PTU-011 Predictors of mortality and rebleeding outcomes after peptic ULCER bleeding

    loading  Checking for direct PDF access through Ovid



Peptic ulcer disease (PUD) accounts for 25%–56% of acute upper gastrointestinal bleeding (AUGIB) cases and is associated with high rates of mortality and rebleeding. We aimed to assess the rates and factors associated with 1 year mortality and rebleeding in patients with bleeding PUD.


This was a single-centre study of patients with AUGIB and endoscopic confirmation of PUD between November 2012–2014. All patients received at least 1 year of retrospective follow-up after endoscopy. Electronic records were scrutinised for outcomes of mortality and bleeding, with time-to-event analyses performed using a Kaplan-Meier plots and Cox-regression.


91 patients (median age 78.4, 65.9% male) were included. 63.7% were admitted with AUGIB and 36.3% bled during their inpatient stay. Mortality at 30 days and 1 year were 12.1% and 34.1% respectively, with 1 year mortality comprising: cardiovascular (20.0%), bleeding (20.0%), other (33.3%), unknown (26.7%). On univariate analysis, predictors of 1 year mortality included inpatient bleeding (hazard ratio [HR] 2.38, 95% CI: 1.18–4.83, p=0.016) [Figure 1], age (HR 1.036 per increase, 95% CI 1.009–1.065, p=0.009), Forrest classification (HR 2.22 for class 1 and 2 vs. 3, p=0.04), Rockall Score (HR 1.64 per increase, 95% CI 1.28–2.10, p<0.001), Charlson index (HR 1.34 per increase, p<0.001), aspirin use (HR 3.05, p=0.003), rebleeding (HR 5.52, p<0.001). The effect of inpatient bleeding on mortality was not significant (p=0.19) after adjusting for Charlson index. H. pylori was positive in 35.7%; eradication was associated with reduced mortality even after adjusting for Forrest classification and age (HR 0.30, p=0.007). Multivariable analyses to account for age are shown in table 1. The 1 year rebleeding rate was 7.8%. Higher haemoglobin on discharge (HR 0.940 per 10 g/dL increment, p=0.04), Forrest 3 ulcers (HR 0.18, p=0.02) and H. pylori eradication (HR 0.214, p=0.02) were significantly associated with reduced rates of rebleeding.


Increasing age, higher Rockall and Charlson scores, Forrest 1 or 2 lesions, inpatient bleed, and rebleeding were factors associated with mortality in bleeding PUD. Higher rates of inpatient mortality may be explained by age and co-morbidities. Eradication of H. pylori was associated with improved outcomes and should be considered in all cases of bleeding PUD.

Related Topics

    loading  Loading Related Articles