Commercial airline pilots are subject to strict rules about the number of consecutive hours they may be on duty, as well as the number of hours that must elapse between their sessions actively piloting planes. The reason, obviously, is to ensure the safety of airline passengers, because the physiological and psychological effects of sleep deprivation are well known. It therefore comes as a surprise to learn of a recently launched study designed to compare the effects of increasing medical residents’ consecutive duty hours with the shorter, currently prescribed limits on their shifts—particularly for first-year residents—caring for patients.
The study, Individualized Comparative Effectiveness of Models Optimizing Patient Safety and Resident Education, or iCOMPARE, is sponsored by the National Heart, Lung and Blood Institute, and is being carried out in more than 60 U.S. internal medicine training programs.
Two main ethical questions arise in this research: Do the risks—both to patients and to medical residents—outweigh the potential benefits? Is the research ethically acceptable without obtaining informed consent for participation either from patients or from residents? In this blog post I address the first question, about risks and benefits; in a subsequent post I will discuss the issue of lack of informed consent.
According to the study protocol, the primary hypothesis of the research addresses the safety of patients: it assumes that 30-day patient mortality under the increased residents’ duty hours (allowing shifts of 30 or more consecutive hours) will not significantly exceed patient mortality under the current mandated standard for medical residents on duty (for first-year residents, a maximum of 16 consecutive hours). In other words, will more patients die under the care of medical residents who work longer stretches of time without sleep than is now permitted under current prescribed limits? To be sure, the study includes numerous other measures, including the perceived effects of sleep deprivation on the residents themselves (although there is no rigorous measure of resident safety outcomes), patient safety indicators other than mortality and factors related to residents’ professional education.
A great deal is already known, not only about the risks of sleep deprivation in general but also about the adverse consequences of long duty shifts on both residents and the patients they care for. For residents, studies have found that long duty-shift duration significantly increases the risk of motor vehicle accidents, needle-stick and other similar injuries and depression.
One study showed that medical interns made 35.9 percent more serious medical errors during a schedule with shifts of 24 or more consecutive hours than on a reduced schedule with shifts of 16 hours or less. Interns also made more serious medication errors and diagnostic errors when on the lengthy schedule than on the reduced schedule. Another study looked at sleep and attentional failures of interns in intensive care units and concluded that reducing work shifts from 30 to 16 consecutive hours significantly increased sleep and decreased attentional failures during night shifts. A compelling statistic compares the effects of sleep deprivation with the effects of alcohol blood levels. And another study found that moderate sleep deprivation, of less than 20 hours without sleep, produced impairments equivalent to legal levels of alcohol intoxication, including a higher level than is legally permitted to drive a car. Finally, an Institute of Medicine study concluded that it is unsafe for the average resident physician to work for more than 16 consecutive hours without sleep.
If these are some of the risks of increasing residents’ shift hours without sleep, what are the anticipated benefits the study seeks to determine? According to the iCOMPARE protocol, the potential benefits to residents of the longer shifts are increased contact time with patients, increased educational opportunities and residents’ greater satisfaction with their training experience. It is hypothesized that there will be fewer medical errors with the longer shifts, since it is known that errors often occur during “handoffs,” when residents transfer patients to the next shift of residents.
But is a study like this the only way—or even the best way—to find a method to reduce errors that occur during patient handoffs?
The first point is that the number of errors can obviously be reduced by decreasing the total number of handoffs by mandating longer shifts. But the number of errors per resident during a transition is not likely to decrease and may well increase, given the longer hours residents are deprived of sleep. When well rested, residents may be less likely to make errors during handoffs than when sleep-deprived. So even if the number of handoffs goes down with longer shifts, the quality of each handoff may deteriorate.
Before embarking on a risky and unethical study to increase the time medical residents spend on long, sleep-deprived shifts, it would behoove the medical training establishment to look to other professions that place people at risk to see how they have succeeded in reducing those risks. A great deal is known about how to reduce human error in a variety of occupations where the safety of operators is a major concern, such as the airline pilots mentioned earlier. The underlying assumption—that the only choice is between long shifts with fewer handoffs or shorter shifts with more handoffs—is a false dilemma.
It may appear cynical to question the motivation for this study. Yet I cannot help thinking that the underlying (and unstated) purpose is to provide better continuity of underpaid labor in today’s highly stressed hospital-based medical system in the United States. Why are program managers of residency training programs eager to climb on board? They have staffing problems; patients need care; attending physicians are home on weekends; medical residents are low in a hierarchical system in which they can be commandeered to work longer hours. Why not join a study that appears to benefit everyone—patients, the residents themselves and the hospitals that employ them?
I conclude that this is an unethical study, based on risk–potential benefit considerations, and will follow up in a subsequent post to suggest that absence of informed consent from patients and medical residents signifies a dangerous trend in so-called “comparative effectiveness” research.