"Not much attention has been paid to the management of complications," he said. "That is where we should be looking."
The pattern could be seen in management of specific complications. For example, all the hospitals had roughly similar incidences of post-surgery bleeding. But the death rate from that complication was 50 percent higher in some hospitals than in others.
Though the new study does not directly address ways to manage surgical complications, "we can speculate," Dimick said. "A number of things in previous studies have been associated with mortality."
Hospitals with lower surgical death rates tend to have fully-staffed, 24-hour intensive care units with physicians trained to handle post-surgical emergencies. They have high nurse-to-patient ratios in their intensive care units and their wards. "The more nurses the better because there are fewer patients per nurse," Dimick said.
A less tangible factor -- "just the culture of the hospital" -- also is involved, he said. "Do people feel afraid to call the surgeon at night, are they afraid to go up the chain of command, are they not afraid of pushing the button?"
Dr. Peter Pronovost, a professor of anesthesiology and critical care medicine at Johns Hopkins University School of Medicine, said that the study's findings don't lessen the need to prevent surgical complications and errors. Pronovost was an originator of the checklist concept.
"We have to try to prevent complications when we can," he said. "But when we can't, we must make sure that they are properly recognized and treated. This study shows that recognition and treatment has a substantial impact on a patient's risk of dying."
Studies already have shown that the availability of intensive care specialists and a high nurse-to-patient ratio improve
All rights reserved