How to convince team members we need IRB approval?

Question

I come from a social science background and have recently moved into more of a public health role. Since I was doing experimental research, pretty much every study needed a review. As far as I can tell, only studies with publicly available data (i.e. secondary analysis) that is totally de-identified can skip review. Some of our studies meet that criteria but most don't. I think even at worst, better to submit and be told it is not necessary than the reverse.

Some team members are convinced that we do not need approval on primary data collection studies of qualitative interviews because they are not "about individuals", or even crazier, that it does not meet the definition of "research" based on the following: "According to federal regulations (45 CFR 46 102.d). Research means a systematic investigation (including research development, testing and evaluation) designed to develop or contribute to generalize knowledge."

Thankfully they are going to go through with it but there are still people who are unconvinced. Any time you gather data about perceptions, decision-making or personal opinions- anything outside of fact-based information about institutions (eg policies), I would argue it is "about individuals." I don't even know how to counter the argument that our work is not "research" or that we don't intend to generalize results, as this seems self-evident. I have tried to illustrate using language from existing documentation but it doesn't get far.

I think the real problem is that prospective study planning is a real weakness in this group, because from my perspective you should have everything you need to submit an IRB before you begin collecting data anyway. It should take an hour to prep an IRB application, not WEEKS like it takes these teams. I just don't see why they feel it appropriate to skip out on ethical oversight when we should already have everything we need to complete the application anyway before the study begins. But everything is so ad hoc that whenever they put together IRB materials, they have to change it later. I can't tell though if the arguments about lack of "human subjects" or "research" are a consequence of this or a separate issue, but I don't know how to address either by myself. I've spoken up on this many times but little progress is being made.

Anyone with similar experiences able to advise? Or am I just totally off base?

Answer 1

In my experience in the US, it is rarely up to researchers to determine by themselves that their research needs full IRB review when it involves human subjects in any way.

Additionally, "exempt" in the world of IRB DOES NOT NECESSARILY MEAN YOU DON'T NEED TO SUBMIT TO IRB, it only means that it is exempt from certain specific requirements of full IRB review, and instead can follow an expedited process.

Low risk, "exempt" research must be exempted by the IRB, not the researcher. You submit a brief application to the IRB, IRB says "we've determined this is exempt, so you do not need a full application" or they say "this is not exempt, complete this process for full submission".

There are other things that are not considered even "exempt" research that the IRB doesn't need to review, but even so, institutions may require that IRB make the determination in these cases, as well, and it's always a good option to the researcher even if it's not required.

This includes considering things as "not research". Things that are "not research" would include internal process reviews, things that you would not consider publishing. If you conduct a poll in your office to decide if they'd prefer cake or ice cream for the department party, that is sufficiently "not research" that you wouldn't even need to do the cursory IRB submission. Other circumstances may be more borderline.

I do think there are cases like department climate reviews (for example, to measure harassment or feelings of marginalization or discrimination among staff) and such where the purpose is not research and IRB need not be consulted, but these situations certainly require very careful attention to participant privacy, and should only be conducted with the involvement of HR and other institutional systems to protect staff.

I assume, though, that the work you are doing involves academic goals in some way rather than internal ones. This is (nearly) always "research", or at least close enough to have an IRB sign off on it; I don't know how to convey this concept to people except to simply state that anything to be published in some form is research because it's intended to contribute to knowledge; the "generalizable" clause there isn't about some statistical ideal. I do not think there is any magic way to resolve this or convince people, you will need to be insistent and persistent. You may be going up against years of entrenched bad procedures, there is no quick fix. I think you can do best to lead by example. Likely your institution has some people specifically involved with research integrity, so they may have further resources or may be able to schedule trainings/seminars on the proper procedures, including helping to facilitate submitting simpler proposals for quick review.

It varies from IRB to IRB, but some IRBs may be willing to approve "blanket protocols" that give you broad leeway to conduct certain types of low-risk research without further consultation. Some examples I've personally experienced are A) protocols that cover the 'research' done as part of a lab course, often where the participants are other students in the class. These will outline certain variables that are collected and certain experimental manipulations, and basically let students play around with research without having to go to IRB for every particular combination of variables they might come up with, and B) for IACUC (animal ethics committee) rather than IRB, basic tissue-harvesting protocols that allow for add-on procedures to already terminal experiments, for example to recover organs for tissue culture without getting further approval when that animal is already due to be euthanized for purposes of another study.

I'm a bit surprised you are having these difficulties in a public health department. In my experience, anyone in the health sciences has had these ethical research requirements drilled into them many, many times since their undergraduate days. There's no real excuse to be unaware. The place where there tends to be less awareness is more on the tech side of things, where people trained as computer scientists or engineers find that they need input from people on their systems and designs, and suddenly find themselves doing human subjects research despite not having any training in the subject. Same thing for business fields, where internal market data analyses are not considered research in industry, but in an academic/social science sense they certainly are.

Answer 2

united-states

One argument I've heard before from my colleagues is "we don't need IRB permission for this study since it is 'exempt' because it fulfills <insert exemption category here>."

In the United States, the Office for Human Research Protection notes:

The regulations do not require that someone other than the investigator be involved in making a determination that a research study is exempt. What they do require is that there be accurate determinations so that non-exempt research ends up being reviewed by an IRB.

So, at least as far as the law concerned in the United States, your colleagues are correct that for exempt research, the IRB is not legally required. However, the OHRP guidance continues:

However, OHRP recommends that, because of the potential for conflict of interest, investigators not be given the authority to make an independent determination that human subjects research is exempt. ... Institutions should implement exemption policies that most effectively address the local setting and programs of research.

This OHRP recommendation frequently leads institutions to enact policies that require investigators submit potential studies to the IRB for certification of exemption. My institution, for example, requires that the IRB make the determination. A quick Google search reveals many other institutions have enacted similar policies, like this (randomly selected) one from Northwestern.

So, how do you use this to convince your colleagues?

1. Check that your institution doesn't have a policy that requires the IRB to make the determination (odds are they might).
2. If there is no such policy, use the OHRP recommendations as an argument. "The Office for Human Research Protection recommends investigators not be given the authority to make independent determinations because of risks of conflicts of interest. If we make the wrong determination, we may be in violation of federal regulations."

Side note: If your institution does not have a policy against investigators making their own indepenent determinations, they should. Try lobbying your IRB to push for the policy.

Answer 3

Academia varies more than you think it does.

I work in an area without human subjects research, so I'll address the interpersonal issues of this only.

First of all, I wouldn't try to persuade your colleagues that IRBs are intrinsically important, something more than a bureaucratic hurdle. It sounds like you believe this and your colleagues don't, and I doubt that you can change their minds on this. Arguments like "the dean will shut us down", "the NIH could pull our funding", etc. might be more persuasive, if true.

Secondly, you say

prospective study planning is a real weakness in this group

If I understand, you like to plan your research projects out in full before you begin, and your colleagues don't. This might indeed be a weakness of theirs, but are you sure? I work in a field (math) where it's common to just "dive in", and where the scope of a project may significantly change halfway through. In math, this flexibility can be a great virtue -- sometimes the final product can be much more interesting than what you were originally hoping for!

In my experience, the people who have successfully changed my mind were those who first listened to me and understood where I was coming from. So, if you'd like to be more persuasive, I'd recommend investing additional time in understanding your colleagues' goals and methods.

Answer 4

Simple: If you do not have IRB approval, journals can, and in my experience will, refuse to publish your work.

Simply from the information that you are using "qualitative interviews" it is not possible to determine in IRB approval is required.