In order to improve patients’ health outcomes, it is important to know the available evidence regarding centralization of surgical interventions for digestive cancer in hospitals with the highest volume of cases. We aim to describe and identify the number of annual interventions recommended by hospitals in order to maximize the health outcomes and efficiency for patients undergoing digestive cancer surgery during 2013–2016 in centers belonging to the Spanish National Health System (SNS).

The study design was a retrospective cohort study (patients aged ≥18 years). Data from Spanish public hospitals’ basic minimum set of data at hospital discharge for esophagus, stomach, liver, pancreas and rectum cancers was used. Age, sex primary/secondary diagnosis and procedures (Charlson index) were included. Reinterventions, hospital stay and in-hospital mortality were considered as the outcomes and measures of efficiency. Hospitals were grouped as low-/medium-/high-volume according to the number of annual procedures. Descriptive analysis and logistic and Poisson regression models with Stata16 were undertaken.

High-volume hospitals performed between 67.4 (rectum) and 88.6 (liver) percent of interventions. The percentage of in-hospital mortality for all cancers was lower in high-volume centers (9.6% esophagus, 6.6% stomach, 7.1% pancreas, 4.2% liver and 2.2% rectum), showing a negative association between center volume and in-hospital mortality, which was statistically significant for esophagus (odds ratio [OR] = 0.48; 95% confidence interval [CI]: 0.28–0.81), stomach (OR = 0.51; 95% CI: 0.39–0.68) and rectum (OR = 0.63; 95% CI: 0.48–0.83) cancers. A non-statistically significant lower in hospital stay was observed in high-volume hospitals.

These results indicate that in Spain there is a negative association between the number of digestive oncological interventions per hospital and in-hospital mortality. This could help to define a threshold or cut-off point for the concentration of digestive cancer surgery in the SNS that might result in an improvement of lower in-hospital mortality and/or hospital stay.