Analysis of 3,294 cases of upper gastrointestinal bleeding in military medical facilities

Upper gastrointestinal bleeding (UGIB) remains a commonly encountered medical emergency with significant morbidity and mortality. Most large studies detailing the specific incidence, demographic, and mortality data were performed more than a decade ago. This study analyzes 3,294 cases of UGIB from 139 military medical treatment facilities over a 12-month period.

A retrospective chart review of Department of Defense military medical treatment facilities for UGIB was performed from October 1990 through September 1991. Core data such as demographic information were analyzed, as well as specific data relating to UGIB.

The incidence of UGIB was 36 per 100,000 population with a male-to-female ratio of 2.18 and a mean age of 52 +/- 19.65 yr. The number of cases increased with age; 44.5% of all patients were > or = 60 yr old. The overall mortality was 7.0% (231 of 3294), and death rates were similar among males and females (7.1% vs. 6.8%) with an odds ratio of 1.03 (CI: 0.77-1.402). Mortality increased with age in both genders; 73.2% of deaths occurred in patients more than 60 yr old. Comorbid illness was noted in 50.9% (1675 of 3294) of patients, with similar occurrence in males (48.7%) and females (55.4%). One or more comorbid illnesses were noted in 98.3% of the patients who died, and in 72.3% of cases, they were the primary cause of death. Bleeding was the primary cause of death in 18.6% of patients. Upper endoscopy was performed in 68.8% of cases, therapeutic endoscopy in 12.6%, repeat endoscopy in 10.7%, and surgery in 4.4%. Blood transfusions were administered in 47.3% of cases, with most patients receiving < 5 units of blood. Rebleeding after initial hemostasis was noted in 7.1% of cases. Factors related to increased mortality include age > 60 (p < 0.001), transfusion requirement > 5 U (p < 0.001), presence of comorbid illness (p < 0.001), rebleeding after initial hemostasis (p < 0.005), surgery (p < 0.001), and UGIB occurring during hospitalization (p = 0.027).

We conclude: 1) The incidence of UGIB is 2-fold greater in males than in females, in all age groups; however, the death rate is similar in both sexes. 2) The mortality rate in this study is slightly lower than in most previous studies and may be more reflective of the average mortality in the community at large. 3) In UGIB patients, comorbid illness and not actual bleeding is the major cause of death. 4) Upper endoscopy was performed less often in this study than in other studies, and there were fewer blood transfusions; however, rebleeding and mortality rates remained similar.