Effect Measure

“Human protection” fiasco: nevermind

Recently we posted on the insanity of requiring informed consent for posting a hygiene checklist in the ICU. This week the New England Journal of Medicine weighed in.

Here’s some background from the NEJM Commentary:

About 80,000 catheter-related bloodstream infections occur in U.S. intensive care units (ICUs) each year, causing as many as 28,000 deaths and costing the health care system as much as $2.3 billion. If there were procedures that could prevent these infections, wouldn’t we encourage hospitals to introduce them? And wouldn’t we encourage the development, testing, and dissemination of strategies that would get clinicians to use them? Apparently not, judging from the experience of Peter Pronovost and other Johns Hopkins investigators who helped 103 ICUs in 67 Michigan hospitals carry out a highly successful infection-control effort,1 only to run into major problems with federal regulators.


The intervention was designed to improve ICU care by promoting the use of five procedures recommended by the Centers for Disease Control and Prevention: washing hands, using full-barrier infection precautions during catheter insertion, cleaning the patient’s skin with disinfectant, avoiding the femoral site if possible, and removing unnecessary catheters. The hospitals designated the clinicians who would lead the teams and provided the necessary supplies. The investigators provided an education program for the team leaders, who educated their colleagues about the procedures and introduced checklists to ensure their use. Infection-control practitioners in each hospital gave the teams feedback on infection rates in their ICUs.

The investigators studied the effect on infection rates and found that they fell substantially and remained low. They also combined the infection-rate data with publicly available hospital-level data to look for patterns related to hospital size and teaching status (they didn’t find any). In this work, they used infection data at the ICU level only; they did not study the performance of individual clinicians or the effect of individual patient or provider characteristics on infection rates. (Dr. Mary Ann Baily, NEJM, hat tip reader KR)

It’s this last part that created the problem. The investigators bothered to evaluate the intervention. It was “research” so the Office of Human Research Protection (OHRP) within the Department of Health and Human Services ruled it violated federal regulations by not requiring informed consent from patients and doctors. A Johns Hopkins review board had already ruled it wasn’t necessary. But OHRP got a written complaint and decided the complaint had merit. Full review and informed consent were required.

The NEJM commentary chose to focus on the fact that this meant the regulations weren’t clear. What wasn’t clear was the thinking at OHRP. “Not clear” is a euphemism. Stupid and devoid of all commonsense is more like it. The OHRP position that the checklist is fine as long as no evaluation of its effectiveness is done says it all.

Well, not quite all. Someone at DHHS got the message. This note prefaced the NEJM Commentary quoted above:

On February 15, 2008, after this article had gone to press, the Office for Human Research Protections (OHRP) issued a statement (www.hhs.gov/ohrp/news/recentnews.html#20080215) expressing its new conclusion that Michigan hospitals may continue to implement the checklist developed by Pronovost et al. “without falling under regulations governing human subjects research,” since it “is now being used . . . solely for clinical purposes, not medical research or experimentation.” OHRP further stated that in the research phase, the project “would likely have been eligible for both expedited IRB review and a waiver of the informed consent requirement.”

In other words, “Nevermind.”


  1. #1 caia
    February 22, 2008

    My question is, who filed the written complaint? Someone angry because use of the checklist put them at less risk? …An antibiotics manufacturer?

    My suspicion would be a doctor or hospital bureaucrat who felt it made their previous procedures look shoddy by comparison, and that the assessment opened them to litigation. Either that, or a misinformed/extremely phobic patient.

  2. #2 jen_m
    February 22, 2008

    I wouldn’t jump to conclusions about the source of the complaint. I can imagine all sorts of political currents involving central line placement and care that could lead to a disgruntled complaint – turf wars among floor, ICU, surgical and specialty care teams; anger because of disagreement about what constitutes an unnecessary central line; offense at the required changes to longstanding habits around line placement, maintenance and removal; interpersonal conflicts between the chosen team leaders and ICU or infection control staff.

    For that matter, since providers don’t have crystal balls, some lines that are pulled as unnecessary may need to be replaced, and central line placement is a big deal – it wouldn’t take being misinformed or phobic for a patient or family member to be distressed at a line being pulled and replaced mainly for what are perceived as inflexible new hospital regulations.

  3. #3 caia
    February 22, 2008

    Perhaps I gave the caregivers too much credit; I assumed that if they saw the benefits to patients, they’d be in favor of the new system. You’re right, the turf and personality conflicts are possibilities.

    But if the complaint were about placement and removal of catheter, wouldn’t it have been about the checklist itself, rather than the study? That’s why I said a misinformed or phobic patient, because only the checklist could have been implemented without the study.

  4. #4 Guy D. LaRue
    December 17, 2009

    The real problems with Dr. Pronovost’s central line infection studies have nothing to do with the IRB process and everything to do with his poor research design, his over-reaching conclusions, and the embarrassing manner in which insurance agencies have used those findings to avoid payment to healthcare facilities.

    Dr. Pronovost claims that every variable in his “bundle”(more commonly known as a “procedure”)is the reason for the decreased infection rates in his studies. This is not a valid or logical conclusion. Any single variable , or any combination of the variables, or even bias (like the Hawthorne effect)may be the cause of his reported infection rate reductions. I suspect that the variable of early catheter removal may be the single reason for the infection rate decrease.

    I assert that covering the foot of patient’s beds with a sterile drape has nothing to do with the reported reduced infection rates and only raised the cost of each insertion by $10 to $20 .

New comments have been disabled.