Association between physician volume and in-hospital mortality in patients with systemic lupus erythematosus

OBJECTIVE: Systemic lupus erythematosus (SLE) is an uncommon, clinically complex disease for which prior experience treating similar patients may be particularly important. This study was undertaken to determine if physician volume is associated with the outcome of hospitalization of patients with SLE. METHODS: Data on in-hospital mortality in a population-based sample of 15,509 patients with SLE ages 18 years or older who were hospitalized in 2000, 2001, or 2002 in New York or Pennsylvania were obtained from state health planning agencies. Risks of in-hospital mortality were examined in relation to the average annual number of patients with SLE hospitalized by the admitting physician. RESULTS: Physician volume was inversely associated with mortality. Mortality was 4.1% among patients of physicians who treated <1 hospitalized patient with SLE per year, 3.5% among patients of physicians who treated 1-3 patients per year, and 2.5% among patients of physicians who treated >3 patients per year. After adjustment for demographic characteristics, severity of illness, and hospital characteristics, the mortality risk was 20% lower among patients in the middle category of physician volume (odds ratio 0.80, 95% confidence interval 0.66-0.96, P =0.02), and 42% lower among patients in the highest category of physician volume (odds ratio 0.58, 95% confidence interval 0.42-0.82, P = 0.002), compared with patients in the lowest category. The association was stronger among patients with nephritis (n = 2,673), for whom the adjusted odds of mortality were approximately 60% lower among those in the highest category of physician volume. CONCLUSION: Our findings indicate that higher disease-specific physician volume is associated with lower risks of in-hospital mortality in patients with SLE.