It usually happens around 5 a.m. after the night-shift adrenaline has worn off. A certain ER doctor sits in front of his computer finishing charts, when his shoulders slowly start to sag, his head lolls, and then he has nodded off to sleep, his loud snores reverberating through the ER.
After several minutes of amusing commentary from the staff, someone usually elbows the doctor in the side (or, once, put smelling salts under his nose) and he awakes with a jolt. “Great power nap!” he smiles, leaning back and running his hands through his hair. Then he bounds off energetically to see the next patient.
On its face, it seems scary to think that your medical provider might still be foggy with sleep while he’s caring for you. But it could be argued that such mid-shift catnaps may actually be protective. Napping may be the answer to certain hospital safety problems that have vexed patient-safety experts and sleep scientists for years: the medical errors caused by overtired doctors and nurses on long shifts.
Probably the most famous example is Libby Zion, the 18-year old daughter of a well-connected newspaper reporter, who died in 1984 after being treated by a sleep-deprived medical intern at New York Hospital in Manhattan. Although the extent to which Zion’s hospital treatment contributed to her death was never determined, the publicity that followed led to an overhaul of the structure of medical training in the United States.
At the time of Zion’s death, doctors in training regularly worked 100-plus hour weeks, punctuated by 36-hour blocks of “call,” during which they worked all day, through the night, and then another full day without any meaningful rest. For those of us who trained this way, call nights were brutal demonstrations of perseverance, fueled by raw adrenaline, gallons of coffee and the knowledge that every experienced doctor had trained this way before us. We got through those nights with pride — but were we making mistakes because we were too tired?
New York State didn’t wait to find out the answer. In 1989, interns and residents who trained in New York were restricted to 24 consecutive hours in the hospital and 80 total work hours per week (on average). In 2003, nearly 20 years after the Zion case, the national organization that regulates medical education implemented similar, but somewhat watered-down, rules for the entire country.
Many experts expected medical care would be improved in hospitals after the change in work rules. That assumption makes sense considering that sleep-deprived people tend to perform worse than the well-rested: one study found that after 24 consecutive hours of wakefulness, people’s motor skills and judgment are as impaired as if they had a blood alcohol level over the legal driving limit.
However, the data on whether the new work limits save lives isn’t so clear. Clever studies have compared death rates before and after the regulations were instituted in teaching hospitals, and also compared those trends to non-teaching hospitals, which presumably shouldn’t be affected by work-hour restrictions for doctors in training. In large national investigations of surgical patients and medical patients, death rates didn’t drop after the rules were implemented.
The findings seem counterintuitive: if sleepy doctors make more mistakes, then shouldn’t restricting work hours help save patients? The theories as to why not vary. First, perhaps the new rules don’t restrict hours enough: they still allow residents to work up to 30 hours in a row and 80 hours per week — double the time the average American worker spends on the job. Second, some hospitals may not be following the rules (they really aren’t well enforced). Last, shorter shifts mean more potentially dangerous handoffs, wherein doctors and nurses transfer the care of their patients to a new shift worker who is not as familiar with the patients’ histories and may be less emotionally invested in their care. Handoffs are notoriously fraught with miscommunication and are known to create opportunity for mistakes.
A strong case can also be made that tired doctors and nurses do make more mistakes, but that they’re typically not serious enough to lead to higher rates of death. This is where performance studies become relevant. For example, in 2011, a thorough examination of more than 700,000 ambulance care episodes explored how long it took for paramedics and emergency medical technicians (EMTs) to care for and transport emergency patients to the hospital. The results: the emergency care workers took significantly longer to do nearly everything when they were at the end of a long shift. This and other studies suggest that the fatigued-health-care-worker problem is real and worrisome, but may not be measurable by looking at aggregate death rates.
Perhaps the work-hour cap was just too blunt a fix for such a nuanced problem. To really ensure that residents get rest when they need it, which would help them think clearly when they need to, maybe the solution is to institute napping on the job. Sleep scientists have demonstrated that naps (even as short as one hour of real sleep) can prevent performance error. One 2006 study randomly assigned medical residents to take optional naps and found that on average, they actually did manage to get extra rest. A newer, hopefully more definitive, study is currently ongoing, testing the effect of required naps: residents hand off all duties — the pager gets forwarded and the phone is turned off — for short periods of time in the middle of the night. In this study, the residents who are assigned to the mandatory nap don’t have the option to “tough it out” in order to help their team or demonstrate their mettle.
For those of us who trained under the old never-sleep, always-take-care-of-your-own-patients-at-all-costs, tough-it-out system, restrictions on work hours seem soft. Napping in the middle of a shift? That’s a sign of downright weakness. But this persistent macho attitude is part of the problem. Sleep science and studies of shift workers in non-medical disciplines have repeatedly shown that tired workers not only make more mistakes, but also often fail to identify their own fatigue. Letting tired doctors and nurses take naps, or even forcing them to, may be a workable solution.
In the coming years, as new data emerge from the ongoing studies, we will see how napping impacts patient safety and outcomes. But until then, on your next visit to the hospital when your doctor shows up with bed-head and bloodshot eyes, don’t be too alarmed. Hopefully she’s just woken up from a nap.
Meisel is a practicing emergency physician and an assistant professor of emergency medicine at the Perelman School of Medicine at the University of Pennsylvania. Follow him on Twitter at @zacharymeisel. Pines is the director of the Center for Health Care Quality and an associate professor of emergency medicine at George Washington University. Follow him on Twitter at @DrJessePines.