP guidance does not mean that there is no IRB oversight of the ethics research study. The ethics study must still be formally reviewed by an IRB (or multiple IRBs, if the ethics investigators are from multiple institutions). Second, there is little likelihood of exploitation of or harm to the subjects. All of our subjects were competent adults who would have given specific, documented permission to pass along their names and contact information to us. The risk in being told about our ethics study and having contact information forwarded to us is virtually nil. No breach of confidentiality is involved (since permission is given to pass along the information).

Third, in regard to decliners of research, the IRBs were even more restrictive. Data from decliners may be particularly important because they help provide a more complete picture of what potential subjects understand and how they make their decisions. Thus, there should be a countervailing ethical consideration to justify special restrictions on interviewing decliners of research. Here, the IRBs may have reasoned that because the subjects had already declined to participate in the clinical study, to be asked again about participation in another study would be an ethical infringement. One of the reviewers of this paper pointed out that the decliners “may view it as an infringement of their autonomy to be asked to participate in what could be perceived as a further stage in the same research. The principle of respect for persons can be interpreted to support this position.” The counterargument for this moral intuition would be that it seems unreasonable to assume that simply because a person declines to participate in a particular clinical trial, he or she should not be contacted further for other research; since decliners of the clinical trials we studied had already shown interest in participating in research (although ultimately declining), it is unlikely that they would have wanted to be protected in this way.

Which moral intuition regarding the attitudes of decliners is correct is an answerable empirical question. A definitive answer could be provided by conducting an empirical research study interviewing decliners of clinical trials. However, if we restrict such research, we will never know which moral intuition is correct and IRB decisions would continue to rely on plausible, yet competing, intuitions rather than on data.

Some may doubt that our experience adequately reflects the practices of IRBs by pointing out some of the limitations of our report. First, we relied on the reports of parent study site staff about their communications with their IRBs. It is possible, for example, that the parent study site staff, being unfamiliar with the issue of “engagement in research,” may not have accurately presented the issue to their IRBs. Thus our experience may not reflect “considered judgments” of IRBs. However, we did provide documentation to the parent study site staff to use in their interactions with their IRBs. Further, none of the 5 sites’ IRBs that conducted a review of the protocol—which would have given them ample opportunities and information to determine that the parent site was not “engaged” in research according to OHRP guidance—determined later that such a review was unnecessary. Also, those IRBs that offered a choice between an opt-in letter versus an official IRB review obviously understood that an official IRB review was not necessary (since they were willing to accept an opt-in letter option over a formal review).

Second, the anecdotal nature of our report may limit the generalizability of our findings. This is undoubtedly true. However, the incidence of IRB practices being more restrictive than OHRP guidance was not just one or two cases but 12 of 26 cases. Also, these were highly experienced IRBs, from well-known academic medical research centers and hospitals. Thus, although our report cannot be relied upon to provide precise quantitative estimates, an overall impressionistic conclusion that IRBs often impose conditions more restrictive than the OHRP guidance is warranted.

Empirical research in research ethics is crucial to developing evidence-based policies and practices to protect research subjects. How IRBs oversee such studies has particular salience for ethicists, regulators, and human subjects protection systems. Our experience with the specific issue of interpreting what constitutes “engagement in research” suggests that practices more restrictive than OHRP guidance are not unusual. From an ethical point of view, there does not seem to be a compelling reason to override the OHRP guidance on this issue. And the long term consequences for the human subject protections systems could be significant, since obstacles to gathering high quality data on the practices and policies we use to oversee human subjects research would effectively cut off an important feedback loop to improve the system.