Abstract
BACKGROUND: Due to the lack of detailed clinical information, existed evidence regarding a link between psychiatric factors and adverse cancer prognosis was inclusive. RESULTS: We identified 1,340 patients (48.8%) with perioperative psychiatric morbidity. Preoperative psychiatric morbidity was significantly associated with both general and surgical complications within 30 days (RR = 1.3, 95% confidence interval [CI] 1.1-1.5), and the risk of death within 90 days (RR = 1.6; 95% CI 1.1-2.2) after surgery. The hazards for mortality beyond 90 days was approximately 2-fold increased among patients with perioperative psychiatric morbidity (HR = 2.0, 95% CI 1.7-2.3 for overall mortality). MATERIALS AND METHODS: Based on the Swedish National Registry for Esophageal and Gastric cancer (NREV), we constructed a nationwide prospective cohort containing 2,745 surgically treated patients in 2006-2012. Perioperative psychiatric morbidity was defined as a clinical diagnosis of psychiatric disorder, from two years before to two years after surgery. Using propensity scores, we applied inverse probability of treatment weights (IPTW)-weighted Poisson regression model to evaluate relative risk (RR) of short-term surgical outcomes in relation to perioperative psychiatric morbidity. Further, IPTW-weighted Cox proportional hazards model was used to estimate hazard ratios (HRs) for mortality that occurred after 90 days of surgery. CONCLUSIONS: Perioperative psychiatric morbidity could worsen both short-term and long-term surgical outcomes among patients with gastric or esophageal cancer.