MSS13: LOCALIZED PANCREATIC CANCER AMONG MILITARY BENEFICIARIES IN THE NATIONAL CANCER DATABASE
R. Connor Chick, MD1; Patrick W Underwood, MD1; Robert W Krell, MD2; Joal D Beane, MD1; Timothy M Pawlik, MD, PhD, MPH, MTS, MBA1; 1Division of Surgical Oncology, Department of Surgery, The Ohio State University Wexner Medical Center and James Comprehensive Cancer Center; 2Department of Surgery, Uniformed Services University of the Health Sciences
Objectives: Pancreatic ductal adenocarcinoma (PDAC) outcomes are influenced by sociodemographic factors including health insurance. Military beneficiaries (MB), which include servicemembers, veterans, and their families, are sometimes treated in civilian settings with Tricare or Veterans Affairs (VA) acting as an insurance payor. To our knowledge, no prior study has used the National Cancer Database (NCDB) to evaluate outcomes for MB treated outside VA hospitals. The aim of this study is to examine outcomes among MB with localized pancreatic cancer.
Methods: Patients age <65 with stage I-III PDAC from 2010 to 2020 with private insurance (PI) or Other Government insurance, i.e. MB or Indian Health Service, with no prior malignancies and Charlson-Deyo score <3 were included. Descriptive analyses were performed using chi-square and Kruskal-Wallis tests. Overall survival was estimated using Kaplan-Meier method and analyzed using a multivariate Cox proportional hazards model with stratification by stage. Additional logistic regression models were generated for postoperative mortality. Propensity score matching (PSM) was performed based on results from descriptive analyses and Cox proportional hazards model.
Results: A total of 21,691 patients met inclusion criteria, among whom 781 had Other Government insurance. MB were more likely than PI to be male, Black, and to live in low income, low education, and rural areas (all p<0.05); in addition, distance traveled for care was longer (22 vs. 15 miles, p<0.001) and treatment was started later (26 vs. 24 days from diagnosis, p<0.001). Median OS was 18.5 months for PI vs. 14.7 months for MB (p<0.001) (Figure 1). On multivariate analysis, MB was an independent predictor of OS (hazard ratio (HR) 1.17, p=0.001), as well as Charlson-Deyo ≥1. MB were also less likely to undergo surgical resection (34.0% vs. 43.4%, adjusted odds ratio (AOR) 0.63). Among MB who did have surgery, postoperative mortality was higher at 30- (5.1% vs. 1.2%, AOR 3.84) and 90-days (8.5% vs. 2.5%, AOR 3.65). After PSM, insurance payor (adjusted HR 1.16, p=0.046), age, geographic location, and surgical resection remained independent predictors of OS.
Conclusions: Military beneficiaries had worse risk-adjusted OS compared with individuals with PI. MB were less likely to undergo pancreatectomy for stage I-III PDAC, however, among MB who did undergo resection, the risk of postoperative mortality was higher. These data suggest MB with PDAC face additional challenges compared to the civilian population. Further studies and policies are needed to close this gap in outcomes for MB with PDAC.
.png)