News Release

UMHS study finds surgery safer at teaching hospitals

Higher procedural volumes account for fewer deaths, shorter hospital stays

Peer-Reviewed Publication

Michigan Medicine - University of Michigan

ANN ARBOR, Mich. – Having a less experienced surgical resident or even a medical student involved in complex surgical procedures may cause some patients to feel uneasy.

But a new study by University of Michigan Health System researchers has found patients undergoing complex gastrointestinal surgery at teaching hospitals are less likely to die or experience complications than those patients at non-teaching hospitals, primarily because teaching hospitals tend to perform these surgeries more often.

The study, published this month in the Archives of Surgery, looked at 6,685 people who underwent surgery in 1996 or 1997 to remove part or all of their esophagus, pancreas or liver. The researchers compared death rates and length of hospital stay for teaching hospitals with those from non-teaching hospitals by looking at patient data from 1996 and 1997 from the Nationwide Inpatient Sample, which tracks 20 percent of all hospital discharges in the United States.

Death rates overall for all three procedures were under 10 percent. But patients undergoing pancreatic resections at non-teaching hospitals were more than twice as likely to die following surgery compared to patients at teaching hospitals. Patients undergoing liver resection incurred an 8 percent mortality at non-teaching hospitals, compared with 5.3 percent at teaching hospitals. Patients undergoing esophageal resection fared slightly better at teaching hospitals, with 7.7 percent death rate, compared with 10.2 percent at non-teaching hospitals.

Similarly, patients were less likely to have lengthy hospital stays at teaching hospitals, suggesting fewer complications from surgery, the study found. Patients who had their esophageal or liver surgery at a non-teaching hospital were 30 percent more likely to have prolonged hospital stays.

"One of the fears that many people have is that going to a teaching hospital with medical students and residents might hinder their care because the attending physicians have to teach rather than perform surgery. This study showed that perception is completely wrong," says study author Gilbert Upchurch, M.D., assistant professor of surgery at U-M Medical School.

When researchers looked at how often each hospital performed these complex surgeries, they found experience was an important factor. Each surgery had a small threshold for high volumes. Teaching hospitals were more likely to see high volumes: 79 percent of teaching hospitals performed nine or more pancreatic resections, 75 percent performed 10 or more liver resections and 70 percent performed more than six esophageal resections.

Non-teaching hospitals, however, tended to perform these surgeries less often, with only 17 percent of non-teaching hospitals performing large numbers of the pancreatic surgery, 11 percent seeing high volumes of the liver surgery and 20 percent frequently performing the esophageal surgery.

Previous studies have shown that high hospital volumes lead to fewer deaths or complications following surgery, particularly complex gastrointestinal procedures. This could be because the physicians have more experience or advanced training or because the hospital has better diagnostic equipment, more intensive care unit staffing or other beneficial resources.

"Undergoing surgery at teaching hospitals is perhaps more safe than at non-teaching hospitals because of the increased volumes of complex cases seen at these centers," Upchurch says.

In addition to Upchurch, study authors include Justin Dimick, M.D., John Cowan, M.D., and Lisa Colletti, M.D., all of the UMHS Department of Surgery.

###

Contact:
Nicole Fawcett, nfawcett@umich.edu, or
Kara Gavin, kegavin@umich.edu
734-764-2220


Disclaimer: AAAS and EurekAlert! are not responsible for the accuracy of news releases posted to EurekAlert! by contributing institutions or for the use of any information through the EurekAlert system.