Effect of Hospital Volume on Prosthesis Use and Mortality in Aortic Valve Operations in the Elderly.

Abstract

This study was designed to examine the effect of hospital procedural volume on outcomes in aortic valve replacement (AVR) in the elderly. The study included 277,928 Medicare beneficiaries who underwent AVR from 2000 through 2009 at one of 1,255 participating hospitals. Operative mortality and the use of mechanical prostheses were analyzed according to hospital annual procedural volume. Annual AVR volume was divided into 5 different categories: the smallest volume group with less than 10 AVRs per year to the largest group averaging more than 70 AVRs per year. The overall observed operative mortality rate was 7.3%; for isolated AVR it was 5.5%. Lower-volume hospitals exhibited increased adjusted operative mortality: 10 cases or fewer per year--odds ratio (OR), 1.55; 95% confidence interval (CI), 1.39 to 1.72; 11 to 20 cases per year--OR, 1.35; 95% CI, 1.23 to 1.47; 21 to 40 cases per year--OR, 1.15; 95% CI, 1.06 to 1.25; 41 to 70 cases per year--OR, 1.10; 95% CI, 1.01 to 1.20 relative to those hospitals performing more than 70 cases per year. The discrepancy in operative mortality between low- and high-volume hospitals diverged during the study. Mechanical valve use decreased with increasing hospital volume (p = 0.0001). Mechanical valves were used in 64.5% of AVRs in hospitals with an annual AVR volume less than 10 in contrast to only 25.4% in hospitals with an annual AVR volume more than 70. After adjustment, the use of mechanical valves was independently associated with increased operative mortality (OR, 1.15; 95% CI, 1.11-1.19). Low-volume centers were characterized by increased adjusted operative mortality and greater use of mechanical prostheses, a trend that persisted during the 10-year course of the study. These data would support the center-of-excellence concept for AVR and may be particularly relevant in the elderly population.