Volume-outcome relationship of gastrectomy for cancer in Texas

6091 Background: The association between hospital volume and outcomes has been well described in a variety of surgical resections in cancer patients, but not for gastrectomy. Our goal was to assess the mortality differences for gastrectomies between low and high volume hospitals. Methods: We conducted a population-based study of mortality following gastrectomy in Texas. A database was constructed of every cancer patient that underwent a gastrectomy for adenocarcinoma between 1/1/99 and 12/31/01 from the Texas Hospital Inpatient Discharge Public Use Data File. We evaluated mortality rates for high volume (HV, >15 gastrectomies/yr) versus low volume (LV, ≤15 gastrectomies/yr) hospitals and developed a multivariate model of risk of mortality using a random effects model to account for clustering of patients with hospitals. All data obtained were based on individual hospital reporting of mortality and all diagnoses were obtained from hospital billing codes. Results: There were 1864 gastrectomies performed for...