Effect of Comorbidity on Mortality in Patients with Peptic Ulcer Bleeding - A Systematic Review and Meta-Analysis

Purpose: To systematically review and meta-analyze data from the literature on the impact of comorbidity on short-term mortality in patients with peptic ulcer bleeding (PUB). Methods: Systematic searches of PubMed and Embase (1989 to January 2010) were conducted to identify randomized trials, case control and cohort studies that provided patient numbers indicating the short-term death rate in PUB patients with and without comorbidity. Mortality rates were pooled across studies using RevMan 5.0 software using a random effects model and reported as risk ratios (RRs) with 95% confidence intervals (CIs). Results: Of 1572 references that were screened, 10 articles were eligible, all of which were observational studies (eight European, three Asian, and one Middle Eastern). Overall, 42.4% (6236/14694) of patients with PUB had at least one comorbidity (range: 29.2-80.9%). Of the patients with PUB who had at least one comorbidity, 13.4% (833/6236) died during hospitalization or within 30 days of admission or treatment (range: 5.7-25.0%), compared with 2.6% (224/8458) of PUB patients without comorbidity (range: 0.0-13.5%). This equated to an RR of 4.04 (95% CI 2.51-6.49) of death in patients with PUB with a comorbidity compared with those without (Figure). The RR increased slightly to 5.44 (95% CI 3.89-7.61) when heterogeneity was reduced by excluding two studies that included a minority of patients with bleeding that was not related to peptic ulcer. Conclusion: Short-term risk of mortality appears to be higher in patients with PUB with comorbidity compared with those without. The high prevalence of comorbidity in patients with PUB may explain why mortality rates have remained relatively constant (5-10%) in this population, despite significant advances in treatment. Further analyses will aim to assess the relative contribution of individual comorbidities to mortality in patients with PUB. Disclosure: This abstract was developed with the aid of medical writing support sponsored by AstraZeneca.Figure: [106] Meta-analysis of risk of death in PUB patients with comorbidity compared with those without. M-H, Mantel-Haenszel.