Objective: To investigate the impact of center volume on postoperative mortality (POM) according to patient condition.
Background: Centralization has been shown to improve POM in esophageal and, to a lesser extent, gastric cancer surgery; however, the benefit of centralization for patients with low operative risk is questionable.
Methods: All consecutive patients who underwent esophageal or gastric cancer surgery between 2010 and 2012 in France were included (N = 11,196). The 30-day POM was compared in terms of the center volume (low: <20 cases per year, intermediate: 20-39, high: 40-59, and very high: ≥60) and stratified according to the Charlson score (0, 1-2, ≥3). The consistency across the esophageal (n = 3286) and gastric (n = 7910) subgroups, and variations between 30-day and 90-day POM were analyzed.
Results: Low-volume centers treated 64.2% of patients. A linear decrease in 30-day and 90-day POM was observed with increasing center volume, with rates of 5.7% and 10.2%, 4.3% and 7.9%, 3.3% and 6.7%, and 1.7% and 3.6% in low, intermediate, high, and very high-volume centers, respectively (P < 0.001). Comparing low and very high-volume centers, 30-day POM was 4.0% versus 1.1% for Charlson 0 (P = 0.001), 7.5% versus 3.4% for Charlson 1 to 2 (P < 0.001), and 14.7% versus 3.7% for Charlson ≥3 (P = 0.003) patients. A similar linear decrease was observed in the esophageal and gastric cancer subgroups. Between the low and very high-volume centers, an almost 70% reduction in the relative risk of POM was systematically observed, independent of Charlson score or tumor location.
Conclusions: To improve POM, esophageal and gastric cancer surgery should be centralized, irrespective of the patient's comorbidity or tumor location.