Medical expenses in treating acute esophageal variceal bleeding A 15-year nationwide population-based cohort study

Acute variceal bleeding in patients with cirrhosis is related to high mortality and medical expenses. The purpose of present studies was to analyze the medical expenses in treating acute esophageal variceal bleeding among patients with cirrhosis and potential influencing clinical factors. A total of 151,863 patients with cirrhosis with International Classification of Diseases-9 codes 456.0 and 456.20 were analyzed from the Taiwan National Health Insurance Research Database from January 1, 1996 to December 31, 2010. Time intervals were divided into three phases for analysis as T1 (1996-2000), T2 (2001-2005), and T3 (2006-2010). The endpoints were prevalence, length of hospital stay, medical expenses, and mortality rate. Our results showed that more patients were <65 years (75.6%) and of male sex (78.5%). Patients were mostly from teaching hospitals (90.8%) with high hospital volume (50.9%) and high doctor service load (51.1%). The prevalence of acute esophageal variceal bleeding and mean length of hospital stay decreased over the years (P < 0.001), but the overall medical expenses increased (P < 0.001). Multiple regression analysis showed that older age, female sex, Charlson comorbidity index (CCI) score >1, patients from teaching hospitals, and medium to high or very high patient numbers were independent factors for longer hospital stay and higher medical expenses. Aged patients, female sex, increased CCI score, and low doctor service volume were independent factors for both in-hospital and 5-year mortality. Patients from teaching hospitals and medium to high or very high service volume hospitals were independent factors for in-hospital mortality, but not 5-year mortality. Medical expenses in treating acute esophageal variceal bleeding increased despite the decreased prevalence rate and length of hospital stay in Taiwan. Aged patients, female sex, patients with increased CCI score from teaching hospitals, and medium to high or very high patient numbers were the independent factors for increased medical expenses.