In-hospital mortality after stomach cancer surgery in Spain and relationship with hospital volume of interventions

BACKGROUND There is no consensus about the possible relation between in-hospital mortality in surgery for gastric cancer and the hospital annual volume of interventions. The objectives were to identify factors associated to greater in-hospital mortality for surgery in gastric cancer and to analyze the possible independent relation between hospital annual volume and in-hospital mortality. METHODS We performed a retrospective cohort study of all patients discharged after surgery for stomach cancer during 2001-2002 in four regions of Spain using the Minimum Basic Data Set for Hospital Discharges. The overall and specific in-hospital mortality rates were estimated according to patient and hospital characteristics. We adjusted a logistic regression model in order to calculate the in-hospital mortality according to hospital volume. RESULTS There were 3241 discharges in 144 hospitals. In-hospital mortality was 10.3% (95% CI 9.3-11.4). A statistically significant relation was observed among age, type of admission, volume, and mortality, as well as diverse secondary diagnoses or the type of intervention. Hospital annual volume was associated to Charlson score, type of admission, region, length of stay and number of secondary diagnoses registered at discharge. In the adjusted model, increased age and urgent admission were associated to increased in-hospital mortality. Likewise, partial gastrectomy (Billroth I and II) and simple excision of lymphatic structure were associated with a lower probability of in-hospital mortality. No independent association was found between hospital volume and in-hospital mortality CONCLUSION Despite the limitations of our study, our results corroborate the existence of patient, clinical, and intervention factors associated to greater hospital mortality, although we found no clear association between the volume of cases treated at a centre and hospital mortality.