Lucian L. Leape, M.D.
Harvard School of Public Health, Boston, MA
DR. LEAPE: Thank you, Harvey. It is a pleasure to be here. Within the past month there were two news items with immense significance for quality and safety. I think they are the most important information I have seen in at least two years.
One was a report in Health Affairs from a Commonwealth Fund cross-national study. The headline grabber was that among Americans with at least one medical condition requiring treatment, 34 percent reported a medical error in the past year or so, higher than in any other country. It won’t surprise most of you to hear that that didn’t surprise me, but what did surprise me was another finding that got much less attention. That was that among these fellow citizens of ours with a medical problem, 51 percent forewent medical care because of the cost. That is, more than half did not get a lab test, did not get a prescription, or did not see a doctor because they couldn’t afford it.
I was quoted as calling that a moral outrage, and it certainly is. There is nothing as troubling to me—and I am sure to all of you—as the persistent, shameful unwillingness—not inability, unwillingness—of our country to provide health care for all our citizens. Lack of access is the most serious threat to quality and safety and also the one most easily remedied if we had the will. I am supposed to talk about safety and I will. But I don’t think they are so different, and I would even go so far as to say that until we solve the access problem, we can’t solve the quality and safety problems.
It was almost exactly to this day in November six years ago that the IOM report To Err Is Human shocked the world with the statement that
almost 98,000 people die annually because of preventable medical error. People today ask, “Well, is there any evidence that things are any better? Is health care any safer?” In fact, it has become quite fashionable to say that health care is not safer, but that is not true. We have made a lot of progress. I think there is no question that things are safer—not safe enough—but we have certainly accomplished a great deal.
I think our progress over the past six years can be thought of in three phases. The first phase was awareness. At the time of the IOM report of 1999, the vast majority of people—I would suspect probably most of the people in this room—were not aware of patient safety as a significant problem. There was a small group of people who had been working on the issue for 5 or 10 years, but as we used to say, we could hold our meetings in a phonebooth. It was really in 1999 that the magnitude of the problem became evident and people became aware of the fact that this was something serious.
The response was predictable, very reminiscent of the Kübler-Ross framework for grief. There was shock. There was disbelief. There was denial. There was anger. There was rejection. And then finally there was reluctant acceptance.
You may remember several papers attempting to show that the figures were greatly exaggerated. Sadly not so. If anything, they were an underestimate, but that has all pretty well subsided. That phase, phase one, has pretty much come to an end, at least for most people, with the acceptance of the fact that we really have a problem.
The second phase, which for lack of a better word I call the phase of definition, was where we set out to try to determine what we were going to do about the problem. When the IOM report came out, we had a lot of great recommendations. We could tell you all the things to do, but what we put forth was based largely on human factors theory, and it was based on experience in other industries because there had been very limited experience in health care and therefore very little evidence. So one of the major tasks of the past five or six years has been to develop practices in health care, to bring these into focus and identify safe practices. Indeed, we have done a tremendous job. I say, “we,” but I mean mostly the National Quality Forum (NQF), with help from the IOM and support from the Agency for HealthCare Research and Quality (AHRQ), and with some superb work done by Bob Wachter’s team at the University of California, San Francisco. They identified safe practices that really are effective and for which there is good evidence. As most of you know it was about two years ago that the NQF came out with a list of 30 safe practices that it recommended all hospitals implement. The Joint Commission was reading off that page, as they have increased the requirements for hospitals each year since.
We made a lot of progress in defining practices. We made a fair amount of progress in research. The only tangible thing that the federal government did to improve safety was an important thing, and that was to appropriate $50 million a year for research.
Now, in the world of research $50 million is barely a blink of the eye, compared, for example, to $28 billion for NIH, but it was $50 million more than we had, and it enabled us to start doing research on these practices. And even more important, I think, it allowed us to attract people and develop a cadre of young investigators who will be our leaders of the future.
So getting started in research was a very important part of this phase of definition. Then we also began to develop some measures. Both the other speakers mentioned the fact that we don’t have enough measures. However, we have a lot more than we had five years ago, and so we are making some progress.
The measure that is the most interesting to me and that probably most of you haven’t heard about since it hasn’t been published is the one that Ben Sachs and his colleagues developed to study the effect of team training in labor and delivery. They decided to train all their people in teamwork, but they faced the very simple question, how would they know if it made any difference? So they developed a measure called the adverse outcome index, in which they assigned a value to every bad outcome that could occur after pregnancy and gave it a weight. Mother death, for example, would warrant a high score. By assigning a score to every adverse outcome and then dividing the total score by the number of deliveries, then came up with an average outcome index.
They found that when they did team training they were able to reduce that measure by 54 percent for pregnancy and about 18 percent overall. So they had a measure of the global measure of safety. We did a lot more of those, but they showed us the way to do it.
This demonstrates the progress I think we have made in the past five or six years, and the end point is that we now know what to do. We have an agenda. We don’t really know exactly what to do, but we have a big enough agenda to keep us busy, to get us started. So the question now is moving into the third phase, which is implementation.
How many people in the room actually work in a hospital and take care of patients? The rest of you might not understand what I’m talking about here. What I am trying to get at is that trying to implement changes like safe practice, which seems so simple on the surface—for example, disinfecting your hands before seeing a patient—is not simple because making changes in patterns is incredibly difficult.
The implementation challenge actually has two steps to it. The first is defining what it is you want to do, and the second is figuring out how to
do it, getting people to put it in place, then making sure that it happens every single time. That is difficult.
We had an experience during these past two years in Massachusetts, for example, where we had a statewide effort to implement two safe practices. One was reconciling medications. That means making sure that when patients enter the hospital they are prescribed the same medicines they were getting outside the hospital. That sounds pretty simple, but it’s not. Eighty percent of patients don’t get the same medications. It’s a big deal to reconcile them.
The second safe practice was communicating clinical test results so that when a patient had an important test result their doctor got it. How do you make sure the doctor gets it to the patient? These things turn out to be very difficult to do. We had a year-and-a-half in which we had 80 percent of the hospitals in the commonwealth participate and we had people helping us. We gave them training in PDSA cycles and change. We had hotlines. We had a web site. We had tool kits. We had meetings. We had scores. At the end of that time, 20 percent were succeeding. Another 30 percent had begun to make some progress. That is with the full court press. So putting these things in is not easy, but there has been a remarkable development that I want to tell you about. It is the work of Peter Pamazosh at Johns Hopkins University and some others who decided they would see what they could do about reducing infection in the intensive care unit.
You may have seen the data from the CDC about a year ago. The CDC estimates that annually 2 million people in the United States become infected in the hospital. That is about 1 out of 17, and it is estimated that 90,000 people a year die from nosocomial infections acquired in the hospital. That is almost as big as the number for the total number of preventable deaths. Peter and his team decided to see what they could do about preventing one particular kind of infection—that associated with central venous lines, the plastic tubes that are put in the large vein for all kinds of treatments for patients who are very sick. Many patients in intensive care units, probably about 30 percent, will have them, and about 5 percent to 10 percent of patients will get infected and about 5 percent to 10 percent of the patients who get an infection will die from it. I am going to give you the bottom line: They really eliminated that infection. They went 8 months and then on to 12 months without a single central line infection.
This is of immense significance because it demonstrates something that everybody thought impossible. One of the responses to the IOM report and one of the responses to the stuff that I have been talking about for 10 years is that errors, like the poor, are always with you. We will always have mistakes. But it is not necessary. The whole concept that we could have error-free performance, that we could have defect-free perfor-
mance, what I call “getting to zero”—meaning zero defects—is a very controversial, dramatic, hard-to-grasp concept for most people in medicine. But Peter showed it could work. As a matter of fact, a number of others have done it. Jodius has had his programs going now for several years. A little hospital in De Sota, Mississippi—Memorial Baptist Hospital—has eliminated ventilator-associated pneumonia, a similar nosocomial infection. That is a giant step forward in terms of moving ahead and breaking through the conceptual barrier and saying, “You know, we really could have a defect-free environment.” But first you have to concede that it is possible.
There was another aspect of implementation that hasn’t done so well, however, and that is the bigger issue of how we change the culture. Medicine has a very dysfunctional culture. I don’t know how many of you have read Peter Sanger’s work or the work of people who talk about organizations. Most hospitals are certainly not learning organizations as Sanger defines them. They tend to be very hierarchical and autocratic. They tend to have a lot of trouble with unclear lines of authority. Many people have talked about the massive underinvestment in information technology in hospitals, but over and above that hospitals have a very peculiar characteristic—the most important player on the team, the physician, usually doesn’t work for the hospital. In 90 percent of hospitals, not true of academic medical centers but essentially all the rest, physicians are independent contractors, meaning they do not necessarily have to do anything anybody wants them to do. Good teamwork and developing a cohesive approach to the problem become very difficult to achieve in that kind of environment.
We are making progress. There have been a lot of interesting things and I think there is no question that the situation is improving, but we have a long way to go, and until we see the CEOs and the physicians move ahead to address issues of safety, we are not going to take the quantum leap forward that we need to.
There have been several recent developments that are worth mentioning. First, this summer Congress, at long last after five years, passed the Patient Safety Act, which provides protection from discovery information about safety errors and adverse events that is shared across state boundaries and therefore escapes state peer review protection. The hope is that this will stimulate the development of reporting systems throughout the country that will improve our learning. I think it probably will. I am not sure it is going to improve safety as much as the people who proposed it think it will, but it certainly won’t do any harm and it will make life a little bit safer for some people.
I think we clearly have the will to do something about the informa-
tion gap. I remarked to George Isham earlier that I think it is such a shame that somebody with the talent and commitment of David Brailor has so little in the way of financial backing. The amount of money committed by the federal government for improving information technology, specifically for developing electronic medical records, which is what we desperately need, is off by three zeroes. It is $50 million instead of $50 billion. We’re not going to get there that way, but we are going to get there another way and that is when payers suddenly realize that it is in their interest to have an electronic medical record. Many of us believe that the electronic medical record is the single thing that would make the most difference in the improvement of quality and safety. It is only a tool, but it is a tool that we desperately need for a number of reasons. The problem has been that the up-front expense as well as the time and effort that it takes have been sobering for physicians, who have long wanted to do it. There have been a number of other problems, too, but that has been a very major area.
Recently payers have begun to realize that they are able to benefit from this. We did a in-depth study that some of you may have seen that showed that for every dollar invested in the electronic medical record, the payers would get back $3 in three to five years. We have convinced the payers in Massachusetts about this, and we now have a pilot study funded at $50 million by BlueCross BlueShield in three communities. If it is successful, and I have no doubt it will be, it will then lead to wiring the whole state so that all patient medical records become electronic medical records in all doctors’ offices and hospitals. There are other experiments going on throughout the country, and therefore I predict that in five years we will be well along and that well over 50 percent of patients, maybe 60 percent or 70 percent, will have electronic medical records. Probably the project will be completed within the decade. It is about time, but it is coming.
The most remarkable development, though, came out on October 13. It was a report of my friend Peter Corngros’s activity in Michigan because Peter, again, got BlueCross BlueShield to put up the money and talked almost all—over 70—hospitals in Michigan into joining a project in the intensive care unit to eliminate central line infections and ventilator-associated pneumonia. The news item that came out on October 13 told us the following: Sixty-eight hospitals have now gone more than six months without a central line infection or ventilator-associated pneumonia. On the basis of their previous history, they estimate they have saved 1,578 lives, 81,020 hospital days, and $165 million. This is big stuff. It is no longer an idea. It is no longer one or two hospitals. This is something we all can do, and the rewards are fabulous.
If every hospital would do just that, we would reach Don Berwitz’s
“100,000 Lives.” So that is a turning point. We are beyond theory now. We are into practice and the payoff is huge, and it is going to be very exciting to watch it go on.
Probably everyone has their own ideas about where we move forward from here, and I don’t pretend to have a corner on the market. We clearly need better data. We clearly need to have better measures, and as I said before I think the electronic medical record is the most important thing and I am pretty optimistic that we are going to have it.
How we accelerate the implementation of safe practices is an interesting question. Maybe pay for performance is the way. Pay for performance means that we pay for results, and not whether you give debatable arguments to 100 percent of the patients. We pay on whether you actually eliminate the infection. For example, when I said this to BlueCross they said, “Why don’t you give half of that $165 million back?” We need to think about that. I am sure many of you have read writers like Brent James who have noted that as you improve the quality of care you lose money. The classic example is treating patients with asthma. If you do a good job treating patients with asthma you don’t see them anymore. They don’t get admitted to the hospital anymore, and both doctors and hospitals lose money. There is something wrong with that. That gets at this whole issue of needing to have a different way to pay for health care. Instead of paying for services, we pay for outcomes, and I would suggest it is time to figure some creative ways to pay for results, to pay for having no infections, to pay for having no care of asthma, and so forth. People are thinking about this and so it is an interesting frontier.
Finally, how do we get to the CEOs and physicians? This is one of the tough ones and we haven’t got a whole lot of answers, but I think one thing we have to do is create environments where it is the norm to be honest and open to make it safer for doctors to talk about things. That is perhaps a subject for another meeting, but we need to create a different environment. I think paying for results will get their attention. I think there is no question about that, but at the hospital level I have long thought that it is time to seriously consider making hospitals accountable for the cost of injuries.
They are doing some interesting work in Minnesota that George Isham may or may not want to talk about. They are saying that we no longer will pay for “never events,” but hospitals by definition don’t often have never events. They have a lot of other kinds of problems, and the patients pay for it. The regional medical practice study that generated the number of 98,000 for people dying from medical error found that patients pay one-fourth of the expense of medical injuries out-of-pocket. So it is a huge amount of money, and if we want to do something about improving our systems, it seems to me we ought to hold accountable the organiza-