Iconoclast or bridge builder? Pronovost Q&A
[Editor's note: Peter J. Pronovost, MD, PhD is a professor in the departments of anesthesiology and critical care, surgery, and health policy and management at the Johns Hopkins University School of Medicine. He also is the medical director of the Center for Innovations in Quality Patient Care at Johns Hopkins. Though his background is not in health care epidemiology, he is a lead author of a groundbreaking study on catheter-related bloodstream infections (CR-BSI) that is frequently cited as proof that health care-associated infections (HAIs) are preventable rather than inevitable.1 The CR-BSI prevention implemented at Johns Hopkins and in ICUs throughout Michigan emphasizes:
1. Hand hygiene
2. Full-barrier precautions during catheter insertion
3. Skin cleansing with chlorhexidine
4. Avoiding the femoral insertion site
5. Removal of unnecessary catheters.
A checklist used by clinicians to ensure aseptic technique during catheter insertion has become one of the more well known aspects of the program. Having driven CR-BSI rates to near zero where it has been implemented, discussions are currently underway to expand the program into 10 other states and several countries in Europe. We recently conducted an in-depth interview with Pronovost for this issue of HIC.]
Q. You testified at an April 16, 2008 Congressional hearing on preventing HAIS that "one of the basic issues is that we have failed to view the delivery of health care as a science." Can you elaborate on that?
A. The advances in biomedical science have been nothing short of breathtaking over the last decade. We have sequenced the human genome, we cure most childhood cancers, and AIDS has now become a chronic disease. This is really awe-inspiring, and yet we have around 100,000 people dying from infections annually. I try to reconcile that and say, "How is this possible?" Though it is a complicated problem, the simple answer is that the biomedical community in health care has had a very myopic view of science. Science is studying a new gene, finding a new antibiotic that is effective. Looking at the use of that antibiotic is viewed as the "art of medicine." We don't invest — either financially or in human capital — in really studying the delivery of health care as a science. It is a science — and it is not just epidemiology. It is behavioral change. It is economics. Because we haven't viewed it as a science, we haven't made the kind of progress that we could have. We have looked for quick short-term answers — a kind of just-go-do-it attitude in quality and safety. I think that has been misguided.
Q. You have said that infection preventionists can not "own" this problem. When you use a checklist tool — like the one that has been so successful in your program — is it essentially to expand "ownership' by improving compliance?
A. It is all about changing behavior. The health care community has been challenged to change evidence into practice. Part of that challenge is we haven't summarized evidence in a way that reflects how busy clinicians think or that is particularly useful to them. Practice guidelines are often 100- to 300-page documents with hundreds of conditional probabilities. They are a necessary synopsis of the evidence, but they can't be the end product because they are not useful to the busy clinician or nurse at the bed side. What we need is a more functional format, and the checklist is elegant in its simplicity. The science of converting both tacit and empiric evidence from a guideline to a checklist is immature. We took the CDC guidelines [for CR-BSI prevention] and made them into five items. Where they the right five things? Well, who the heck knows, to be quite honest? I won't say we pulled them from the hip, but I'm a clinician and a clinical researcher so I looked at how strong the treatment effects were, how strong the evidence was, and [weighed the] barriers to putting that research into practice. We simply said this is likely to work; this isn't likely to work. It looks like we got it right.
Q. One aspect of the program is that team members can speak up and stop the procedure. More to the point — nurses can remind physicians if they missed an important step on the checklist. Has that been a critical element?
A. It's been absolutely key. The checklist simplifies the evidence but if you have a checklist that nobody uses it's not going to do very much good. Look, I am a struggling practicing doc; I still do a lot of ICU time. I know that I forget things. I know that I am in a hurry and may forget to wash my hands or use full barrier precautions. We have to work as a team. But when I first said this you would have thought it was World War III. The nurses rolled their eyes and said, "My job isn't to police the docs. If I do, I'm going to get my head bit off." And the docs said, "Peter, you can't have a nurse question me in public. It makes it look like I don't know things." To which I said, "Welcome to the human race. We all don't know things."
The striking thing was the debate wasn't about the evidence. The debate was about the power and political struggle between physicians and nurses. We took that discussion to a higher order and made patients our North Star. I pulled the groups together and said, "Is it tenable that we harm patients here at Johns Hopkins or in Michigan?" People said "no." And I said, "Then how could you as a nurse see someone not wash their hands and keep silent?" That's not acceptable so you need to speak up. At the same time you can't be hung out to dry [for doing this], so docs, "Let me be really clear. The nurse is going to stop you and unless it is an emergency you are going to go back and fix the defect. If you give them flak, nurses page me any time of day or night. There is not going to be discussion on this. You will be supported. Period." The remarkable thing was I was never paged. We are all the same team working toward a common goal, and that is the best patient care possible. It's a huge culture change. Once you have that it's easy to fix things. Now our ICU team is saying what's next? Let's work together. But if we didn't change that culture it would have never been possible.
Q. Under testimony during the hearing you agreed a uniform national infection rate reporting system could be beneficial if it was accompanied by ongoing efforts to prevent infections. Will legislative mandates have to be the answer?
A. I'm uncertain about this because the Michigan results happened without public reporting. When we were working with many hospitals in New Jersey they implemented public reporting during our intervention. It was striking that the discussion with hospital CEOs in many cases switched from doing good to looking good. When it was public, the focus went from using checklists to arguing about whether that numerator (infection) really belongs in there. I mean literally hospital CEOs were arguing with people that that wasn't really an infection. The challenge is that we are using surveillance definitions to monitor these rather then clinical definitions. And because they are a surveillance tool they have some bias. There is some subjectivity to them. These definitions were developed in a much lower-stakes environment when the focus was just on monitoring improvement. I don't know if they are going to be valid enough for the kind of high stakes environment where you are not paid for infections, where there is public reporting, and quite frankly, where you may subject yourself to lawsuits. I think that those things have more risk for harm than good. Because as long as a measure has any kind of noise or subjectivity to it — which it will, when you start increasing the stakes — its economics 101: The measures are going to be gamed. You can imagine in what direction.
Q. You have certainly shaken up traditional perceptions of infection prevention. Do you see yourself as something of an iconoclast?
A. No, I view myself more as a bridge builder or a consensus builder. I live in two communities. I live in an academic research community and I live in a quality improvement, real-world practice community. For too long those two communities have been running in opposite directions. The quality community has avoided science and said we don't need it — just go do things. The research community has had a very myopic view. I see myself laying down the planks to connect those two worlds. Then we can go from basic science to clinical science and improved health care delivery is actually realized. Right now, there is a huge gap in that translation. We're not going from research to practice. Insurance, regulators, IPs, clinicians — we are all looking for the same thing. We all want improved health care at reduced costs. That's achievable if we learn to play in the sandbox together, so maybe me and you have to give up our politics and our hierarchies but the goal is something greater. When it works like in Michigan it is truly breathtaking because patient outcomes improve, costs drop and, more importantly, it puts joy back into the daily work life of many clinicians and administrators. People are energized about this. It's rewarding and they feel good about their work. As they should, because they are doing really sacred work."
Editor's note: Peter J. Pronovost, MD, PhD is a professor in the departments of anesthesiology and critical care, surgery, and health policy and management at the Johns Hopkins University School of Medicine.You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content