Did reducing hours for residents increase medical errors?

There was a sigh across the country in 2011 when a Quebec arbitrator ruled that the 24-hour shifts required of medical residents violated the Charter of Rights and Freedoms. Why were young doctors working inhumane hours with vulnerable patients at their semi-lucid mercy? Doctors shared stories of car accidents after long shifts and of mistakes made in the daze of exhaustion.

Key thinkers on health care called it a common-sense ruling. A few months later, Quebec capped on-call duty for residents at 16 hours, down from 24. With mounting evidence about the adverse effects of long-call shifts, it seemed like a move the rest of Canada should follow.

Or maybe not, new research suggests.

Two studies in the Journal of the American Medical Association argue that the 16-hour limit might not actually improve the lives of residents or the safety of patients. In fact, it may make matters worse.

A longitudinal cohort study, published in JAMA, used a first-year resident health survey to track the effects of 2011 duty-hour reforms in the U.S. (which, like Quebec, limited call to 16 hours but only for first-year trainees). Dr. Srijan Sen, a psychiatry professor at the University of Michigan who led the study, told Science-ish that his data busts assumptions. Shorter shifts did not lead to more sleep — nor did it improve the well-being of residents. In fact, they made more medical errors. “There are unintended consequences—negative consequences—associated with capping hours,” he said.

For example, most hospital residency programs didn’t have the resources to hire new physicians or physicians’ assistants, so residents were expected to do about the same amount of work in less time. Dr. Sen linked the resulting “work compression” to medical error. “There’s clear evidence that working so long isn’t good for cognitive functioning. But it looks like we’re creating new problems by cutting down those shifts.”

The second JAMA study looked at medical house staff at the Johns Hopkins Hospital, randomly assigning them to a 30-hour cap, or to one of two groups with a 16-hour cap. The investigators used wrist watches that measure movements to find out if the groups working fewer hours managed to catch more sleep. The lead author, Dr. Sanjay Desai of Johns Hopkins, told Science-ish that the interns who worked less slept an average of three hours more around the period during which they took call, but otherwise got no more sleep than the control group.

“Is three hours enough sleep to change levels of fatigue and response times?” he asked. “For us, this introduced potential flaws in the logic that if you cap hours, people will sleep more, and meaningfully more. That doesn’t seem to hold true based on the data we have.”

There were other alarm bells. Dr. Desai says residents who worked night shifts felt their education was compromised, since learning and educational activities generally slow down at night. Hand-offs in the 16-hour groups increased between 130 to 200 per cent compared to the previous 30-hour model. As a result, patients were juggled between more doctors than before. (It’s well known that transitions of care can be akin to a game of broken telephone, a major source of medical error.)

So what now?

Dr. Chris Landrigan, a professor at Harvard Medical School, who did a systematic review on the effects of the 16-hour call, said most literature points to reducing shifts, and notes serious limitations in the designs of the new studies. The first by Dr. Sen relied on self-reporting of medical errors, well-being and sleep patterns. Dr. Desai’s study also made no direct measure of patient safety. Instead, investigators looked at sleep and hand-offs, which weakens the argument that caps lead to more errors.

Still, some of the findings in the JAMA studies have been illustrated by the Quebec experience.

Dr. Charles Dussault is president of the Fédération des médecins résidents du Québec, which represents the province’s medical residents. The FMRQ is trying to find ways to supplement education in the OR after surgical residents complained they weren’t learning enough. “OR time is precious,” he explained. “Some of the models we proposed limited the number of hours residents can spend in the OR.”

Quebec doctors have also reported concerns about the increase in hand-offs. “People are feeling the fact that there is more patient transfers than there were before could increase the risk of mistakes.” Dr. Dussault said hospitals are looking for ways to improve hand-offs. “We are still in transition,” he added. “People had the same debate when they went from a 72-hour cap to 36, from 36 to 24, and now from 24 to 16.”

And that’s exactly it: shift length is a systems challenge that requires a systems solution. Resident duty hours need to be more humane, but cutting hours without redesigning the hospital infrastructure—standardizing hand-offs, addressing work compression or paying attention to resident education—won’t get doctors very far.

Both Drs. Sen and Desai pointed out that it wasn’t just the shrinking of shifts that may have increased errors; it was the lack of planning and organization around the changes. Importantly, they cautioned against the 16-hour limit, noting it was too prescriptive — that a one-size-fits-approach is not suitable for every hospital.

In Canada there is currently no consensus on the regulation of duty hours. We can do better. Somewhere between the U.S. and Quebec is a model for the way forward. 

Science-ish is a joint project of Maclean’s, the Medical Post and the McMaster Health Forum. Julia Belluz is the senior editor at the Medical Post. Got a tip? Message her at [email protected] or @juliaoftoronto on Twitter.