In-hospital mortality and associated factors among patients with cirrhosis in Martinique and Kinshasa: a retrospective comparative cohort study

Abstract Background and objectives: We compared in-hospital mortality and associated factors among patients with cirrhosis at Martinique University Hospital (CHUM), France, and the University Clinics of Kinshasa (CUK), Democratic Republic of the Congo. Methods: This retrospective cohort included adults hospitalised during 2019–2024 at CHUM and 2019–2025 at CUK. One selected admission per patient was analysed. Mortality proportions, Kaplan–Meier estimates, cumulative incidence with discharge alive as a competing event, and exploratory centre-specific Cox models were examined. Results: Among 404 patients, 302 were treated at CHUM and 102 at CUK. Patients at CHUM were older and had more comorbidities. Alcohol-related cirrhosis predominated at CHUM; viral hepatitis and undetermined aetiologies were more frequent at CUK. Mortality was 42.1% versus 42.2% (absolute difference 0.1 percentage points, 95% confidence interval [CI] −10.6 to 11.2; p = 0.985). Day-28 cumulative mortality was 36.1% versus 37.3%. At CHUM, chronic kidney disease (adjusted hazard ratio [aHR] 2.06, 95% CI 1.34–3.15), pneumonia (aHR 2.04, 95% CI 1.21–3.43) and Child–Pugh C versus A (aHR 3.44, 95% CI 1.14–10.38) were associated with death. At CUK, hepatic encephalopathy retained an adjusted association (aHR 3.81, 95% CI 1.42–10.20). Conclusions: Mortality was high in both centres, without a statistically significant difference in proportions. The findings support early evaluation of hepatic severity, renal function, infection and neurological status.