To improve patient safety, try treating problem physicians
By Edward Doyle
NEW ORLEANS—Talk about impaired physicians, and most people think of doctors battling substance abuse. At an Annual Session presentation on patient safety, however, Lucian Leape, MD, painted a much less dramatic—but more pervasive—picture of the problems that afflict physicians.
It's true that medicine has its share of physicians who suffer from drug and alcohol problems. Dr. Leape estimated that about 15% of physicians will be impaired by one of these problems at some point in their career, the same percentage affected in the general population.
He was quick to add, however, that the bigger threat comes not from chemical dependency, but from behavioral and competency problems that can lead to medical mistakes—and from the reluctance of American medicine to intervene. That reluctance undermines physicians' professionalism and creates a culture in which mistakes can thrive.
Take the example of Hank Goodman, MD, the fictional name of an orthopedic surgeon who was eager to build a successful practice. After working hard for 10 years, he enjoyed an excellent reputation and a busy practice.
But that ambition eventually proved to be his undoing. "He had a not totally atypical flaw," Dr. Leape explained during his talk on avoiding medical mistakes. "He couldn't say 'no.' " When Dr. Goodman became overwhelmed, he took the tried and true route, squeezing more and more patients into his already-packed schedule.
"He began to cut corners, have complications and went into a downward spiral," said Dr. Leape, adjunct professor of health policy at the Harvard School of Public Health. The surgeon eventually lost his staff privileges at the local hospital, and his career was in a shambles.
Dr. Leape used the story of Dr. Goodman, which is recounted in "Complications: A Surgeon's Notes on an Imperfect Science" by surgeon Atul Gawande, MD, to illustrate how the workaholic, go-it-alone attitudes of some American physicians can sabotage patient safety.
He also said that the tale of Dr. Goodman illustrated an even more widespread problem. For years, the surgeon's colleagues had overlooked the telltale signs, turning a blind eye to problems that should have been obvious. When his peers finally did decide to discipline him, they didn't address the deeper patient safety and behavioral issues.
Instead, they waited until he missed a morbidity and mortality meeting and, based on that technicality, revoked his hospital privileges.
The telltale signs
According to Dr. Leape, a long list of conditions can cause sub-par performance. While drug and alcohol problems top that list, he pointed to the threat from physicians with depression and other forms of mental or physical illness.
Performance problems can be caused by a variety of factors, including substance abuse, stress from overwork and cognitive decline due to aging.
Then there are doctors who, like Dr. Goodman, suffer from what Dr. Leape called "overworked stress syndrome," as well as physicians whose competency is slipping because of age or loss of cognitive function. And there are physicians who are disruptive or rude to patients, colleagues, residents or co-workers. Studies have shown that when co-workers feel intimidated by physician behavior, they are much less likely to question that physician's orders, eliminating a key safeguard in preventing medication mistakes and other errors.
Then there are physicians whose performance is impaired because of exhaustion. "Data show that if you are sleepless for 24 hours, your performance level is equivalent to that of a person with a blood alcohol level of 0.1," he said. "In effect, you are legally drunk." He said he faults the current resident work hour rules, which don't address the issue of lack of sleep within a 24-hour period—which is the critical metric. According to Dr. Leape, rules should restrict work hours to 12 hours within a 24-hour period, not 80 hours over the course of a week.
Altogether, he concluded, "We have an array of people who for various reasons may function in a way that threatens patient safety, perhaps as many as 30% or more of physicians over the course of their careers."
Part of what makes it so hard to help physicians is that there are no systems for dealing with problem doctors until their performance has become a serious threat to patients.
"We have a nonsystem," he said. "We tend to use a very personal one-on-one approach, which means the physician in question can immediately come back and claim you're out to get him because you're both competing for the same patients."
While it may seem obvious to blame individual physicians for poor performance, Dr. Leape said that's not the right approach. "Errors are not caused by bad people, they're caused by bad systems," he explained, echoing the conclusion of the Institute of Medicine's 1999 "To Err Is Human" report, which Dr. Leape helped write.
"Our objective is not to punish physicians for their mistakes," he said. "Our objective is to prevent mistakes, and the way you prevent mistakes is to change the system."
The good news is that there are telltale signs to help identify problem doctors, as well as simple and effective strategies to intervene.
Dr. Leape cited research published by Gerald Hickson, MD, in the June 12, 2002, Journal of the American Medical Association showing that patient complaints can play a critical role in pinpointing physicians' behavioral problems. (An abstract is online.)
Researchers found that most physicians—80%—receive no complaints from patients. However, physicians who regularly receive complaints, on the order of two or more a year, have a much greater chance of being sued for malpractice.
While the correlation between patient complaints and malpractice suits may come as no surprise, researchers found a surprisingly simple solution. When colleagues discussed the complaints with physicians, half of those physicians never received any more letters. Apparently, Dr. Leape explained, the discovery that there was a problem shook them up so much that they changed their behavior.
The role of performance measures
As far as larger-scale approaches to identifying physician performance problems, Dr. Leape said he is optimistic that the growing trend toward implementing performance measures—by groups such as the Accreditation Council for Graduate Medical Education, the American Board of Internal Medicine and others—will help.
"I want to survey or measure everyone, and I want to do it proactively before people get hurt," he said. "We need to enable people to realize their potential and correct their problems, and for that, we need data on performance. To get that data we need to develop measures."
Veterans Administration (VA) hospitals, for example, collect a standard set of data for every patient undergoing major surgery, he said. Experts then give physicians a profile of their risk-adjusted mortality and complication rates.
"As I understand this system," Dr. Leape said, "when physicians discover they are at the bottom end of the curve, they go to work on improving their results. The VA is not above giving them a little nudge if they don't. If you're at the bottom of the pile and you don't show improvement after a year, you may have a visit from someone at headquarters. They think they're doing a pretty reasonable job of assessing complications and mortality in a scientifically legitimate risk-adjusted way." The VA program has been so successful, Dr. Leape added, that the American College of Surgeons has enlisted a number of non-VA hospitals to begin similar programs.
To help identify problem physicians in teaching hospitals, Dr. Leape recommends a routine "360-degree" evaluation process, with each physician being assessed by residents, nurses and colleagues.
While the VA has the resources to collect and analyze vast amounts of data, Dr. Leape suggested a simpler route to gather data and help identify physicians with problems: a process he called a "360-degree evaluation."
Teaching hospitals, for example, could collect confidential evaluations of physicians from a group consisting of three residents, three nurses, three colleagues and a department chair. The idea is to get a well-balanced view of a physician's performance.
"You would get 10 assessments," he explained, "with everyone filling out the same form with questions about whether the person has ever been disrespectful to a colleague or a patient, whether there are concerns about the person's competency, or signs of chemical abuse."
If one of those evaluations flags a problem, Dr. Leape said, you don't have much to worry about. But if three or four reports cite issues, that's an indication that it's time to talk candidly with the physician and intervene.
"The idea is to measure everybody by the same yardstick and identify the people who are on the wrong end of the distribution curve," he said. "Then we give those people help."
Dr. Leape pointed out that physicians need more remedial programs, and he called on medical societies—which have taken the lead in developing alcohol and drug abuse programs for their members—to design programs to address behavioral impairment as well.
When the Institute of Medicine (IOM) issued its landmark report on medical errors in 1999, many health care organizations reacted by creating reporting systems to track adverse events. For many of those organizations, however, that's where their efforts to prevent errors begin and end.
Lucian Leape, MD, adjunct professor of health policy at the Harvard School of Public Health, was part of the committee that issued the IOM report.
At an Annual Session lecture on avoiding medical mistakes, he pointed out that in New York state, health officials receive more than 30,000 adverse event reports a year. Instead of helping prevent errors, however, those reports sit around in boxes because administrators don't have an adequate budget to analyze the information.
Dr. Leape attacked what he called the myth of adverse event reporting. People in and outside of health care seem to believe that the mere act of reporting an error or a near-miss will in and of itself help improve patient safety.
The rationale behind that belief, he said, seems to be that requiring physicians to report adverse events will shame them into doing better: "If we make them report, it will be so embarrassing that they wouldn't make so many mistakes." The problem with that approach? "It assumes that errors are caused by not trying hard enough—and we know that isn't true."
While he acknowledged that reporting can be effective in select instances, such as identifying hazards like a complication of a new drug, the reporting process by itself is unlikely to bring about major change.
To understand an adverse event and prevent future problems, he said, requires analysis. "You have to examine, investigate and talk to people," he explained. "The contextual information and the understanding of the various contributing factors are the part that leads to the ability to redesign the system."
Reporting systems also require serious funding to be effective. The aviation industry, for example, requires pilots to report near-misses. The industry spends about $100 to analyze each report, for a total annual budget of about $3 million.
Dr. Leape said that the mandatory reporting systems being used by about 20 states may help improve patient safety by holding individuals accountable and identifying those that need help. However, these systems operate at what he described "a gross level" and produce limited information.
"If a hospital has a real commitment to safety," he said, "it is going to get so much information that you don't need a reporting system. I can talk to any three nurses in a unit for an hour about what bothers them, and come out with a safety agenda that will keep me busy for a year."
Internist Archives Quick Links
New Leadership Webinars
The ACP Leadership Academy is offering FREE webinars covering the core tenets of leadership, leadership in hospital medicine, finance, and more.
Join ACP Today!
ACP membership connects you with like-minded colleagues and provides access to a variety of clinical resources, practice tools, and ways to earn MOC and CME.