Consequences of Refusing Surgery for Esophageal Cancer: A National Cancer Database Analysis.

Abstract

Given the potential morbidity of esophagectomy, patients may pursue other treatments. We sought to determine predictors and outcomes of esophageal cancer patients who refused esophagectomy. The National Cancer Database (2004 to 2014) was queried for locally advanced esophageal cancer patients. A unique field allows identification of patients recommended to have surgery but who refused. Comparisons between the entire cohort and between propensity matched groups were performed using analysis of variance and χ² tests. Survival was compared using Kaplan-Meier curves. Logistic regression was performed to identify predictors of refusing surgery. We identified 18,459 patients with esophageal cancer meeting criteria, including 708 (3.8) who were recommended but refused surgery. By multivariate analysis, elderly, female, nonwhite race, squamous histology, early year of diagnosis, absence of insurance, treatment at nonacademic centers, lower income, and clinical stage I/II predicted refusal of surgery. Median survival was worse for patients who refused surgery compared with patients undergoing surgery. Among propensity matched groups (n = 525 each), median survival was better for patients undergoing surgery versus patients who refused (32 versus 21 months, p < 0.001). Although patients may be reluctant to undergo esophagectomy for esophageal cancer, refusal of surgery when offered comes at the expense of decreased survival. These data allow for a discussion of alternative outcomes with those patients in the context of shared decision making.