Variation in the cost of 5 common operations in the United States

Abstract

Health care costs are an important policy focus in the United States. The magnitude and drivers of variation in the costs of common operative procedures are not well understood. We sought to characterize variation in costs across hospitals. We used data from the Nationwide Inpatient Sample from 2001-2011 for 5 elective operations: colectomy, coronary artery bypass graft, total knee arthroplasty, cesarean section, and lung resection. Hospitals were benchmarked for each using hierarchical risk- and reliability-adjustment methods to generate an observed-to-expected cost ratio, which was adjusted for patient demographics, comorbidity, wage index, and procedure complexity. Hospitals were divided into quintiles. Characteristics of high- and low-quintile hospitals and their adjusted outcomes were examined. Cost observed-to-expected ratios ranged widely for all 5 procedures: 14.9-fold for colectomy, 5.5-fold for coronary artery bypass graft, 12.5-fold for lung resection, 10.6-fold for total knee arthroplasty, and 28.0-fold for cesarean section. Comparing highest to lowest cost quintiles of hospitals, high-cost hospitals were more likely to serve minority and Medicaid patients. Mortality was elevated significantly in high-cost hospitals for colectomy, coronary artery bypass graft, and lung resection (adjusted odds ratio 1.99, 1.32, 2.57; respectively). Service lines were correlated at low-cost hospitals. There was a significant association between greater procedure volume and low-cost hospitals for colectomy, coronary artery bypass graft, and total knee arthroplasty. Despite robust adjustment, there is wide cost variation for common operative procedures in the United States. High-cost hospitals may need to focus on cost reduction at the hospital level to reduce cost across service lines. Benchmarking costs may identify significant opportunities to promote value, or the balance between cost and quality, in operative care in the United States.