Factors influencing the volume-outcome relationship in gastrectomies: A population-based study

Background: A relationship between hospital procedural volume and patient outcomes has been observed in gastrectomies for primary gastric cancer, but modifiable factors influencing this relationship are not well elaborated. Methods: We performed a population-based study of 1864 patients undergoing gastrectomy for primary gastric cancers at 214 hospitals. Hospitals were stratified as high-, intermediate-, or low-volume centers. Multivariate models were constructed to evaluate the effect of institutional procedural volume and other hospital- and patient-specific factors on the risk of in-hospital mortality, adverse events, and failure to rescue, defined as mortality after an adverse event. Results: High-volume centers attained an in-hospital mortality rate of 1.0% and failure-to-rescue rate of .7%, both less than one-fifth of that seen at intermediate- and low-volume centers, although adverse event rates were similar across the three volume tiers. In multivariate modeling, treatment at a high-volume hospital decreased the odds of mortality (odds ratio [OR], .22; 95% confidence interval [95% CI], .05-.89), whereas treatment at an institution with a high ratio of licensed vocational nurses per bed increased the odds of mortality (OR, 1.96; 95% CI, 1.04-3.75). Being treated at a hospital with a greater than median number of critical care beds decreased odds of mortality (OR, .46; 95% CI, .25-.81) and failure to rescue (OR, .53; 95% CI, .29-.97). Conclusions: Undergoing gastrectomy at a high-volume center is associated with lower in-hospital mortality. However, improving the rates of mortality after adverse events and reevaluating nurse staffing ratios may provide avenues by which lower-volume centers can improve mortality rates.