IRBs

In 2003, I discussed the one-size-fits-all approach to both review and consent forms in a paper in JAMA. As an epidemiologist working on a multi-center genetic study, I found the regulations inappropriate, the reviewers were inexperienced with the field, and the consent forms were designed for a clinical trial. All of this promoted problematic variability in review and consent. In 2006, I wrote a paper suggesting that a virtual centralized IRB process would result in a review by individuals with expertise in the field of the study and in a consent form much more appropriate for the study. We need to discuss this in a forum which can be read by all researchers. Thus far, the humanities have been underrepresented in this type of literature so I suggest that you do ‘review the reviewers’ and submit the review to “IRB".

Rita McWilliams, Dr, at 10:40 am EST on January 19, 2007

IRBs — research misconduct

I did a talk on this topic in my own field (social work) this past year. My conclusions after a cursory review (e.g. not a full systematic review) of the literature regarding research misconduct and regulatory interventions (IRBs): — We apparently have no credible, published evidence that a problem exists. — We apparently have no credible, published evidence that current regulatory interventions produce statistically and practically significant effects in a cost effective manner. — We apparently have some evidence from other fields that current regulatory methods produce substantial negative effects.

As Stanley Milgram said many years ago: “erection of a superstructure of control [by the federal government] on sociopsychological experimentation is a very impressive solution to a non-problem”

The lack of evidence re: a problem raises a more revealing question. Why is the ethics business doing so well?

Gary Holden, Professor, at 11:25 am EST on January 19, 2007

confusion about IRBs

As someone who has recently looked into federal IRB regulations, I can say with great confidence that much of the information provided in this article is inaccurate, as is the information with which many IRBs seem to operate. For instance, this article starts out by saying “An IRB must approve any study that involves human subjects.” That’s what most people, and unfortunately most of those on IRBs, seem to think. But its wrong. Federal regulations (see the OHRP web site) state that only human subjects *research* needs to be reviewed, and the regulations specifically define research as a *systematic* investigation that is *designed* to create (i.e., doesn’t just happen to create) generalizable knowledge about people). The OHRP has gone on to clarify in further documents, including in a 2003-ish letter to the Oral History Association that substantially approved the Association’s new policy on research ethics, that it does not regard most oral histories as systematic investigations that are specifically designed to produce generalizable knowledge. These documents further indicate that such knowledge is generally only produced when members of classes of people (e.g., males and females) are studied as representatives of such classes, as is done in statistical, but not ethnographic, research. Thus, the OHRP did *not* contradict its own edicts when it did not seek institutional review of its oral history project. Unfortunately, the OHRP has not circulated these documents very widely or posted them very prominently on its site. I could only find one in a Google search I just conducted while writing this, and only happened upon them accidentally the first time I found them.

Z, at 11:35 am EST on January 19, 2007

IRBs

Perhaps I’m more fortunate than most, but I’m currently having a very positive experience with my IRB. In fact, they’ve been very helpful in helping me design my project and think through how I want to use my data. The OHRP guidelines aren’t difficult to find — they are right on their website under human subjects (here’s the link, http://www.hhs.gov/ohrp/humansubjects/guidance/decisioncharts.htm)

I would add that as a historian of medicine, I think that while these rules are cumbersome they do serve an important purpose.

Heather Munro Prescott, Professor of History at Central CT State University, at 1:20 pm EST on January 19, 2007

IRB related mortality

I noted the lack of evidence re: research misconduct in social work above. In terms of the effect of IRBs the following is worthy of consideration.

Collins, Doll and Peto (1992) stated regarding the ISIS-2 trial of streptokinase in acute myocardial infarction that was carried out in the UK and in the US that:

“If the USA had recruited as fast as the UK then the trial would have ended six months earlier, and since the eventual ISIS-2 results have transformed medical practice. . . that six-month delay means that about 10,000 unnecessary deaths are directly due to whatever it was that slowed recruitment into ISIS-2 in the USA. Apart from consent procedures, all other aspects of the study were equivalent. It seems likely, therefore, that the US consent procedures were directly responsible for at least a few thousand unnecessary deaths, in the pursuit of “informed consent” that many doctors would consider inhumane” (p. 54)

Do the proponents of IRB intrusion into the research process and the diversion of research dollars to this unproven regulatory oversight think that such human tragedy is a justifiable outcome of their approach?

Collins, Doll and Peto (1992). Ethics of clinical trials. In C. J. Williams (Ed.) Introducing new treatments for cancer: Practical, ethical and legal problems. John Wiley & Sons.

Gary Holden, Professor, at 5:05 pm EST on January 19, 2007

IRBs

As a behaviorist/clinician who happens to chair a biomedical IRB, I’ve seen both side of this argument, and to put it bluntly, I am ashamed of the position taken by many of my colleagues in their diatribe against IRBs. I’ve seen an incredible amount of bad behavior from my own field- no wonder the “hard” scientists look down on us. The IRB members are your peers; if not,then complain about the membership— it should include a sufficient variety of members to provide a fair review and should include members who understand how research works in the area being reviewed. You can help this situation yourself by volunteering to serve on the IRB. The IRB members should know the regulations (if your IRB thinks that a historical record is always “research” under the federal definition, they have not been reading the regulations) and those stupid tutorials are to help you understand as well. Granted, many of the tutorials are too heavily weighted toward biomedical concerns, but if you are smart enough to earn a PhD, you should be able to figure out how the core ethical principles translate to your field. If your study does require review, present a well written study and come prepared to answer questions in an intelligent and respectful manner. I rarely see a well designed study with a clear rationale for why the investigator wants to conduct the study a certain way get turned down. The “turn downs’ or delays are usually because the board could not figure out what the investigator was doing and on a few, luckily rare occasions, because the investigator approached the board in an adversarial manner. Again the IRB members are your peers. They don’t get paid. If you are rude and condescending, or waste their time with badly written material, they tend to react in kind. To poster CW; apparently neither you nor your IRB understand what the regulations allow or how to apply Belmont principles in your study. The regulations DO allow for waiver of documentation of informed consent when the written consent itself poses risk of harm, and, ethically speaking, while you are bound to obtain informed consent from your subjects, you are also bound to do so- and collect your data, in a way that protects your subjects from retaliation— an all too real possibility when working in a totalitarian state. Finally; to you sociologists and historians who don’t want to deal with the IRB at all, are you aware that if you use the IRB “system’, there is a device called a “certificate of confidentiality” that you can obtain to give you legal protection so that you can make your tapes, store them, and guarantee your subjects— and yourself—that you cannot be forced to divulge the contents of the tape?

Jerry, at 8:45 am EST on January 20, 2007

Jerry:"I’ve seen an incredible amount of bad behavior from my own field- no wonder the “hard” scientists look down on us.”. . . . “I rarely see a well designed study with a clear rationale for why the investigator wants to conduct the study a certain way get turned down.”

Heather"what if the drug in question had caused some patients to have heart attacks? Or caused birth defects when given to pregnant women? Or caused vaginal cancer in the daughters of women given the drug during pregnancy? There are numerous examples of why this “intrusion” is needed.”

These quotes from Jerry and Heather are typical of discussions I have been involved in on this topic. Unsupported assertions or tangents of negative possibilities usually unrelated to the specific issue being discussed. I presented a specific example of IRB related mortality and the reference above. No one responded to it.

Gary Holden, at 10:55 am EST on January 20, 2007

Sorry, I should add that I was granted approval for a verbal consent form for a survey, but not for interviews.

CW, doctoral candidate at California, at 9:35 pm EST on January 21, 2007

Jerry:

A quick reply. I don’t have time to read your book at the moment- but it would seem fair to assume you ‘hit me with your best shot’. Two points:

a. Deception. To the best of my knowledge – deception has always been acceptable in ethical research if there is an appropriate assessment of short and long term / subject v. societal risks and rewards, combined with immediate post-study debriefing and provision of psychological support if desired by the subject. There are instances in which deception during the course of the study is required if clear results are to be obtained. Without deception the world would have been denied such important work as the Milgram obedience studies – perhaps the most valuable behavioral science research of all time.

b. My connection to my library server is down so couldn’t get to the COST study. Your explanation of the problem is bit unclear. You seem to be saying that the study surgeons exposed subjects to undue risk. I can not tell if this risk was too high without understanding the full context.

What I find curious is that I present a case (BTW: with a clear reference) regarding the possibility that as many as 10,000 people in the US died as a result of overzealous IRB intrusion into a study.

You respond with a sweeping unsupported assertion – “Anyhow, the book details numerous situations like this, affecting hundreds of thousands of research subjects (such as most children with cancer—this issue is discussed in detail in the final chapter, with real studies examined)” – which merely proves my point about the type of arguments often marshaled in these discussions.

I am not an expert in this area – by any stretch. But the argument seems to boil down to a very simple idea.

Does some (yet to be determined) number of individuals experiencing a lack of truly informed consent (a goal that may not be obtainable) and thereby perhaps experiencing greater physiological or psychological risk, justify thousands of deaths (who knows how many more may have died as a result of IRB intrusion beyond the ISIS-2 study)?

In terms of relaxed rules for social science research, I repeat my prior point. My conclusions after a cursory review of the literature regarding research misconduct and regulatory interventions (IRBs): We apparently have no credible, published evidence that a problem exists; we apparently have no credible, published evidence that current regulatory interventions produce statistically and practically significant effects in a cost effective manner; and, we apparently have some evidence from other fields that current regulatory methods produce substantial negative effects.

Can society afford to waste precious resources on an unproven intervention, that seeks to solve a problem that has not been proven to exist, and appears to produce iatrogenic effects?

Gary Holden, Professor, at 8:15 pm EST on January 22, 2007

As someone who has been an IRB member and administrator, I recommend that those frustrated with the process explore any or all of the following options:

1. Talk with your IRB administrator and/or Chair. Most are knowledgeable and try very hard to facilitate the review process, often with insufficient resources and within a regulatory framework that can be less than flexible. 2. As noted by previous posters, if you think your IRB is clueless about your area, get on the board or volunteer to serve as a consultant. IRBs depend on the expertise of their members. If neither you nor anyone else in your area are willing to serve or provide consultation, don’t expect a smooth review process. 3. For those conducting cross-cultural/international research, many IRBs use consultants to help evaluate such studies and assess local context. If your board doesn’t do this, discuss this option with the Chair or IRB administrator. 4. If you’re getting nowhere with your Chair or IRB administrator, contact the institutional official at your institution, who has ultimate responsibility for the IRB. IOs can be helpful in addressing problems, particularly at higher administrative levels. 5. If you hate your required training because of a biomedical focus, keep in mind that there are several training modules available for social-behavioral sciences that institutions can use, including the one offered through CITI. The Office of Research Integrity also has training resources that do not have a biomedical focus. Talk to your IRB or IO about using one of these other training modules. 6. Hate the regulations? Slamming your local IRB is of very limited utility in changing the regulations. Whether through your professional organizations or independently, try contacting SACHRP, the advisory committee for the Office of Human Research Protections. Members of SACHRP come from various backgrounds and are very attuned to the difficulties posed by the regulations to both IRBs and researchers.

Edith, at 12:16 pm EST on January 23, 2007

Over-regulation of human subjects research has been implicated in the deaths of humans.The zeal of regulators and their aspirants is never justified when such events occur. If this moral position does not persuade, then we occupy different moral universes and should probably stop wasting each other’s time.

I remain hopeful that some of the bystanders to these exchanges will reconsider IRB regulations.

Gary Holden, Professor, at 5:25 am EST on January 24, 2007