What I am going to talk about is serious. I am going to talk about errors in medicine and what we can do about it. I have a little temerity — not too much, but a little — in coming to you, because I know that you are already overburdened. You have a full plate, and there is not a single state board that I am aware of that has adequate staff or adequate funding to do the things that you are asked to do; and then here comes this guy from Harvard telling you that you should be doing something else. So, I have a little bit of humility about that but, like I said, not too much.
I recently had the experience of being at the 4th Annual National Patient Safety Foundation Annenberg Conference on Medical Errors. It was really an exciting time. I heard Ed Stoltzenberg, CEO of Westchester County Medical Center, talk about the terrible experience they had when a child was killed in their MRI machine because of a tank of oxygen getting loose, how they dealt with it, how he felt about it, and how it changed his life and his whole way of thinking about medical errors. It was a powerful presentation. I heard Ken Kaiser, whom many of you know was such a dramatic and dynamic leader of the Veterans Administration (VA) health system and who turned a stodgy old bureaucracy around to become a much more effective, although still stodgy, bureaucracy. It is a much more effective organization because of the person who made safety “job one,” which is what we want all CEOs to do. The VA is the leader in safety in many ways, and I am sure many of you know that.
I heard Kaiser talk about what the National Quality Forum is doing in terms of defining best practices. We have some standards, specifics that hospitals can look to, and that is going to make a big difference. Then I heard Paul Euling speak. Euling is a cardiac surgeon in Concord, New Hampshire, and he noticed that the major problems they had with service in his area were communication problems. Sound familiar? He had an idea. His idea — which he demonstrated using the people there — was to get everyone involved in the care of the patient to see the patient at the same time, to get them all in the same place at the same time. That way, when they make rounds on their cardiac service, they have 15 people in the room: respiratory therapist, social worker, pharmacists, nurse, all the usual suspects — including the patient and the patient’s family. They talk about the patient, make their decisions with everyone’s input, and then move on.
There have been some interesting results: Patients think it is terrific, of course, because someone is listening to them and they know what is going on, whereas normally they do not. The staff thinks it is terrific and, interestingly enough, mortality and complication rates have dropped. These were exciting stories to hear. I am telling you all of this to illustrate the point that the safety movement is really accelerating; and I think that we will see tremendous changes during the next few years as we get a lot of talented and motivated people thinking in ways that before would never have occurred to us.
An obvious question is, “Where do medical boards fit into all of this?” I did not see many people at that meeting from medical boards. The obvious answer is this: Safety is what you are. Everything you do is related to safety, from licensing of doctors to continuing education to all of the programs boards have for dealing with problem doctors — everything focuses on safety. It is central to your mission; you could say it is your whole mission. How are we doing in the area of safety? I would say that we do not do well with the “problem doctor” problem, and that is what I would like to talk about with you today.
As some of you know, I am an outspoken advocate of a non-punitive approach to medical errors. I am fond of quoting the comment that the problem is not preventing bad doctors from making mistakes, it is preventing good doctors from making mistakes. I believe that 95 percent of medical errors are made by good people trying to do a good job. There is another 5 percent that I would not call bad doctors; I would call them problem doctors. We do not deal with problem doctors well, and I think we need to look at that.
I would like to start first with a brief review of what I mean when talking about a new approach to patient safety. Some of you are no doubt familiar with this. The first thing is, let us be clear on what we are talking about. Safety is freedom from accidental injury. It is not freedom from all complications, because we do not know how to prevent all complications, but it is freedom from accidental complications, accidental things that should not have happened. It is not freedom from error. We will never have freedom from error. As a matter of fact, we really are not too concerned about errors that do not hurt people. We are concerned about are errors that hurt people. Safety means that the person feels safe and is safe from accidental injury.
Figure 1 contains a lot of information. We see the work of Rene Amalberti, a French physician, aviator, and safety expert. He examined various industries and put them on a scale in terms of their hazards. On the left side of Figure 1 is a scale that shows the number of deaths per year. It is a logarithmic scale, so the first measurement is for 10 deaths per year, followed by 100, then 1,000, then 10,000, and then 100,000. The bottom indicators show the number of encounters for one fatality. So if you look at this you see over in the lower right-hand corner the industries that we think of as safe. Let us choose scheduled airlines. Most of us flew on an airplane to get here. If you board a scheduled airline in the United States, your chances of dying in an accident are one in three million. If you look at scheduled airlines in Figure 1 on the left side you can see it comes out at 250 to 300, because that is the average number of people who die each year in the United States while traveling on scheduled airliners. Actually, in the past 10 years there were two years in which no one died and some years in which more people died. So, when you fly a scheduled airline in the United States, your chances of dying are about one in three million.
How Hazardous Is Health Care?
If you segue all the way over to the left side of Figure 1 you can see health care, along with some interesting companions. The number that I used for health care was the number from the 2002 Institute of Medicine report “To Err is Human: Building a Safer Health System.” The research shows that your chances of an accidental death when you go into a hospital are one in 200, which comes out to nearly 100,000 deaths. If you do not like those numbers and you think that they are exaggerated (you might think there are only 40,000 to 50,000 deaths) the figures for deaths caused by health care will drop down a whole quarter of an inch. If you think there are twice as many deaths it goes up a quarter of an inch. Even if the numbers are off by 100 percent, we are still sitting down there in that corner. We have a big problem, and the people who work in the safety industry do not have any question about whether or not there is a need to do something.
Why do people make mistakes? What we have learned is that mistakes do not come out of the blue; they are not acts of God; they are not lightening bolts, and they do not just happen. People make mistakes for a reason. Here are some of the reasons, the results of a number of physiological studies, although you do not have to do physiological studies to know that these are true. If you have lived on this planet for more than 10 years you are aware of the fact that you make more mistakes when you are in a hurry — like when you lock yourself out of the car. How many of you have had the experience, as I have more than once, where you set off on the weekend to drive to the cape or the mountains, and you end up driving to the hospital? You set out to do something and then you forget what it is that you set out to do because you were interrupted somewhere along the line. We know we make more mistakes when we are tired, angry, or bored. It is not a mystery. Mistakes do not just happen; they happen for a reason.
The other thing is that you do not have a lot of control over this. We try to control mistakes by being more careful, but even when we are careful mistakes will still sneak up and bite us. Knowing that people make mistakes, and that we all make a lot of them, does not get us far other than the realization that we do not make mistakes because we are bad, or because we are stupid. So I would say that Alexander Pope only got it half right. It certainly is human to make mistakes, but it is not a sin, and so forgiveness is not required. If you want to improve patient safety, you cannot accomplish this by going to church and praying.
Improving patient safety comes from trying to understand a little bit better, and the big understanding — for me at least, and I think most people would agree — was the breakthrough that came from James Reason, a psychologist from Manchester, England, who about 10 or 15 years ago coined the term “latent errors.” Reason said that we, as the individuals who make mistakes, are led to make them because of the situations in which we are put. The design of our work, the conditions, training, and other factors set us up to make mistakes. Therefore, if you want to prevent mistakes, you must pay attention to these precursors, these latent errors. This is a powerful concept. Let me put it another way, as illustrated in Figure 2, by using the terms “blunt end” and “sharp end.” The person who makes the mistake — the provider, the nurse, the doctor, or the pharmacist — is at the sharp end. This is the person interacting with the patient, the person who has the capability to hurt someone because they are providing the treatment. The sharp end is affected by many things going on above it: the so-called blunt end that involves management decisions, training, equipment design, process, and so forth. Let me paint a picture for you. Reason’s theory says there are latent errors that may cause a system defect that will then lead people to perform unsafe acts or make errors and that, in turn, can cause an accident. There may be triggering factors for an event and there are certainly defenses that keep accidents from happening, but if the defenses are not working or have weaknesses an accident can occur. For example, we have a latent error in which we say that it is all right for surgeons on our staff to determine their own working hours and workloads. We also have a systems defect where the orthopedic service decides that its 13 orthopedists will alternate full weekends. That means that when orthopedists are on, they are on from Friday night until Monday morning — about 60 hours. That is a systems defect. It works fine if you are not too busy. But consider a triggering factor of 30 cases during the course of the weekend. Do you think that maybe after being awake for two nights and working like that an orthopedic surgeon might possibly be more likely to make a mistake? You bet. He’s not thinking straight; he wants to go to bed; he wants to get the work done.
Levels of Safety
What happens is that he starts to cut corners. Maybe he has a patient with a broken leg and he takes him to the operating room but forgets to look at the chest X-ray. Unbeknownst to him the patient has a hemothorax and dies on the operating table. That really happened. It did not have to happen, because we have defenses. Maybe someone else saw the X-ray and noticed the patient had a hemothorax, or maybe the anesthetist said, “There’s something going on here,” and he tapped the patient’s chest and listened and said the patient does not have any breath sounds and we had better take a look. There are lots of defenses, and the fact is that we set the system up for the accident and the defenses work 99 percent of the time. However, when they do not work, the patient suffers. This is the concept: That person made a mistake, a serious mistake, but unless you understand why he made the mistake and do something about it, someone else is going to make that mistake later. So what we are talking about is getting the focus off punishing the person who made the mistake, because that will not accomplish anything in terms of keeping the mistake from happening again. We need to change the system so it never happens again. That is a little harder, but it is the only way to ensure safety.
BAD SYSTEMS, NOT BAD PEOPLE
The idea that medical errors are caused by bad systems and not by bad people is a transforming concept. It truly turns on its head most of what we have all been trained to think and the way we naturally react. We tend to think that when a person makes a mistake they are bad. We were taught that when we were three years old. This theory says that is sometimes true, but not most of the time. It is a system problem, and if you want to do something about it you have to shift your spotlight off the individual and really look at the system, and that is absolutely profound. The implications are straightforward. For the physician, it means that the individual doctor is a part of the system, but only a part, and in some cases not the most important part. You can achieve safety by just concentrating on the individual, and individuals can achieve safety on their own. We all need to work together to become a team operation.
What is the significance of this as far as what you can do? Let us go back to the blunt end and the sharp end. As the state medical boards, you are part of the environment that has an impact on the hospital, on the team, and on the provider. Let us take, for example, a simple situation that we are all familiar with: wrong-site surgery. If you have a surgeon who has operated on the wrong kidney, you might ask, “Did you have a policy to sign the site in your hospital, and did you follow that?” If you did not, therefore, we will punish you. On the other hand you might also ask what the system was in the hospital for preventing this. What we have learned is that if you really want to prevent mistakes you have to make it everyone’s responsibility so it is not just the surgeon that signs the site, but it is the radiologist, the anesthesiologist, and the nurse in the operating room. Everyone is keyed into this because safety is everybody’s responsibility. Safety is a team sport, not an individual exercise, so even if that doctor totally forgot about it, everybody else would not and the patient would not slip through.
Now, no system is going to be 100 percent. We can have the best system in the world and occasionally we will have a disaster. But the point is that a good system would be more useful when a mistake happens, to enable you to focus your searchlight on the hospital instead of on the doctor and ask how it was possible for a doctor to do this in your hospital. Some of you have done that and some of the investigations have been right down that line, and that is encouraging. I think that then we get to the real question of what is occurring and how we want to have this take place. Let me start with a little story that illustrates the problem. Some of you may have read an article in the August 7, 2000, issue of the New Yorker by Atul Gawande, a surgical resident at Brigham and Women's Hospital, titled “When Good Doctors Go Bad.” It should be required reading.
The fictitiously named Hank Goodman was a prominent and capable orthopedic surgeon who had a big practice. He was a well-trained, brilliant doctor, a doctor’s doctor. He was the one you took your wife to, and he had a busy practice, a practice that was too busy. His practice got bigger and bigger and he worked harder and harder, and after a while he burned out. As he burned out he started to cut corners and started to do things wrong. Things started to go bad, patients started to get in trouble, he was being sued, the malpractice suits were increasing in frequency, he was the constant subject of the morbidity and mortality (M&M) conference — one thing after another. It went on like this for a long time. There were several terribly quiet chats, as they say, during which a senior physician spoke with him confidentially. He tried to straighten out, but he really wasn’t able to and so it went on and on and on — and dozens and dozens of patients were maimed and damaged and hurt.
Finally they got him on a technicality; they got him because he failed to come to M&M conferences and they kicked him off the staff. This was after 10 years of steady downhill progression. What do we see from that? The first thing that we see is that it is all too common. Marilyn Rosenthal says, in a study of more than 200 cases, this is the typical way it is. Typically it is years — not days, not weeks, not months — before it comes to something actually being done when a doctor has really fallen off the edge like that. The second of course is that early warning signs are ignored. When something really bad happens, like a wrong kidney being removed, or something like that. What do you hear around the hospital? I knew that was going to happen someday. Well, if you knew that was going to happen someday, why didn’t you do something about it? Well, you know the reason. The reason is that we do not have a system for doing something about it. It is not easy. So, there are warning signs, but they are often ignored.
Another dramatic case that I am sure everybody remembers was Alan Zarkin, who in 2000 lost his license for carving his initials into the abdomen of a woman in Beth Israel Medical Centre during a caesarian section. Here is this clearly psychopathic behavior. But I thought it was interesting what the CEO of the hospital said. When the department of health, to its credit, slapped him with an $18,000 fine, the hospital CEO said there is no quality improvement system you could have that would pick up something like this because this was such abhorrent behavior. However, investigations by the department of health showed that the operating room nurses observed this bizarre behavior 18 months earlier. So, we tend to not pay attention to the warning signs. The reason, of course, is that it is not easy to do.
It seems to me that clearly we need to do something and we need to do it better. That leads me to my first of four radical ideas. We cannot make progress in dealing with problem doctors if we only deal with problem doctors. That is the fundamental problem with the usual approach. We deal with the person who is the problem, and we have to deal with the system and change the system. What is the system? The system is that it is up to the medical staff to identify these people and get them into some type of corrective action and, if that does not work, refer them to the board. However, medical staffs do not like to do that. You know the reasons as well as I do: It is distasteful and it is emotionally draining. After all, these people are your golfing buddies, your friends, someone you might have known for 20 years. They are a member of the family, and you do not want to hurt them; you would like to help them, but you do not know what to do. It is difficult and it takes a lot of time to put everything together, to document everything. There is a real threat of retribution, that the person will fight back. People do not take this easily. They will attack you personally, and they may sue. There is no good mechanism for doing it. So, it is not too surprising that in most cases not much happens for quite some time, until it gets to the point where there are no alternatives.
We have a situation where no one really wants to deal with it, and it is quite uncomfortable. Taking on one of these things is a good way to lose a lot of sleep at night, to make your life miserable while you’re making someone else’s life miserable. Not surprisingly, people do not sign up.
So, what is the answer? I do not know all of the answers. I do have some questions and some ideas that I would like you to think about. What we need to do most of all is devise a way to take that problem out of the range of the personal, individual, emotional and judgmental, and somehow depersonalize it and make it more objective. Make it based on something other than just our gut feeling and what we all know. There are several criteria that we should have up front when we try to do this that may give us some guidance. The first thing that I want is a system that is objective. That is, I do not want it based on my saying Joe is doing a bad job. Objective means based on data, some kind of evidence. We clearly need a system that is fair. When you say fair, that means that whatever system you are going to use to identify these people you want to do it in such a way that you are not just focusing on an individual, you are not zooming in on just the problem people, but you are measuring everyone by the same yardstick. So, whatever system we have, we all have to be willing to be measured by that system. I would really like it to be proactive, and I would really like it to be a system that would identify people before they hurt someone — when the problem is young, when it is early, when the physician is just slipping off the edge, when it possibly would be easier to help the physician and, of course, most of all, to prevent others from being hurt.
We want hospitals to do this. We really want hospitals to do two things: We want them to identify problem people and then to do something about it. If we want them to do that we must give them the tools; we have to give them the method; and we have to show them how. So, I would say that if we could have a system that is objective, that is fair, that is proactive, then maybe we are ready to get started. When I say based on data, we are talking about performance and about behavior. It is not what you know that counts; it is what you do. We are talking about performance standards, and it seems that there are distinct and separate domains. Clearly we are talking about substance abuse problems, including alcoholism; psychiatric problems; competence (which is a complex array of professional and technical performance issues, but it is doctoring in several dimensions); and interpersonal relations (which are important not just for doctors, but for everyone in the hospital), how they get along with their fellow staff and how they treat their patients. If we want to have an objective system we have to set standards for these kinds of demands. That is a big order. When you think about this you are considering several different categories. This is arbitrary, and it may not be the best way to break it down, but I find it useful in terms of thinking about it. You have certain standards that apply to everyone, whether you are a doctor, a nurse, or the floor sweeper, and these are the policies we have with regard to safety in a given hospital. Then you have the standards that are specific for physicians. Then you have different standards that are specific to individual specialties.
Examples that I have made up for discussion purposes are: In this hospital we expect that everybody will observe the safety practices that we have defined. So, if one of the practices is to sign your site, then we expect 100 percent compliance. We do not punish people for making errors; we punish people for misconduct, but we recognize that most errors are not bad, so we have a non-punitive punishment regarding errors. Non-punitive does not mean that you do not punish; it means that you do not punish for errors. We have a hospital policy that limits the hours and the amount of work you can do. Think about the orthopedist who was on for 60 hours. You cannot do that in my hospital because we limit work hours, have specific limits on staffing ratios and we do not allow a nurse to work more than one shift at a time. These are critical safety issues, therefore we have a hospital policy on it. One of our policies is openness and honesty with our patients. We tell our patients when we hurt them and we apologize. We treat all co-workers with respect. What does that mean? It means we do not allow hostile behavior, you do not throw instruments in the operating room. You do that twice and you are out! We do not tolerate humiliation of residents or nurses. That is not part of good education, it is not part of good safety, it is not part of a good environment. These are just things that come to mind. They are not written in stone, they are not written anywhere. The point is that the hospital specifies these as written policies and everyone knows what they are. Department policies would be more specific.
If we want to have a safe system, what we want the hospital to do is to adopt performance standards and have the understanding with all staff, that compliance with our standards is a condition of your being here. When you come on our staff, you agree to follow the rules. Here are the rules:
Adherence is a condition of employment. We monitor adherence. We have a system for collecting data to find out if a physician is in trouble, and then we try to do something about it. We have a whole repertory that we can use to help people because the purpose is not just to document their problem, but to remediate them, not to get rid of them, but to keep them in practice. If this system were working, you would not see many problems. Therefore, a measure of the effectiveness of your program is not how many licenses you pull, but how few you pull. It is the old ivory tower problem, so how do you get hospitals to do this? It is quite simple. You require it. Which is radical idea number three.
Require hospitals and departments to follow safety practices and policies. That is a neat idea, but I do not have any authority to do that. Maybe that should be agenda item number one: how we get that kind of authority without full legislative authority. It probably should be in statute form, so maybe that is the major agenda item. It is not going to happen otherwise. I am not sure that there is anyone else who is going to do it. That leads me to the fourth radical idea. If we are going to do this, then it does not make sense for every hospital and every clinic and every health care organization to develop their own policies. There is a tremendous benefit for having a national approach, and you, as the Federation of State Medical Boards, are ideally suited to do this. You have done some of this sort of thing already in some of the other programs, and so it is not exactly a new idea, but it is certainly a big job and I do not mean to imply that it is not. However, the real way for this to happen would be for you to take the lead in developing national performance standards. We have a lot of help. The American Board of Medical Specialties and the Accreditation Council of Graduate Medical Education have done a tremendous amount of good work in developing competency standards and measures for each of the medical specialties. A lot of that can just be moved right into this. There has been some good work in terms of monitoring and measuring patient satisfaction. Some of you are familiar with the work of Jerry Hickson, who uses a simple measure of patient complaints and shows that this correlates closely to risk of malpractice — and he has a program for working with doctors. There is a lot of material out there, and it needs to be brought together and reviewed to make sure it is valid and made available to the states. It could play an important role in this regard.
In summary, here are four radical ideas for a different way of thinking about your role, based on the concept that it is the system, and not the person, who is responsible. When it is the person, we need a system to deal with that person, and unless we focus on those systems and develop much better ones we are not going to get there. Before you reject this totally out of hand as another idea from the ivory tower, from someone who is not on the front lines, let me leave you with a thought: At its heart, safety is a moral issue. It is a moral issue because we know a lot about how to prevent injury, and we are not doing it. The moral questions are, “Why are we not doing it?” The moral question directed at you is, “What is your fiduciary responsibility to see that it gets done?” And the final question is, “If not you, then who?”
This article was originally delivered as the Galusha Lecture at the Federation of State Medical Boards 2002 Annual Meeting in San Diego.
