Failure to rescue as a source of variation in hospital mortality for ovarian cancer

Abstract

Purpose: Although the association between high surgical volume and improved outcomes from procedures is well described, the mechanisms that underlie this association are uncertain. There is growing recognition that high-volume hospitals may not necessarily have lower complication rates but rather may be better at rescuing patients with complications. We examined the role of complications, failure to rescue from complications, and mortality based on hospital volume for ovarian cancer.

Patients and methods: The Nationwide Inpatient Sample was used to identify women who underwent surgery for ovarian cancer from 1988 to 2009. Hospitals were ranked on the basis of their procedure volume. We determined the risk-adjusted mortality, major complication rate, and "failure to rescue" rate (mortality in patients with a major complication) for each tertile. Univariate and multivariate associations were then compared.

Results: We identified 36,624 patients. The mortality rate for the cohort was 1.6%. The major complication rate was 20.4% at low-volume, 23.4% at intermediate-volume, and 24.6% at high-volume hospitals (P < .001). However, the rate of failure to rescue (death after a complication) was markedly higher at low-volume (8.0%) compared with high-volume hospitals (4.9%; P < .001). After accounting for patient and hospital characteristics, women treated at low-volume hospitals who experienced a complication were 48% more likely (odds ratio [OR], 1.48; 95% CI, 1.11 to 1.99) to die than patients with a complication at a high-volume hospital.

Conclusion: Mortality is lower for patients with ovarian cancer treated at high-volume hospitals. The reduction in mortality does not appear to be the result of lower complications rates but rather a result of the ability of high-volume hospitals to rescue patients with complications.