Restricting the number of hours doctors-in-training are allowed to work without rest hasn't led to more patient deaths, according to a new study.
Researchers found no increase in deaths over the three years following a rules change that restricted resident doctors to working a maximum of 80 hours per week. In fact, the team reports a decline in deaths during the fourth and fifth years.
"This study is nice, because it shows that late after the 2003 changes there seems to be an improvement in mortality," said Dr. Sanjay Desai, director of the residency program at Johns Hopkins Hospital in Baltimore.
"But we don't know if this outcome is because of these changes or despite these changes. I think we need to know that to inform these types of decisions," Desai, who was not involved with the new study, told Reuters Health.
In 2003, concerns over errors caused by sleepy residents at hospitals led the Accreditation Council for Graduate Medical Education (ACGME) to restrict doctors-in-training to working a maximum of 80 hours per week.
The ACGME again restricted residents' working hours in 2011, when it said shifts can last no longer than 16 hours for the least-experienced doctors.
But reducing work hours for residents increases the number of times a patient's care changes hands, and there were concerns that would lead to more errors.
"Every time that occurs - just like every communication that occurs - there is some loss of information," said Desai. "The more hand offs there are, the more risk for an error to occur."
The new study's authors write in the Journal of General Internal Medicine that the new rules didn't have a great effect on the number of patients who died immediately after the change, but no one had looked at hospital death rates years later.
For the new study, the researchers - led by The University of Pennsylvania's Dr. Kevin Volpp in Philadelphia - used U.S. hospital data from before the first rules change in 2003, and compared that to data for the years 2003 to 2008.
Volpp and his colleagues looked at records for about 14 million people on Medicare, the government-run health insurance for the elderly and disabled, who were admitted to the hospital for heart attack, heart failure, gastrointestinal bleeding or surgery.
Overall, the number of deaths across conditions did not change significantly between the period 2000 to 2003 and the years 2003 through 2006, but deaths did fall during 2007 and 2008.
For example, about 17 percent of heart attack patients died within the 30 days following their hospital admission in 2000, before the rule change. About the same number of patients died immediately following the rule change in 2004, but that number fell to just 14 percent by 2008.
And about 10 percent of heart failure patients died within 30 days of being admitted to the hospital in 2000. As with heart attack patients, that number was about the same immediately following the rules change in 2004, but then fell to about 9 percent by 2008.
Desai said the results are reassuring, but the study cannot prove that the improvements were caused by the rule change.
"It's difficult, because it's retrospective. You can't identify how the change in policy correlates or relates to the changes they observed," he said.
Volpp and his colleagues could not be reached for comment before deadline.
Because the new study only looked at death rates through 2008, Desai told Reuters Health that it's also hard to say how this data would apply to the most recent change in maximum shift lengths dating from 2011.
"We as a community are highly interested in understanding this relationship better. These policies are in effect and changed relatively recently," he said. "I think there is a need to understand these relationships much more clearly."