Cutting hours for hospital interns may not lead to fewer mistakes after all.
A typical call day for first-year residents -- known as interns -- lasted 36 hours before 2003, and 30 hours until 2011. Regulations now limit first-year trainees at teaching hospitals to 16-hour shifts in order to give them a break. However, two new studies in JAMA Internal Medicine on March 25 show that less time at the hospital doesn't necessarily mean fewer errors.
The first study, which was conducted by Johns Hopkins University researchers in Baltimore, Md., showed that despite not having to spend as much time at work, shortening shifts to 16 hours did not increase the number of hours of sleep each medical intern got, and reduced the amount of training time they received. The shorter shift also increased the amount of handoffs of patients from one trainee to another, which could increase risk for errors.
"The consequences of these sweeping regulations are potentially very serious," lead author Dr. Sanjay V. Desai, an assistant professor of medicine at the Johns Hopkins University School of Medicine and director of the internal medicine residency program at The Johns Hopkins Hospital in Baltimore, said in a press release. "Despite the best of intentions, the reduced work hours are handcuffing training programs, and benefits to patient safety and trainee well-being have not been systematically demonstrated."
Desai's team looked at three different work schedules among 43 interns at Johns Hopkins Hospital in the months before the 2011 ruling of 16-hour shifts was put in place. Medical interns worked either under the 2003 model of being on call every fourth night with a 30-hour shift limit, being on call every fifth night and working for 16 hours straight, or a "night float schedule" of working a regular week on the night shift not exceeding 16 hours a time. They had their schedules for three months and were instructed to wear a special wristwatch that tracked arm movements measure sleep patterns.
The interns on the 16-hour limit schedules slept on average three hours longer during their on-call periods compared to the 30-hour group, but there was no difference in the overall amount of sleep they got in a week.
"During each call period, the interns had 14 extra hours out of the hospital, but they only used three of those hours for sleeping," Desai said. "We don't know if that's enough of a physiologically meaningful increase in sleep to improve patient safety."
Those working the 30-hour shift had a minimal number of three patient handoffs between interns. For the 16-hour shift subjects, they had a minimal number of nine. More handoffs leave more room for errors, according to the researchers. The minimal number of different interns caring for a patient during a three-day stay increased from three to five with the shorter shift, but it is unknown if that has an effect on the quality of care.
Trainees and nurses said that people working the 30-hour shift model had a higher quality of care. Those on the "night float schedule" had such a low quality of care that their type of schedule was stopped early. In addition, the time for rounds -- the educational period of an intern's shift when senior physicians quiz trainees about different patients at their bedside and give the interns instructions on how to properly care for different cases -- was cut in half for the 16-hour shift people because schedules were much shorter. Interns were also unable to follow a patient throughout the first day they were admitted because of their shorter schedules.
"Dramatic policy changes, such as the move to 16 hours, without a better understanding of their implications are concerning," Desai said. "Training for the next generation of physicians is at risk."
The second study, which was led by researchers at the University of Michigan in Ann Arbor, looked at 2,300 doctors in their first year of residency at more than 12 hospital systems around the U.S. It showed that the number of medical errors that caused harm to patients went up after the 16-hour shift implementation.
The researchers sent out surveys in 2009, 2010 and 2011 to interns every three months during their first residency year that looked at mental health, overall well-being, sleep habits, work hours and performance on the job. They then compared the results of the interns who were working before the new 2011 implementation (meaning they could not work more than 80 hours a week or 30 hours in one shift) to those placed in the 2011 16- hour shift limitations.
"In the year before the new duty hour rules took effect, 19.9 percent of the interns reported committing an error that harmed a patient, but this percentage went up to 23.3 percent after the new rules went into effect," first author Dr. Srijan Sen, a psychiatrist at the University of Michigan, Ann Arbor, said in a press release. "That's a 15 to 20 percent increase in errors -- a pretty dramatic uptick, especially when you consider that part of the reason these work-hour rules were put into place was to reduce errors."
Fifty percent of the medical errors were medication errors, 20 percent were missed diagnoses, 20 percent were incorrect treatment and 10 percent were surgical or procedural errors. Twenty percent of the residents screened positive for depression as well.
Co-author Dr. Sudha Amarnath, a resident in the radiology oncology program a the University o Washington, said that many interns felt that they had to do the same amount of work that they had in the longer workweek in fewer hours. The subjects said they had less "downtime" than pre-2011 rules.
"The 2011 changes were a pretty radical shift," Sen said. "Doctors have worked 30-hour shifts for decades, and it may just take time for all parts of the health care system to get used to the new rules and adjust."
However, Dr. Christopher Landrigan a pediatrics professor at Harvard Medical School -- who was not involved in the study -- pointed out to the Los Angeles Times that because the interns reported their errors in both studies, the quality of their work could be subjective.
"None of us are very good judges when it comes to determining the frequency of our errors," Landrigan said. "It's hard to tell if the answers are real, or if they might be influenced by personal recall bias."