Abstract
Background: We aimed to investigate the hospital volume-outcome relationship (HVOR) in patients with pleural infection, which is important because outcomes may be improved by volume-based selective referral if an inverse HVOR is present. Methods: We analyzed 24,876 patients with pleural infection in 2,188 hospital-years from Taiwan's National Health Insurance Research Database between 1997-2008. Primary outcome was hospital mortality. Secondary outcomes were hospital length of stay and charges. Hospital volume was measured both as a categorical and a continuous variable (per one case increase per hospital-year); and the effect of which was assessed using multivariable logistic regression models with generalized estimating equations accounting for hospital clustering effect. Adjusted covariates included patient and hospital characteristics (model 1), pleural surgery (model 2) and life-supporting measures (model 3). Results: HOVR was significant only when volume was measured as a categorical variable. Patients treated in the highest volume quartile (≥14 cases per hospital-year) had a 27% lower risk of hospital mortality than those in the lowest volume quartile (1 case per hospital-year) after adjusting for patient and hospital characteristics (model 1: adjusted OR 0.73, 95% CI 0.55-0.96). However, after adjusting for treatment covariates (model 2 and 3), the volume effect on hospital mortality disappeared. Hospital volume only explained a small proportion of variation in hospital mortality (-2 log likelihood%=0.26%). Conclusions: In patients with pleural infection, the effect of hospital volume on patient outcomes is small, depends on volume measures and can be explained by differences in treatment across hospitals.
- © 2012 ERS
