Hospital epidemiologists and the IRB: Two views

Viewpoint 1:
Silvia Munoz-Price

During the past few years, my hospital has experienced high rates of Clostridium difficile infections especially in one inpatient unit. This situation is of major concern to Hospital Administration, the Quality Department, the Infection Control Department, as well as medical and nursing leadership. As the Hospital Epidemiologist, my main duty is to better understand the pathophysiology of the infection control problems so that interventions can be tailored to our specific needs. My initial questions in this particular situation focused on where the acquisition of C. difficile strains were occurring. Are patients acquiring C. difficile at home, in the outpatient clinic, or in the inpatient units? If they were already colonized upon admission to the inpatient unit, then that would certainly explain why infection control bundles had been ineffective in controlling this problem for the past year. Or is it that the hospital environment is acting as a reservoir for C. difficile due to inappropriate disinfection? In order to answer these questions, we started a quality Improvement project, testing the unit with DAZO, culturing the environment, and performing active surveillance cultures of consecutive patients on admission and weekly thereafter. These stool surveillance cultures aimed to detect asymptomatic carriers presenting from the community.

After a couple of months of pursuing these initiatives, I am now getting questioned in regards to my lack of IRB approval. Pondering this issue, to me it is clear that as a hospital epidemiologist I am mandated to investigate the answers to the above questions. Should I get IRB approval ahead of time before each of these projects? I think the answer should be no. If we are not testing any invasive procedure or drugs, but rather we are trying to understand how to prevent infections in our patients, we should not be required to get an IRB approval ahead of time. I do think the IRB should be involved once we determine that the data collected merits publication. THAT is the time when hospital epidemiologists should get the IRB involved. Otherwise, we handicap ourselves on the investigations that we do as part of our duties. Another situation that should warrant IRB approval from the start is multicenter studies of infection control interventions, given that their intent from the start is to publish the findings.

It is true that our job duties fall in the gray zone between quality and research, and some of us tend to publish a great deal of what is done as part of our paid job. However, we should always be guided by this initial question: Is the primary goal of the project to help my patients and my hospital or is the primary goal to publish the results? If it is the former, I say do not ask for permission to do your job. If it is the latter, then by all means, IRB approval is necessary.

Viewpoint 2:

Mike Edmond

I agree with Silvia that this is a problematic issue for hospital epidemiologists, and also agree that the interventions she describes should not require submission to the IRB. My view is that you need IRB approval if your intent is to answer a research question, and that you don’t need IRB approval if the intent is to do quality improvement, particularly if you are implementing interventions that have already been shown to be effective in published studies. I don’t think the intent to publish results has anything to do with the decision to submit for IRB review. For example, if I roll out chlorhexidine bathing to the entire hospital and find that our infection rates fall (or increase, or even if nothing happens at all), I don’t think I need IRB approval regardless of whether I choose to publish the results. On the other hand, if I want to do a trial of chlorhexidine bathing and randomize hospital units to test the hypothesis that chlorhexidine reduces infection rates, that would require IRB approval.

If you would like the US Department of Health and Human Services official opinion on what constitutes research and requires IRB review, I have attached their algorithms below. Once you go through those algorithms, it will all be crystal clear!