Editor's note: Jennifer Daley, MD, is the Chief Medical Officer of Partners Community Healthcare Inc., the organization for the 6000 physicians employed/affiliated with Partners HealthCare System (which includes Massachusetts General and Brigham & Women's Hospitals). From 2002 to 2007, she was the Chief Medical Officer for Tenet Healthcare, one of the nation's largest hospital systems, where she was responsible for the development and implementation of Tenet's Commitment to Quality (C2Q). Her academic background (including her previous directorship of the Center for Health Systems Design and Evaluation in the Institute for Health Policy at Massachusetts General Hospital and Partners HealthCare) and her years of leadership at a huge multistate private sector system provide her with a unique perch from which to view patient safety implementation in complex systems.
Dr. Robert Wachter, Editor, AHRQ WebM&M: You've now been involved in trying to improve safety in both a community-based, very large multihospital organization and at very large academic institutions. What are the major differences and the major similarities?
Dr. Jennifer Daley: Let me start with the similarities. I think they're all equally challenged in improving patient safety. Regardless of whether it's academic or community hospitals, people are so focused on making sure the trains run on time in patient care that they often don't have the time to sit back and decide what to focus on in terms of improving safety. Then they have a hard time assessing in a comprehensive way the clinical and operational systems that need fixing so that they're safer. One of the contrasts between Tenet and Partners is the role of the corporate center. At Tenet, the staff at the corporate center would do the research on what the latest advances in patient safety were. Then we sent "transformation teams" to work side by side with the hospital staff to figure out what the opportunities were at the local level and then work with them on site for a few weeks to help them implement new ways to make patient care safer. We could document the improvements and support achieving the safety targets we had established as a system (like reducing hospital-associated infections). In the system I'm in now, there are so many capable resources at both the community hospitals and the academic centers that at the corporate level we have more of a convening function. Still, it involves a lot of coordination, a lot of dialogue, and a lot of encouragement. Sometimes safety improvements are generated when bad things happen to good people. Creating an environment where people feel safe in sharing a serious adverse occurrence and sharing what other hospitals or other practices in the system have done to close the gap is absolutely essential.
RW: When you're sitting in "mission control," how can you have a handle on what's happening in 100 different hospitals around the country?
JD: We had an occurrence reporting system. If anything was reported to a state Department of Public Health or the Joint Commission, those all came through my office and my staff reviewed them all. The average size of a Tenet Hospital is about 150 beds, so these things happen infrequently. But we could identify patterns that individual hospitals couldn't see. For example, how many wrong-site, wrong-side surgeries did they have, and then come up with the policies, procedures, and oversight system to make sure that changes were made. We took on the NQF Safe Practices. In many community hospitals with only 150 beds, they just don't have the staff to keep up with everything that's happening in the field of safety. So they were appreciative that we had a corporate-wide committee with hospital, corporate, and specialist represpresentation that was proactively coming up with policies, procedures, and implementation and tracking tools.
RW: What's the role of data when you're sitting centrally in a large organization? What are the elements of a dashboard that you might follow centrally?
JD: We built a very comprehensive clinical quality dashboard for the company. It incorporated a lot of things other than safety. But in the safety arena, I'll just pick one example. We had some process and structure measures when we began because we had to make sure that everybody had the infrastructure needed to do this appropriately. When I left, we were installing a very robust risk management and patient safety system from a company called Quantros. It replaced a somewhat rickety risk management system that the hospitals had. The risk managers and the patient safety officers just love this new system. So we did a combination of make and buy. What we did have to make was a hospital-associated infection reporting system, which was built with our IT partner, Perot Systems, which basically tracks device days and hospital-associated infections for central venous lines, catheter-associated urinary tract infections (UTIs), and ventilator-associated pneumonias. When I left, they were building up the modules for MRSA [methicillin-resistant Staphylococcus aureus], C-Diff [Clostridium difficile], and vancomycin resistance. In time, it will also have the antibiotic resistance patterns on them so that we could actually look at patterns across hospitals. Accurately tracking hospital-associated infections per thousand device days created a lot of conversations! I think some hospital administrators believe that hospital-associated infections are the "cost of doing business." But we were able to show them that it was possible to reduce the incidence of these infections. And we gave them an incentive in that 10% of their entire annual at-risk compensation was driven by their ability to drive down the rates of health care–associated infections. Although skeptics in the group doubted that would actually result in improvement, we saw the overall rate of those health care–associated infections drop by 50% within 18 months of implementation. Some hospitals had moved to such a low rate that they were actually measuring days since last infection. We were implementing all the best practices that came from IHI [Institute for Healthcare Improvement] and other sources. But it resulted in a dramatic decrease in hospital-associated infections.
RW: Can you talk about the role of physicians and the challenges of working with physicians in a very large organization like Tenet, where I assume that most of the physicians don't work for the organization?
JD: As a matter of fact, out of 21,000 affiliated physicians, only about 400 were employed by the organization. Many physicians were hard to engage because they're so busy. But the administration of the hospitals engaged them—the intensivists in the ICUs and the surgeons placing central lines. And we were transparent about the results. We also knew that most of the states in which Tenet operates were shifting to public reporting within a short time. I won't tell you that there weren't a few doctors here and there who didn't want a nurse assisting in a central line placement to stop them when they weren't appropriately gowned and gloved. We had good regional leadership on the part of our regional CMOs [Chief Medical Officers], and, in the larger hospitals, a Vice President of Medical Affairs or CMOs. When one of them presented the data and the ways in which one could improve, most of the doctors got on board.
RW: Talk about the role of reporting. I can imagine that you might hear about a case with a terrible outcome from one particular hospital, and your instinct might be that it would be great to have 50 hospitals learn of this case. Yet all of the understandable challenges of disseminating that kind of information—with medical-legal challenges and newspapers—how do you balance that?
JD: Because we could track what was happening within the system and we knew what the incidence of "serious reportable events" was, as I was leaving, we were creating what I call Post-it Notes. They aren't really policies and procedures. They look more like a Post-it Note on an 8 by 11 piece of paper. And it would summarize a case in a de-identified way, and then it would have all the things that front-line clinical staff could do to improve. The intent was to disseminate those broadly to nursing units and doctors' lounges. So rather than a dry policy and procedure, it was more like, "Here's a little vignette of what happens when patients are harmed and here's what you can do about it." You could never trace it to any particular hospital, but it captured all the salient features of some of the terrible things that happened. And then there would be, for example, the four things you or your unit can do, to make sure this doesn't happen at your hospital.
RW: How do you structure things so that, at the level of an individual hospital, people feel comfortable with the central office? That it doesn't feel punitive, it feels supportive. I assume there's some tension there—even though centrally you are trying to help, at the level of an individual hospital, sometimes it feels like Big Brother watching.
JD: When I got to Tenet, people would say—and this is my favorite statement about corporate—"People from corporate are like seagulls. They fly in, eat your lunch, dump on your head, and fly out." And it took us a lot of education and training and feedback to change that. The system had always had 25 or so people who had been going out and working with the hospitals to ensure that they were Joint Commission ready at any time. Those people always had a great reputation, because everyone knew they were there to help. Over the time I was at Tenet, we went from announced to unannounced visits. Then we built the transformation team approach, when we were going in and picking the top things that a particular hospital could improve from a balanced scorecard perspective. Initially people were skeptical, but when you start addressing problems that really bother the doctors and nurses, which they haven't been able to get fixed because there are system problems—when you show them a system approach, most of them really go for it because it's the scientific method applied to hospital operations. But it was never our intent to go in and be punitive or critical. It was to help them provide better patient care. And I think that over time (I just kept saying it until I was blue in the face), most of them believed me.
RW: You talk about the scientific method and people instinctively appreciating that. I can imagine that the reaction to the science of safety probably differs substantially as you do things at Tenet versus doing things at Partners. Can you contrast the differences? I imagine that sometimes it's easier and sometimes it's harder in both places.
JD: In some of the community hospitals in Tenet, I would say that, because people are so busy, they don't actually have the time to read or explore current evidence and "best practices" as much as in academic settings. So we could bring them the latest evidence and we could classify the evidence in such a way that we would bring them the most accurate, reliable, and compelling reasons to change a particular approach. At Tenet, the resistance was often from people who hadn't kept up as much with the latest. My experience was that I would get calls from physicians or nurses who say, "This isn't right." I would send them a précis of the latest evidence and say: "Well can you show me evidence to the contrary?" Sometimes it was awkward, but generally they would come around. I think here at Partners, the level of evidence has to be so pristine that sometimes it gets in the way of meaningful change. Every "i" and "t" has to be dotted. I think the evidence that we need to make improvements in safety probably doesn't have to meet the 95% confidence interval, but it has to be high enough so that it's meaningful and directional. We're willing to make experiments of change and follow what happens and make sure we don't have any unintended consequences. Sometimes requiring absolutely randomized controlled trial evidence that a particular safety intervention works just isn't going to happen.
For example, we're not going to do a randomized trial of washing your hands. I do think that there are some recent experiences here at both the Brigham and the Mass General where they've made extraordinary improvements in hand washing. We all know that that's the right thing to do, but concomitantly they've seen a dramatic decrease in MRSA. It's hard to know in massive tests of change whether that is association or causation, but I'll tell you people feel pretty good about it. It wasn't just the infection control people; the CEO of the hospital, who happens to be a physician, took a personal interest in this issue. And he talks about it all the time. So there is leadership. And then there's a small bonus to every employee in the hospital associated with hand washing, which really—I mean it's nominal, but it empowers everybody. So the anecdote that I heard recently is: There was a chaplain visiting a patient at the Mass General, and a caregiver came in and the chaplain reminded the caregiver to wash his hands! So everybody is invested, from the CEO of the hospital right down to front-line staff, and environmental services and the chaplain service. And they've gone from 60% adherence (the national rate) to 90%.
RW: What was the role at Tenet of shared IT? Was there shared IT across all the institutions, and how important is that to try to centrally manage something this complex?
JD: Tenet has no shared common informatics platform for clinical care. They're working toward that, but it will take a while. The clinical information systems are extremely diverse. We did build a common platform for any important clinical data relevant to either core measures or patient safety or utilization management—whatever was in our portfolio for improvement. We were able to build interfaces with the existing clinical systems so that it powered a real-time quality monitoring system. On the other hand, Partners is deeply invested in electronic medical records for computerized physician order entry in the hospital, and by 2 weeks from now, 96% of our primary care physicians in the network, which is about 2000 of them, will be on an entirely electronic ambulatory medical record. The specialists have a target for next year to achieve 100% use of the ambulatory medical record. Having worked in an environment where there's nothing like that and an environment where there is something like that, I can see they both have their strengths. The ability of smaller community hospitals, critical access hospitals, or even significant good-sized community hospitals to have the capital and resources to install a progressive computerized physician order entry and data repository is out of the reach of many hospitals financially. As much as I would like to see it as the end all–be all of solving all the safety problems that we have, I think it's not realistic. We have to come up with other systems or toolkits to help them achieve standardization of systems that don't rely on having an electronic record.
RW: An organization like Tenet that's a for-profit organization and has been the subject of some media scrutiny, how does that change the burden as it relates to safety?
JD: I was brought to Tenet to transform the level of quality in those hospitals after enormous public scrutiny of their clinical care as well as some of their financial practices. We accomplished a great deal in the 5 years I was there. For Tenet, it created a burning platform, because they were in so much trouble with the public's perception of their quality of care. One of them was the utilization of procedures in patients who didn't need them, and the general perception that the level of care was not as good as it could be. They had to do something about it. As a result, they were very supportive, both financially and in terms of leadership. I had the distinct privilege of having my CEO be the sponsor for the entire Commitment to Quality program. Tenet considered it a wonderful investment because, with the tracking system we created, they were able to see that the quality of the care kept getting better every quarter. We had the data to show skeptics that we were improving. It also made the people at the bedside feel really good that they had been able to change and improve. They could tell their neighbors about it and talk to the patients about it. So it was a transformative experience for them. Here at Partners, people do an exceptional job, but they are very devastated, just like anyone would be or any organization would be, if they have a serious medical error. They take it very seriously and are very committed to systems change, root cause analysis, best practice sharing, and problem solving. Everyone knows that it happens at every hospital at some point, but they are totally dedicated to fixing the system problems when they identify them.