When hospital inspectors are in town, fewer patients die, study says

Every couple of years, unannounced visits by inspectors from the Joint Commission trigger a frenzied response in hospitals, in which clinicians are instructed to do everything by the book to avoid potentially embarrassing violations.

On Monday, a new study revealed the impact of that all-hands-on-deck response when the accreditors are in town: significantly fewer patient deaths during inspection week than the weeks before and after the inspectors’ visit. Researchers at Harvard University found the disparity was particularly pronounced at major teaching hospitals, where the heightened response, if applied for an entire year, would translate to 3,600 fewer deaths.

The findings, published in JAMA Internal Medicine, did not link the lower death rate to adherence to specific best practices or standards. Rather, the researchers said, the improved performance appeared to a result from a more generalized vigilance caused by the arrival of the inspectors.

“The results suggest there are opportunities for quality improvement right under our noses that we take for granted,” said the study’s lead author, Dr. Michael L. Barnett, an assistant professor of health care policy management at the Harvard T.H. Chan School of Public Health. “Some creative thinking around how we react to Joint Commission surveys might lead to quality improvement.”

Barnett, a primary care physician at Brigham & Women’s Hospital, said he was motivated to conduct the study, in part, by what he said was the absurdity of inspection week, when hospitals put on a show for visiting inspectors who aren’t really getting an accurate view of normal operations.

“You’re not really measuring the quality of the hospital,” Barnett said. “You’re measuring the capacity of the hospital to mobilize the troops.”

The Joint Commission performs unannounced inspections of US hospitals every 18 to 36 months to test their adherence to quality and safety practices in a broad range of areas, including infection control, accuracy of documentation, and medication management. The not-for-profit organization accredits 21,000 health care facilities across the country. Failing an inspection could cause a facility to lose its accreditation and to suffer a big hit to its reputation.

Failure is exceedingly rare, but to avoid it, many hospitals quickly notify their staffs of an inspection and hold training sessions to optimize performance. Barnett said more than 98 percent of facilities passed, with only a tiny fraction receiving a citation.

The study, which examined outcomes for 1.7 million Medicare patients, found that patients who happened to receive care during inspection week had a better chance of surviving. Overall, the reduction of 30-day mortality rates across all hospitals was modest — about 1.5 percent.

But the reduction was significantly larger, about 5.9 percent, at major teaching hospitals, which rely the most on maintaining their reputations as prestigious institutions on the vanguard of medicine.

“They have a lot at stake if they don’t get accredited by the Joint Commission, and they also have the resources to mobilize their staff hospital-wide to respond,” said study coauthor, Dr. Anupam Jena, a professor of health policy at Harvard.

While the overall improvement in mortality rates was modest, he said, they can have a significant impact when spread across hundreds of thousands of visits each year from Medicare patients. “If you aggregate over the entire US population of Medicare beneficiaries, it’s a large impact,” he said.

The study examined outcomes in patients admitted from three weeks before the inspection to three weeks after its completion. The researchers adjusted for demographic and clinical differences among the patients included in their examination.

The study also examined infection rates and other quality measures, but did not find a significant difference at the hospitals during inspection week compared to the other periods reviewed.

Jena said that latter finding suggests that the difference in mortality outcomes was not related to any one or two specific practices, but to a more heightened vigilance that occurs during inspection. “When you’re being monitored you behave differently and you do your job differently,” he said, adding that hospitals could adjust their schedules and practices to help harness that increased focus and attention to detail.

“One thing you could do would be to ensure that when key clinical decisions are being made, you have a disruption-free policy, so that doctors and nurses cannot be disrupted unless it’s an emergency,” he said. “Changes in shift work and the amount of hours doctors work, and the staggering of those hours, could also have an impact.”