This post first appeared in a slightly altered form in Axis of Logic as http://axisoflogic.com/artman/publish/Article_64538.shtml
I’m a retired neurosurgeon who served as a chief of staff and governing board member of two separate 400 bed community hospitals over a five year period . I observed several doctors at each institution who repeatedly hurt their patients. Their sins included refusing to show up in an emergency, lying to patients and other doctors, missing diagnoses, getting poor operative results, doing unnecessary surgeries and extracting sexual favors from patients. My efforts were opposed by my colleagues, the hospital administrators and the hospital lawyers. I sensed that protecting bad doctors was a significant cause of the unnecessary loss of 98,000 lives a year from harmful adverse events. To Err is Human, a book by the Institute of Medicine revealed that stark figure to a shocked public in 2000.
I had to do some lifting and probing before I could turn my intuition into facts arguments. My research into why my experience was nearly universal led me to someone who mined the National Practitioner Data Bank, Dr. Robert Oshel, who worked for the data bank before he retired. That repository was established by Congress in 1986 and put into effect at the end of 1990. It lists for the public, by a number only, those physicians who have been sued successfully or have lost their license to practice. Here’s what Dr. Oshel found over a period of 20 years:
1. There is a hard core of 2% of the physicians whose misdeeds result in half of the money paid out in malpractice cases. This is the best measure of the size of the group of doctors who repeatedly hurt their patients that I could find.
2. The average hospital drops only one doctor from its staff every twenty years.
3. About 250 doctors lose their licenses each year, or 0.04% of the total. At that rate it would take 50 years to remove the hard core 2% from practice.
There is a potential cause of undue leniency to doctors who repeatedly hurt patients inTo Err is Human itself. Lost in the furor over the death numbers was that To Err Is Human had blamed bad health care delivery systems, not bad doctors for the deaths of the 98,000. It emphasized “creating safety systems inside health care organizations through the implementation of safe practices at the delivery level. This level is the ultimate target of all the recommendations.” As a result of the publication of To Err Is Human all of the efforts to save the lives of the victims of hospital adverse events have been systems based. The authors had reversed the interpretation of the original data of their own published findings on theoretical, not empirical grounds. No new studies were done so that no new data were gathered to contradict the original.
The Harvard Medical Practice Studies, the primary source on which To Err is Human is based, showed that at least 61% of harmful adverse events in hospitals were the result of blunders by individual physicians and that systems errors were responsible for only 6%. Documentation that the systems correction approach was not working came in two reports that came out in November, 2010, one from the Inspector General of HHS, based on data gathered on Medicare patients and the other from Harvard Medical School based on outcomes in several North Carolina hospitals. That state was chosen because its hospitals had the reputation of rigorously following systems error prevention methods. Both reports said that in the first decade of this century, the period following the publication of To Err is Human, the number of deaths from adverse events was unchanged from the 15 years preceding, perhaps greater, and had stalled at 120,000 per year.
I questioned the prime mover behind both the Harvard Medical Practice Studies and To Err is Human, Dr. Lucian Leape, in a series of e-mails. Dr. Leape of the Harvard School of Public Health is one of the most influential American doctors in the quest to make our hospitals safer. We discussed the virtues of the blame vs the systems approach of dealing with doctors who err repeatedly. He took full responsibility for the reversal of the conclusions of the Harvard Medical Practice Studies. Dr. Leape is convinced of the futility of a shame, blame and punish approach, but he also said that no one in our profession was interested in taking a serious look at reforming poorly performing doctors. The systems method, which he advocates, holds that medical errors are due to flaws in medical care delivery systems, not flawed doctors. The systems approach advocates continuous measurements of things like making sure every heart attack patient gets an aspirin and beta blocking medications on admission to the hospital and polyvalent pneumonia vaccine on discharge for all patients over 65.
Dr. Leape, although he is a systems advocate to the core, takes a more balanced approach than others in the field of medical mistakes, having written on unnecessary surgery and re-educating bad physicians. But one of the original systems proponents, Professor James Reason of Manchester, England, a psychologist, has said that disciplining doctors is irrelevant to making hospitals safer.
In 2008, the state boards of medicine in the U.S. and Canada had bought into the systems approach. They announced in a joint statement that they accepted that doctors are often the victims of flawed medical care delivery systems and that this is a much more frequent source of poor outcomes than the doctors themselves; this must be considered in dealing with the physicians who are reported to them for alleged professional failures. They said, “Systemic sources of risk significantly eclipse professional incompetence as the dominant cause of harm to patients.” The evidence from the Harvard Medical Practice Studies was otherwise.
I told Dr. Leape of my concern that his two most prominent progeny, the Harvard Medical Practice Studies and To Err is Human, contradicted each other. I said that his younger child, To Err is Humansays, “Look for the bad systems, not the bad doctors.” But that is at variance with a key statistical table in the Harvard Medical Practice Studies. The table says that most patients harmed in the system were affected by errors caused by individuals, not systems. He said that I was quoting the table correctly but that’s all I had gotten right. My interpretation would have been right in 1991when his group published the Harvard Medical Studies which contained all the scientific data. Back then they felt like I did, that errors were individual failures, and, sure enough, they found them.
I asked him if there was new data that caused the reversal of opinion. He said no, that he was “blown away” by the human error literature written by psychologists such as Dr. Reason and he changed his mind. And that gave birth to To Err is Human. There had been no subsequent studies that had justified the turnaround. We also discussed the two November, 2010 papers, one from Medicare and the other from Harvard which showed no improvement in the number of fatal adverse events in American hospitals. Dr. Leape felt that the problem was one of compliance, that good leadership and financial incentives were necessary.
It boiled down to this: The reversal of opinion was made on theoretical not empirical grounds. No new data had been gathered to reverse the original interpretation. In Galileo’s time the dogma of the geocentric solar system held sway on ideological grounds only. Galileo made the thousands of meticulous observations that gave rise to the heliocentric concept of the solar system. The systems approach adopted by the authors of To Err is Human was assumed a priori to work in medicine because it had worked in the airline and auto industries. But the systems approach, while it was a huge and rapid success in preventing airline mishaps and poorly manufactured cars, has made only a small dent in the number of unnecessary deaths in American hospitals.
That tiny inroad has been the contribution of Dr. Pronovost of Hopkins who is largely responsible for the reduction in deaths from infected central IV lines. According to the CDC, Dr. Pronovost’s efforts have resulted in a reduction of 3 to 6,000 deaths per year. That sterling result got swamped by all the other causes of adverse events. A fair question to the American medical establishment would seem to be, “Why are we continuing an approach that seems not to work?” We have no time to lose. A million people, at least, will die unnecessarily in the next decade unless we remove incompetent doctors from practice on a much larger scale than the minute number we do now.