The Real Safety Net Hospitals: High-Medicaid Hospitals Confer Worse Survival for Patients Undergoing Pancreas and Biliary Cancer Surgery.

Abstract

Large-database cancer research often defines hospitals servicing a threshold number of Medicaid patients "safety net"; however, counts are generally with the cancer of interest, and because sufficient rare/complex cases such as hepatobiliary/pancreatic (HPB) cancers tend to be managed at larger institutions, "safety net" may be skewed toward academic centers. Recent data show similar survival for "safety net" HPB care. We examine patterns/outcomes of curative intent surgery (CIS) for HPB cancers by a more comprehensive Medicaid-serving (AID) status utilizing all alimentary tract cancers to generate AID quartiles. We identified all alimentary tract cancers (NCDB: 2004-2021) to determine quartile proportions of Medicaid-insured patients. Stage I-III primary HPB cancers were analyzed by highest quartile of AID facility ("high-AID") versus remaining quartiles ("low-AID"). Logistic regression identified predictors for undergoing CIS. Cox proportional regression modeled survival. P < 0.01 was significant. HPB cancer cases totaled 286,873. High-AID patients were more likely to be non-white and have low income/education. Patients were less likely to undergo CIS at high-AID programs (hepatic [H]: OR 0.83, 95% CI 0.80-0.86; pancreas [P]: OR 0.81, 95% CI 0.78-0.84; biliary [B]: OR 0.83, 95% CI 0.79-0.87). Multivariable analyses showed similar survival at 1 and 5 years for hepatic resection at high-AID facilities (1 year: HR 1.01 [0.89-1.14]; 5 years: HR 1.01 [0.95-1.07]). High-AID for pancreatic/biliary patients had greater mortality risk at 1 and 5 years. Including more hospitals to define AID status revealed worse survival at high-Medicaid-serving facilities for pancreatic/biliary, but not hepatic cancer surgery patients. Our "safety net hospital" proxy (high-AID) allows for better care disparity identification for complex underserved patients with cancer.