11

In human subjects research, it's fairly well settled that researchers may not go ahead without the guidance and approval of an IRB, even if their research otherwise lacks ethical infirmities.

What happens in the opposite situation? That is, suppose a researcher gets IRB approval for their project, but later on, ethical issues are found in the research that the IRB missed?

Since researchers are not allowed to trust their own understanding and must seek IRB approval even in the most obvious "there's no way that could be unethical" scenarios, it stands to reason that the opposite is true - that if research has been IRB approved, then researchers may go ahead with a clear conscience even if there could still potentially be ethical issues in the research.

I recognize that there could possibly be civil or criminal liability outside of academia, but I'm not asking about this.

If research turns out to be unethical, but it received IRB approval, who is held responsible? Is it solely the responsibility of the IRB for misguiding researchers, or does the researcher bear some responsibility for not recognizing the ethical infirmity?

Another way of asking this is whether researchers have a duty to police their own IRB, potentially second-guessing their approvals.

To be clear, I'm asking about situations where the average non-expert might be unclear as to whether or not something is ethical. Obviously, if an IRB tells someone, "Yes, it's ok to torture political prisoners as long as you kickback 5% of your grant to us", that's blatantly unethical and no one should accept that, but an IRB saying, "Yes, we reviewed your consent form and we don't think you need to disclose the extremely remote risk of the subject being hyperspace tunneled" might be one where the average person (or even researcher) would just get glazed eyes and defer to the IRB.

11
  • 7
    I don't see how anyone could go ahead with a clear conscience doing something they personally think is wrong and unethical. Maybe they can convince themselves it's fine since the "perfect authority" (IRB in this case) says it's fine, but that's an altogether different scenario.
    – Anyon
    Jul 24, 2019 at 14:15
  • 1
    If you harbor any concerns, you should talk to your supervisor/advisor and re-check with the IRB. Any subsequent issues will be associated with your name as well...
    – Solar Mike
    Jul 24, 2019 at 14:22
  • 20
    I know you ask a lot of IRB questions here, but I am curious what your actual experience is with IRB, because what you are describing doesn't make any sense to me. IRB is not a mechanism for researchers to submit unethical proposals to get a stamp of approval that makes them magically ethical. Researchers don't second-guess IRB approvals, they "first-guess" the protocols before they are ever submitted. IRB approval is a minimum to doing research in human subjects.
    – Bryan Krause
    Jul 24, 2019 at 15:04
  • 3
    One suspects quite different outcomes between intentionally misleading the IRB and the IRB not operating to standard. With shades of gray in between.
    – Jon Custer
    Jul 24, 2019 at 17:37
  • 5
    @Bryan Krause : I think what he is after here is not that "IRBs" "make unethical things magically ethical", but rather to what extent that the emphasis put on IRB usage and their relevant "expertise" shifts blame away from the researcher when the IRB, in such position of expertise, makes an error of judgment and the researcher relied on it in good faith? I also believe OP's idea of a situation is less one where that the researcher harbors doubts and does not disclose them to the IRBs, but rather where the researcher does not have the competence to assess the ethical dimensions. Jul 25, 2019 at 11:00

3 Answers 3

24

Since researchers are not allowed to trust their own understanding and must seek IRB approval even in the most obvious "there's no way that could be unethical" scenarios, it stands to reason that the opposite is true - that if research has been IRB approved, then researchers may go ahead with a clear conscience even if they personally feel that the research still has ethical issues.

I don't see it that way. The IRB doesn't replace the author's judgment, it supplements it, and the ultimate decision should be based on the "logical AND" of the author's and IRB's approval.

Since conducting unethical research seems clearly worse than not conducting ethical research (a Type II error is worse here than a Type I error), it makes sense to have an asymmetry between approval and rejection.

So in my view, even if the IRB has approved the research, the author should not go ahead if they do not believe it is ethical.

If research turns out to be unethical, but it received IRB approval, who is held responsible? Is it solely the responsibility of the IRB for misguiding researchers, or does the researcher bear some responsibility for not recognizing the ethical infirmity?

Practically speaking, the researcher would have some cover because of the IRB approval, but I would still hold the researcher primarily responsible. After all, it is their name on the paper.

Another way of asking this is whether researchers have a duty to police their own IRB, potentially second-guessing their approvals.

Well, the researcher ought to make sure they think the work is ethically OK before submitting to the IRB in the first place. But if they start having doubts later, or they are ethically uncomfortable with changes recommended by the IRB, they should not proceed until those issues are resolved. So in short: yes.

5
  • 1
    Actually, the IRB does more than supplement the author's judgement in the case that the IRB points out ethical issues. In general, you need approval, but, as you say, approval may not be sufficient. Necessary, but not sufficient.
    – Buffy
    Jul 24, 2019 at 15:00
  • "But if they start having doubts later, or they are ethically uncomfortable with changes recommended by the IRB, they should not proceed until those issues are resolved. So in short: yes." However, what happens if they have already started "proceeding" - the research is off the ground and into the air, and only then is the issue discovered, which to me was how I read OP's post? Jul 25, 2019 at 10:57
  • 1
    @The_Sympathizer I don't see how this changes anything. The responsibility to conduct ethical research should be blind to where in the research process one is. Jul 25, 2019 at 12:53
  • @BryanKrause Ah, so it does -- my bad, I missed that.
    – anon
    Jul 25, 2019 at 20:44
  • @DremConspiracy : I did not say it removed the responsibility, but rather because you phrased it as "should not proceed" regarding doubts that come up "later", but "later" suggests they have already "proceeded" into the research to at least some extent even if not all the way, and it's impossible to act on doubts that have not yet arisen. Do you mean "proceed further"? Jul 26, 2019 at 7:19
11

While the exact advice will depend on your institution's IRB, researchers have a duty to follow up with IRB if something goes wrong in research.

For example, here is the FAQ at the University of Tennessee at Knoxville:

What if I have an “adverse event” occur during my data collection?

Adverse events must be reported immediately to your advisor and the Compliance office.

IRB training and policy covers what an adverse event is. Here is the University of Michigan glossary definition:

Adverse Event (AE)

Any experience or abnormal finding that has taken place during the course of a research project and was harmful to the subject participating in the research, or increased the risks of harm from the research, or had an unfavorable impact on the risk/benefit ratio. The FDA also includes in its definition abnormal preclinical or laboratory findings which may not yet have resulted in direct harm to subjects (e.g., a bacteria is identified in a culture from the same batch of cells used to produce a vaccine which has been administered, even if no cases of infection have been reported). The event may or may not be caused by an intervention (e.g., headache following spinal tap, death from the underlying disease, car collision). Adverse Events also include psychological, social, emotional, and financial harms. See also Serious Adverse Event and Adverse Drug Reaction.

Michigan's Office of Research has an entire timetable to determine how soon one should report an Adverse Event, as well as policies for reporting an ORIO (Other Reportable Information or Occurrence).

Generally, your IRB training should cover when and how you report issues, and if you have any questions, you should ask your institution's IRB. In other words, researchers definitely have a duty to recognize when harm or risk of harm is incurred by their research, even if their project is already approved.

1

There are really two kinds of answers to "what happens?"

  1. IRBs are regulated by the OHRP which is a federal agency. If an IRB is actually approving things that should not be approved there are a variety of possible sanctions, which can range from mandated retraining to all research at an institution being shut down until the situation is remedied. Yes, this has happened when the situation is bad enough.

  2. IRB approval is not supposed to be a one time thing, there is supposed to be ongoing monitoring, dealing with adverse events and often continuing review on a regular schedule. If the IRB discovers a problem (or is informed of one) they will shut down the research or take other actions and will report to OHRP if appropriate.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .