Long-term mortality after gastrointestinal bleeding, a retrospective cohort study
Aipi Forsberg-Puckett, Isabel Drake, Gabriele Wurm Johansson, Sara RegnérAbstract
Introduction
Gastrointestinal (GI) bleeding is a common reason for emergency hospital admission, with an in-hospital mortality of 3.9–14%. Less is known about long-term mortality after a bleeding event and the factors influencing it.
Methods
A cohort of consecutive adult patients admitted for GI bleeding at Skåne University Hospital from January 1, 2018 to June 30, 2019 was validated using medical records. Mortality data were collected from the Swedish National Cause of Death Register for 3 years after the index bleeding event. Long-term mortality was estimated using Cox proportional hazards regression.
Results
A total of 582 patients were included, 58% male, with a median age of 76 years. During follow-up, 228 (39.2%) patients died, with a median time to death of 263 days. Patients who died of GI bleeding had a shorter median time to death (47 days, IQR 10.5–272, n = 31) compared to those dying from GI tumors (192 days, IQR 63–679, n = 29) or cardiovascular causes (260 days, IQR 94–639, n = 62). Antithrombotic treatment at admission did not significantly affect 3-year mortality (HR 1.09, 95% c.i. 0.09–1.01, P = 0.31), even after adjustment for Charlson Comorbidity Index and sex. No difference in 3-year mortality was found between upper and lower GI bleeding (HR 0.94, 95% c.i. 0.75–1.19, P = 0.61).
Discussion
Long-term mortality after GI bleeding is high. Patients dying from new bleeding events have shorter survival than those dying from tumors or cardiovascular causes. Further studies are needed to identify patients at risk and factors influencing rebleeding and mortality.