Association between hospital competition and quality of prostate cancer care

Abstract

Background: Hospitals account for approximately 6% of United States' gross domestic product. We examined the association between hospital competition and outcomes in elderly with localized prostate cancer (PCa). We also assessed if race moderated this association.

Methods: Retrospective study using Surveillance, Epidemiology, and End Results (SEER) - Medicare database. Cohort included fee-for-service, African American and white men aged ≥ 66, diagnosed with localized PCa between 1998 and 2011 and their claims between 1997 and 2016. We used Hirschman-Herfindahl index (HHI) to measure of hospital competition. Outcomes were emergency room (ER) visits, hospitalizations, Medicare expenditure and mortality assessed in acute survivorship phase (two years post-PCa diagnosis), and long-term mortality. We used Generalized Linear Models for analyzing expenditure, Poisson models for ER visits and hospitalizations, and Cox models for mortality. We used propensity score to minimize bias.

Results: Among 253,176 patients, percent change in incident rate of ER visit was 17% higher for one unit increase in HHI (IRR: 1.17, 95% CI: 1.15-1.19). Incident rate of ER was 24% higher for whites and 48% higher for African Americans. For one unit increase in HHI, hazard of short-term all-cause mortality was 7% higher for whites and 11% lower for African Americans. The hazard of long-term all-cause mortality was 10% higher for whites and 13% higher for African Americans.

Conclusions: Lower hospital competition was associated with impaired outcomes of localized PCa care. Magnitude of impairment was higher for African Americans, compared to whites. Future research will explore process through which competition affects outcomes and racial disparity.

Keywords: Hospital competition; Localized prostate cancer; Older adults; Outcomes of care; SEER-Medicare.

Fig. 1

Cohort selection for prostate cancer cases diagnosed between 1998 and 2011 The mean annual HHI score between 1998 and 2011 remained mostly stable (eFigure 1). In Table 1, we show the comparison of demographic and clinical characteristics between four quartiles of HHI score