Characterizing geographic variation in postoperative venous thromboembolism

Abstract

Objective: Venous thromboembolism (VTE) after major surgery remains an important contributor to morbidity and mortality. Despite significant quality improvement efforts in prevention and prophylaxis strategies, the degree of hospital and regional variation in the United States remains unknown.

Methods: Medicare beneficiaries undergoing 13 different major surgeries at U.S. hospitals between 2016 and 2018 were included in this retrospective cohort study. We calculated the rates of 90-day VTE. We adjusted for a variety of patient and hospital covariates and used a multilevel logistic regression model to calculate the rates of VTE and coefficients of variation across hospitals and hospital referral regions (HRRs).

Results: A total of 4,115,837 patients from 4116 hospitals were included, of whom 116,450 (2.8%) experienced VTE within 90 days. The 90-day VTE rates varied substantially by procedure, from 2.5% for abdominal aortic aneurysm repair to 8.4% for pancreatectomy. Across the hospitals, there was a 6.6-fold variation in index hospitalization VTE and a 5.3-fold variation in the rate of postdischarge VTE. Across the HRRs, there was a 2.6-fold variation in 90-day VTE, with a 12.1-fold variation in the coefficient of variation. A subset of HRRs was identified with both higher VTE rates and higher variance across hospitals.

Conclusions: Substantial variation exists in the rate of postoperative VTE across U.S. hospitals. Characterizing HRRs with high overall rates of VTE and those with significant variation across the hospitals will allow for targeted quality improvement efforts.

Keywords: Deep vein thrombosis; Pulmonary embolism; Variation; Venous thromboembolism.