Long-term mortality and causes of death in endoscopically verified upper gastrointestinal bleeding: comparison of bleeding patients and population controls

Abstract

Objective: Upper gastrointestinal bleeding (UGIB) is a common emergency, with in-hospital mortality between 3 and 14%. However, the long-term mortality and causes of death are unknown. We investigated the long-term mortality and causes of death in UGIB patients in a retrospective single-centre case-control study design.

Methods: A total of 569 consecutive patients, aged ≥18 years, admitted to Kuopio University Hospital for their first endoscopically verified UGIB during the years 2009–2011 were identified from hospital records. For each UGIB patient, an age, sex and hospital district matched control patient was identified from the Statistics Finland database. Data on endoscopy procedures, laboratory values, comorbidities and medication were obtained from patient records. Data on deaths and causes of death were obtained from Statistics Finland.

Results: In-hospital mortality of UGIB patients was low at 3.3%. The long-term (mean follow-up 32 months) mortality of UGIB patients was significantly higher than controls (34.1 versus 12.1%, p < .001). During the 6 months following UGIB, the risk of death compared to controls was highest (HR 19.2, 95% CI 7.0–52.4, p < .001) and remained higher up to 3 years after the bleeding. Beyond 3 years’ follow-up, there was no difference in mortality between the groups (HR 0.7, 95% CI 0.4–1.6, p = .436). During the first 3 months after the UGIB episode, mortality was related to gastrointestinal diseases; after 3 months, the causes of death were related to comorbidities and did not differ from causes of death in controls.

Conclusions: UGIB patients have three times higher long-term mortality than population controls.

Keywords:

• Upper gastrointestinal bleeding
• long-term mortality
• cause of death
• bleeding complications
• comorbidity
• survival

Acknowledgements

The authors wish to thank Marja-Liisa Sutinen, RN, and medical students Olli Pöntinen, Laura Ryhänen, Jukka Voutilainen, Sami Heikkinen, Jenna Ilmavirta, Anna Salminen, Johanna Weitz and Jenni Sipola for their efforts in collecting data from hospital electronic databases. We also want to thank research director Hannes Enlund, PhD, and statistician Vesa Kiviniemi, Lic Phil, from the Finnish Medical Agency for their help in planning the study concept and data collection.

Disclosure statement

No potential conflict of interest was reported by the authors.