P.Mean: IRB approval of studies with less than adequate research integrity (created 2008-11-14).

This page is moving to a new website.

How should an  IRB view its job with respect to research approval if the study design is "less than adequate" to provide valid results? Is the IRBs job only to assess for patient risk? If a study's design is poor and will not yield useful results should the IRB approve the study if there is minimal risk? Does it matter if the study is minimal risk vs greater than minimal risk?

In many research studies, volunteers will cite the money as a primary motivator, but another common reason cited is the desire to help other patients who have the same disease that they do. If a study has a flawed research design, then you are violating an implicit assumption that the volunteers are making when they join a study.

That being said, pretty much every study design is "less than adequate." A study that requires informed consent, for example, is less than adequate because the process of informed consent produces selection bias. You obviously can't demand perfection. Two standards that I use are

1. Is the study design good enough so there is a reasonable probability of publication? Don't demand a level of rigor sufficient to publish in an "A" journal of course.
2. Is the study design good enough so there is a reasonable probability of inclusion in a systematic overview/meta-analysis.

If it fails to meet those standards, then you are subjecting volunteers to an amount of inconvenience, at a minimum, and possibly some level of pain and discomfort, and maybe even an increase in risk, without any prospect of benefits for future patients.

Both of those standards require some degree of speculation, of course, and a familiarity with the standards used by peer-review journals and meta-analysts.

If the study is good enough in your opinion to get published or included in a meta-analysis, then certainly offer any comments that might improve the research design, but don't consider failure to make those improvements a deal breaker. Often in research, you have to make compromises between the ideals of the scientific method and the pragmatic constraints of the real world.

An important question is whether you should hold a study with greater than minimal risk to a higher research integrity standard than other studies. I would argue that your job is already hard enough. Adopt a pass-fail standard. Is the research design so fatally flawed that it is unpublishable? If the answer is "yes" then it doesn't matter what the risk to the patients is. If the answer is "no" then I would not demand "A" level research integrity even for a greater than minimal risk study. In fact, the real-world constraints that might force you to approve a study with greater than minimal risk might be the same constraints that prevent you from adopting an "A" level of research integrity

A competing question is whether a study with zero risk should be allowed to have a zero level of research integrity. The answer to that is obviously "no" because volunteers still suffer a level of inconvenience when they participate in a study, and we have to respect their time and effort. I'm all for letting volunteers think for themselves, but if you have to word an informed consent statement with something like "this study is unlikely to ever get published and is being done just to help a medical fellow meet a regulatory requirement to conduct a research study" you might as well not bother asking.

A third question is how do you choose between two research designs, one with greater scientific integrity but greater risks to the patients and another with less scientific integrity but also fewer risks to the patients. Again, I would adopt a pass-fail standard. If both research designs are adequate and both levels of risk are acceptable from an ethical perspective, then trust the researcher to choose the right design. Certainly you can offer an opinion, but don't force your preference on the researcher.

This work is licensed under a Creative Commons Attribution 3.0 United States License. This page was written by Steve Simon and was last modified on 2010-04-01. Need more information? I have a page with general help resources. You can also browse for pages similar to this one at Category: Ethics in research.