AMA Journal of Ethics^®

Michael Swango was a physician—and—a serial killer.¹ Swango killed hospitalized patients, those under his care or others just in the wrong place at the wrong time. To say that a serial killer is too personally culpable and too extreme to provide meaningful lessons on patient safety ignores the critical patient safety and system issues that we need to acknowledge and address to prevent harm to patients.

Clinical case 2 in this issue, "Problem Peers," brings up the common scenario in which a physician represents a safety concern that peers are aware of. This scenario occurs everywhere, at every level of medical education. The Swango case illustrates how, even when the worst happens, the system is not effective at dealing with problem physicians.

Michael Swango was a problem student and a problem resident. His medical school classmates feared for the safety of Swango's patients, calling him "double-O Swango," meaning he had a "license to kill."² The high mortality among Swango's patients was noted by the students and residents he worked with as soon as he began clinical training.³ He was dismissed from a residency position after the hospital suspected him of poisoning a patient—not through poor medical care but by injecting something into the IV of a patient who was not under his care.⁴ Swango was later convicted of poisoning coworkers with arsenic and served prison time.⁵ Swango was able to secure 2 subsequent residency positions where he continued to poison people. He finally went to Africa to practice and is suspected of committing additional murders there.⁶

Most problem physicians aren't depraved people like Michael Swango. But many good doctors provide care while impaired, most often by sleep deprivation.⁷ There is evidence that after a night on call, an event that happens dozens of times in every resident's training, residents are impaired to the level of people with sleep disorders.⁸ Sleep deprivation is a well recognized source of cognitive impairment in physicians.⁹ As the "Problem Peer" case and commentaries indicate, however, dealing with a peer who is routinely unsafe is difficult.

Most physicians probably have contact with an unsafe physician at least once in the course of their career. As a profession, medicine enjoys a great degree of self-policing and self-regulation. With the privilege of self-regulation comes the responsibility to regulate adequately. The "Problem Peer" case discusses how difficult adequate regulation is in common situations, while the Swango case shows that it isn't easier (or even more likely to occur) in the most extreme case.

Although confronting a colleague is never easy, we need to do more of it. Every health care professional needs to feel personally responsible for any patients cared for by impaired or negligent clinicians. I believe that every professional who stays quiet when faced with impairment or incompetence is morally culpable for any injury caused by the impaired or incompetent individual. If I know my colleague is too tired to be performing a particular procedure, but I don't intervene, I am responsible for any injury that results. If a patient gets hurt it is my fault, just as it is the fault of my impaired colleague and my other colleagues who knew of the problem but didn't intervene.

The blame-free approach to patient safety has become very popular, and it is crucial to detecting and improving system deficiencies, reporting "near miss" incidents, and correcting errors before a patient is injured.¹⁰ However, not all errors should be protected from blame or punitive measures. Protecting patients from problem colleagues requires that all physicians and other providers accept and act on the personal, individual duty we have to every patient and every future patient.

Sometimes physicians are viewed by the public as being more interested in protecting each other than protecting patients. Physicians have been described as having a "brotherhood of silence" where no member of the profession will speak out against another member of the profession, at least not outside the profession. While this has not been true in my experience, the public perceives it to be true at times. Incidents like the case of Michael Swango, where a serial killer remained responsible for patients years after colleagues began to be concerned about him, foster public distrust.

Fear of litigation is one argument offered for keeping quiet about problem colleagues. Physicians fear that speaking up creates a basis for a lawsuit. If one physician is vocal about another physician's perceived impairment or incompetence, does the first physician's voiced concern create a basis for the second physician's patients to sue? Does documentation of a problem create a legal record that a problem exists? This commentary does not purport to explore these legal intricacies nor to answer these questions. Instead I propose that we have an obligation to be vocal even if it does generate liability.

Whatever your feelings are about how the tort system functions in reality, the underlying basis for tort lawsuits is just compensation to a victim for harm caused by another. The purpose is to shift the financial consequences of injury from the person injured to the person at fault for the injury. If a patient is injured by a physician who is drunk or who is incompetent, justice demands that the physician, and not the patient, bear the financial burden of that injury. Failing to vocalize concerns about problem peers in order to minimize liability means keeping quiet for purposes of denying injured patients just compensation for the financial burden of their injury. It means valuing the impaired physician's financial interest over the patient's financial interest. The tort system may be flawed in practice, but a case where a patient is injured by an impaired or incompetent physician is exactly the type of wrong that the tort system should address.

A system of self-regulation requires a high level of commitment to effective self-regulation. Improving patient safety requires a commitment to meaningful self-regulation as much as addressing systems issues and latent errors. Terrible failures in this area only serve to highlight the importance of making health care safer.