Hospital variation in treatment for synchronous metastatic esophageal and gastric cancer: A nationwide population-based study in the Netherlands

Care for metastatic esophageal (EC) or gastric cancer (GC) includes a large variety of treatment modalities. Data on treatment variation across centers are unknown. This study investigated treatment variation across hospitals and its effect on overall survival (OS) in the Netherlands by conducting a nationwide retrospective cohort study with population-based data from the Netherlands Cancer Registry. Patients diagnosed with synchronous metastatic EC/GC between 2015 and 2022 were included. Multilevel logistic regression assessed treatment patterns according to hospital of diagnosis. OS was analyzed using Cox regression analysis after categorizing hospitals into tertiles based on their adjusted odds (low/medium/high) for systemic treatment (chemotherapy, targeted therapy, and immunotherapy). Among 8406 EC and 3871 GC patients, the proportion receiving systemic treatment varied substantially: 19.8%-69.6% for EC and 15.8%-81.3% for GC across hospitals. Hospital of diagnosis was significantly associated with the adjusted probability of receiving systemic treatment (p < .0001). Ten out of 78 EC (12.8%) and 7 out of 73 (9.6%) GC hospitals had significantly lower systemic treatment probabilities. EC patients with OS ≥4 months diagnosed at hospitals with lower probabilities had significantly worse OS compared to high-probability hospitals (hazard ratios [HR] 0.87 [0.79-0.95] p = .002). GC patients from low-probability hospitals had significantly worse OS than from medium- (HR 0.86 [0.76-0.96], p = .011) or high-probability hospitals (HR 0.73 [0.64-0.82], p < .0001). In conclusion, this study showed substantial hospital variation in treatment for metastatic EC and GC. Hospital of diagnosis was not only associated with the probability of receiving systemic treatment but also OS. This reflects the challenge of ensuring equal healthcare access.