CHICAGO – Surgical patients receiving care at the least complex hospitals had a 50% increased risk of death compared with those treated at more complex hospitals, based on data from more than 400,000 patients.
Data from previous research suggest that the characteristics of individual hospitals predict surgical outcomes, but the relationship between hospital complexity and surgical mortality has not been well studied, Dr. Marta McCrum said at the annual clinical congress of the American College of Surgeons.
She and her colleagues reviewed Medicare administrative claims data from 2008-2009 and information on hospital characteristics from the American Hospital Association Survey 2009 on 2,695 hospitals. The hospitals were divided into quintiles based on complexity, which for purposes of this study was defined as the number of unique diagnoses and procedures performed at each facility.
"Our research suggests that outcomes for certain surgical procedures are better at more complex hospitals."
"Hospitals that see a wide variety of conditions (common problems but also very rare or complicated diagnoses) would be equipped with the wide range of services and resources needed to support them – that is, they would be complex. Similarly, hospitals that perform the widest range of unique procedures would also have the greatest diversity of services and technology," Dr. McCrum of Harvard University, Boston, explained in an interview.
"We therefore ranked the hospitals separately based on the number of unique diagnoses and procedures they saw, and then summed the ranks to assign a numeric value to the complexity of that hospital in comparison to the others. For the analysis, we separated the hospitals into quintiles based on this value."
Not surprisingly, low-complexity hospitals tended to be smaller, more rural, and located in lower-income areas, and the more complex hospitals tended to be larger, urban, and in higher-income areas, she said.
"Of note, the surgical mortality rate of the highest-complexity hospitals was 7.3%, versus 12.6% at the lowest-complexity hospitals," for an absolute risk reduction of 5.3%, she said.
The researchers controlled for hospital and population characteristics, including total number of hospital discharges, public/private ownership, percentage of Medicare patients, urban location, and county income. Hospital complexity remained a significant predictor of mortality between each quintile compared with the highest-complexity quintile.
"The average aggregate mortality rate at the lowest-complexity hospitals is 46% higher than that of the highest-complexity hospitals," said Dr. McCrum.
Overall, the research model explained 28% of the variability in mortality rates, and within the model, hospital complexity explained the greatest proportion of variability in mortality rates. Although hospital volume was a statistically significant predictor of mortality, the effect was small, she noted.
The study was limited by the fact that approximately two-thirds of the variability remained unexplained, likely due to a combination of patient factors and hospital factors, she said. Additional limitations included the lack of an existing metric to measure hospital complexity, and the limitations of using administrative claims data.
"Our research suggests that outcomes for certain surgical procedures are better at more complex hospitals," Dr. McCrum said in an interview. "This might be due in part to the expanded capabilities and systems of care present at these centers. By identifying these lifesaving elements that are cultivated in complex centers, and making them available in lower-complexity hospitals, we can ensure that all surgical procedures take place in facilities with the appropriate systems to support them," she said.
Dr. McCrum had no financial conflicts to disclose.