Checklists Can Save Lives

Reader Comments
Back to blog

I don't think it builds my confidence to think that these doctors, who apply for grants for a living, don't understand the regulatory environment in which they apply. Imagine the surgeon staring at the tray: "gosh I don't remember what all these different clamps do." Especially at the alleged premier research institution in America.
And it certainly does not build my confidence that people want to regulate an exception for this activity. Suppose that checklist, inadvertently and in good faith, unfortunately had a step in it that actually increased infection rates? And that step was never noticed because it is mixed in with other steps that offset the damage? Since it is exempt "quality improvement" research, there would be no oversight whatever.
And to bev, two questions: if it was so confusing whether to apply to the IRB, why did Pronovost submit the study to his own IRB in Baltimore? No I am sorry there is no diagnosis that explains ALL of the symptoms other than arrogant duplicity.
Second question: should a hospital without an IRB be doing human subjects research? I don't think so. I am sure the small community hospitals are full of good people but it is good for them to have an undiluted focus on patient care, not that stupid triangle that shortchanges everyone in the end.

Lars of MD 10:22AM January 27, 2008

Lars;
I had a different take on the letters than you did. I think Provonost told the OHRP that he was doing human subjects research because there was no other box on the form to check to indicate what he was really doing. If you read their regs, (see 'regulations' on the left side of their website) you will see what I mean; it's like wandering in a dark forest. Even the AHRQ director to whom I wrote, Dr. Francis Chesley, agreed that there is confusion among researchers and the feds alike over what constitutes "human subjects research". According to him, HHS seems to be talking this over and trying to provide some guidance as to what constitutes "research" vs. "quality improvement activity." I personally think they are using the wrong terms. The distinction should be between "human subjects research" and "quality improvement research."
As for the IRB requirement, see Dr. Gawande's article about how he is working with the WHO on similar "research" - if some of these hospitals, who are small and don't even have IRB's, have to form them and then obtain informed consent from all patients, his efforts will just die.
This is not a simple issue; I think everyone has to read all the articles, letters and regs themselves and make their own judgement.

bev M.D. of MD 7:57PM January 25, 2008

Orac: actually no, there is no disagreement over whether this is human subjects research, except in the sensational accounts in the press. If you read the actual record you find:
Pronovost told the HHS (the AHRQ specifically) that he was doing human subjects research. The OHRP based its conclusion in no small part on that statement, and others MADE BY THE INVESTIGATORS.
Repeat: Pronovost told the HHS (the AHRQ specifically) that he was doing human subjects research.
This is getting worse and worse, this is really irresponsible. Gawande is suggesting in the NY Times that people write to HHS, but it is not their fault and there is nothing they can do now. People who really care about patients over subscriber counts should write to the Michigan hospitals, tell them to put on a pot of coffee, page their IRB members off the golf course, and convene sessions to properly review and approve the study at their hospitals. They may (gasp) have to do it on an evening or a Saturday.

Lars of MD 1:49PM January 25, 2008

"You mean that there are circumstances where a multi-site, multi-state study only needs to be approved at one site in one state?"
I'm unaware of any such circumstance.
A human subject research study needs to be approved by the local IRB at each institution where the study is being conducted. What may have happened is that there was a disagreement over whether this study constituted human subjects research.

Orac of NY 4:50PM January 23, 2008

Just found time to read the OHRP letters. The press is grossly misrepresenting the regulators' position here. It was Pronovost who told the AHRQ that he was conducting human subjects research. He also told his IRB he was doing research. Then he told the IRB and the OHRP that he was entitled to an exemption for retrospective studies, while at the same time he told the AHRQ and his journal editors that he did a "prospective study" of an "intervention." His words, not mine.
Both of those statements cannot be true. He tried to pull a fast one and is now using righteous indignation as a shield. Gawande should disclose any honoraria he received for the times he spoke at Johns Hopkins.

Lars of MD 5:47PM January 22, 2008

You mean that there are circumstances where a multi-site, multi-state study only needs to be approved at one site in one state? Please tell me that isn't so. That sounds like a loophole you can fly a 747 through. Will we start offshoring IRB approvals in the future? I will do my multi-site trial here and in Lesotho, don't worry their board says my protocol is fine.
I remember that Ellen Roche case, it was closer to seven years ago. A healthy young person not a terminal cancer patient. Johns Hopkins stroked a presumably big check and went right back to business. The problem there was not a lack of IRB approval but the IRB never dug into the application and it became a rubber stamp. It will become one again if everyone sends their "dinner menu" to it.

Lars of 5:55PM January 20, 2008

You wrote that "[i]f Kristina C. Borror, Ph.D., Director of the Division of Compliance Oversight at OHRP, who wrote the letters to Johns Hopkins, reported to me, this is what I would have said: “The purpose of our office is to protect people in the health care system. The checklists aren’t research. They are part of a managerially sophisticated effort to make ICUs accountable for doing what all ICUs agree should be done, in order to protect people in the health care system. It is great that Johns Hopkins is studying that care management process. Please show them how their application could be corrected so that we don’t erroneously subject the project to a misguided requirement for individual informed consent.”
I'd add one more sentence: "We won't demand or suggest suspension or termination of any part of the effort while further action is pending."

Avery of DC 8:52AM January 19, 2008

The issue is a little more complicated than just sending it to the Michigan IRB's as well as the Hopkins IRB. The dissemination of the project to Michigan was part of an AHRQ-funded statewide project called the Keystone project. (see the New England Journal article for details; i don't remember them offhand) It was the Hopkins people who were the "researchers", so it was unclear to them that the Michigan IRB's also needed to review the work. Also, the Hopkins IRB felt that the project fell under an exemption provided for in the OHRP regs, in which informed consent was not necessary. OHRP disagreed with the interpretation of the regs, for somewhat arcane reasons. You would have to read the regs and the letters from the OHRP (both on their website) to get all the details.
In summary, there was more than one violation alleged - first, the Michigan IRB's did not review it and second, informed consent was not obtained from BOTH the patients and the caregivers!
The only ray of hope I see is that both AHRQ and OHRP now seem to realize they need to provide "guidance" on what constitutes "human subjects research" as opposed to quality improvement activity. I suggested to them in my letters that they convene some sort of consensus conference including members of the medical community, so that all may agree on this definition - rather than just coming up with some pronouncement on their own.

bev M.D. of MD 6:47PM January 18, 2008

Chris, you've made Pronovost's points exactly. But if even a couple of elements are now embedded in the routine, that's much better than just having one or two people conform.

Avery of DC 2:07PM January 18, 2008

I'm a pediatric ICU doc and am sorry to say we don't use a list, at least not formally. On the other hand, like many places we found that at least in the case of venous line placement, enforcing "full barrier" rules (mask, hat, gown) plus surgical scrub of the site reduced our infections. The measures are largely enforced by the nurses, though, which is not optimal, and the list is a mental one. After a while, though, it becomes the new routine. That's a weak way to do it.
All this has made me talk to my colleagues about getting some lists going. Of course now we can't tell anyone whether they helped or not.

chris johnson of NM 1:54PM January 18, 2008