Procedure volume of gastric cancer resections versus 5-year survival

Aim: We used nationwide, population-based data to examine associations between hospital and surgeon volumes of gastric cancer resections and their patients' short-term and long-term survival likelihood. Methods: The study uses 1997-1999 inpatient claims data from Taiwan's National Health Insurance linked to cause of death data for 1997-2004. The total cohort of 6909 gastric cancer resection patients were categorized by their surgeon's/hospital's procedure volume, and examined for differences in 6-month mortality and 5-year mortality (post 6 months), by procedure volume, using Cox proportional hazard regressions, adjusting for surgeon, hospital and patient characteristics. We hypothesized that surgeons' case volume and age but not hospital volume will predict short-term and long-term survival. Results: Adjusted estimates show that increasing surgeon volume predicts better 6-month survival (adjusted mortality hazard ratio = 1.3 for low-volume surgeons relative to very high-volume surgeons; p < 0.01) and 5-year survival (adjusted mortality hazard ratios = 1.3; p < 0.001 for low-volume; 1.2 with p < 0.01 for medium volume) and increasing surgeon's age (adjusted hazards ratio = 1.4 for age < 41 - years relative to 41-50 years; p <= 0.001; 0.8 for >= 51 years relative to 41-50 years; p < 0.05). In hospital volume regressions, surgeon's age is a consistent and significant predictor, not hospital volume. Findings suggest a key role of experience in surgical skill and sensitivity for early stage diagnosis in gastric cancer survival. Conclusions: Although a key study limitation is the lack of cancer stage data, the pattern of findings suggests that experienced surgeons have relatively better survival outcomes among gastric cancer patients. (c) 2007 Elsevier Ltd. All rights reserved.