Research Activities, March 2014
Composite measure better for explaining hospital-level variation in bariatric surgery than procedure volume or complications
Patient Safety and Quality of Care
Using a statewide registry of patients who have undergone bariatric (weight loss) surgery, Justin B. Dimick, M.D., M.P.H., of the University of Michigan, Ann Arbor, and his colleagues found that a composite measure for laparoscopic gastric bypass surgery (LGBS) explained 89 percent of the variation in performance between hospitals. This contrasted with only 28 percent variation when hospitals were ranked solely on complication rates adjusted for patient risk factors.
Procedure-specific measures have become quite important in ranking hospital quality and safety, both as information for prospective patients and their families, and for payers (specifically, Medicare and Medicaid) looking to improve a hospital's quality and safety to reduce the costs of care. Previous studies have shown that bariatric surgery centers that received accreditation as centers of excellence, based on performing more than 125 bariatric procedures annually, did not have better outcomes than other hospitals. This led Dr. Dimick and his colleagues to look for better indicators of quality bariatric care.
Direct outcome measures, such as rates of serious problems and reoperation have been suggested, but vary too much to serve as accurate performance measures. The Michigan researchers chose to test composite measures, which combine a number of different quality indicators into a single score, using specific weights for each indicator. When the researchers used their composite measure to rank hospitals doing LGBS as 3-star (top 230 percent), 2-star (middle 60 percent), or 1-star (lowest 20 percent), they found that 1-star hospitals had a twofold difference in the serious complication rate compared with 3-star hospitals. In contrast, the differences in serious complication rates were much smaller if hospitals were ranked using either serious complications (1.6-fold difference) or hospital volume (0.85-fold). The findings were based on data from the Michigan Bariatric Surgery Collaborative for the years 2008 and 2009. The study was funded in part by AHRQ (HS17765).
More details are in "Composite measures for profiling hospitals on bariatric surgery performance," by Dr. Dimick, Nancy J. Birkmeyer, Ph.D., Jonathan F. Finks, M.D., and others in the online October 16, 2013, JAMA Surgery.