The study, “Evaluation of Surveillance Bias and the Validity of the Venous Thromboembolism Quality Measure,” found the validity of the quality measure questionable because it scores hospitals based on the rates of patients with postoperative venous thromboembolism (VTE or intravenous blood clot) they report, regardless of how proactive their clinicians are in VTE testing and prevention. The authors argue that the measure doesn’t accurately reflect the quality of care provided because it contains a surveillance bias where hospitals that are more vigilant in performing VTE testing will likely find more instances of postoperative VTE, resulting in what appears to be a higher rate of blood clots and a worse hospital ranking.
“It is very possible that patients are being misled by this measure when they look at publicly available rankings,” said Karl Y. Bilimoria, MD, MS, surgical oncologist at Northwestern Memorial Hospital, director of the Surgical Outcomes and Quality Improvement Center at Northwestern University Feinberg School of Medicine and lead author for the study. “Our results indicate that patients who review hospital rankings or quality reports that use this VTE rate measure may actually be guided away from hospitals with higher levels of safety and quality care, toward lower quality hospitals.”
To assess the impact of the surveillance bias on the validity of using reported VTE rates as a quality measure, researchers compiled performance results from approximately 2,800 hospitals using 2010 Hospital Compare, American Hospital Association, and Medicare claims data for 954,926 surgical patient discharges following major surgery. VTE imaging and VTE event rates were calculated for each hospital. A summary score of characteristics reflecting a hospital that provides a higher quality of care – such as a hospital’s number of accreditations, inpatient beds and quality initiatives – was then applied to each hospital in the study.
Northwestern researchers found that hospitals with high rates for implementing VTE preventions also surprisingly had high VTE event rates. The study also revealed that hospitals with more characteristics reflecting a higher quality of care unexpectedly had higher VTE event rates. “Most importantly,” the authors write, “hospital VTE rates were associated with the intensity of detecting VTE with imaging studies.” Hospitals that performed fewer tests to look for blood clots (those in the lowest imaging rate quartile) diagnosed 5 VTEs per 1,000 patients, but hospitals in the highest imaging rate quartile found more than double that amount with 13.5 VTEs per 1,000 patients.
“The more you look for something the more you will find. Therefore, under the appropriate clinical circumstances, evaluating patients for VTE is indicative of high quality care,” said Gary A. Noskin, MD, senior vice president and chief medical officer at Northwestern Memorial and professor of medicine at the Feinberg School. “Hospitals like Northwestern Memorial that do an excellent job of identifying VTE to ensure patients receive the best treatment should be recognized for thorough care.”
“Although our study suggests that the blood clots after surgery measure does not measure VTE rates accurately, we need to continue measuring the hospitals’ quality of care for the prevention of blood clots after surgery,” said Bilimoria. “Our team is currently developing some novel approaches to solve this problem.”
“Evaluation of Surveillance Bias and the Validity of the Venous Thromboembolism Quality Measure,” is currently available on JAMA’s website and will appear in the October 9, 2013 print edition of JAMA.
Senior Media Relations Associate