Main Article Content

Factors influencing outcome in patients with perforated peptic ulcer disease at a South African tertiary hospital


J.J. Nanack
L. Ferndale

Abstract

Background: Perforated peptic ulcer (PPU) is associated with significant morbidity and mortality, particularly in low to middle income  countries. This study aimed to scrutinise the clinical course of patients diagnosed with PPU and identify modifiable factors to improve outcomes.


Methods: A retrospective review of the hybrid electronic medical record (HEMR) database at Grey’s Hospital was performed.  All patients diagnosed with PPU between January 2013 and December 2020 were entered into the study. The variables collected include  age, ethnicity, comorbid profile, Boey score, type of surgery performed and complications. These factors were analysed to determine the  factors responsible for morbidity and mortality.


Results: One hundred and ninety four patients were diagnosed with PPU during the  study period. Six patients were treated non-operatively, all of whom survived. In the surgically treated group, omental patch repair was  performed in 159 (84.5%) patients, and primary closure in 26 (13.8%) patients. The leak rate was 32% in the cohort that underwent  relaparotomy and the overall mortality was 14%. There was no significant relationship between the type of repair performed and  outcome. All patients had a Boey score of 1 or more. The following factors were found to increase the probability of in-hospital mortality:  age > 40 years (OR: 8.49, 95% CI 2.46–29.29 p < 0.01), female gender (OR: 2.509, CI 0.98–6.37, p = 0.048), need for relaparotomy (OR: 0.398,  CI 0.17–0.91, p = 0.027) and Boey score > 1 (OR: 46.437, CI 6.13–350.28, p < 0.01). A Boey score > 1 was the only variable that increased the  likelihood of finding a leaking repair at relaparotomy (p < 0.01).


Conclusion: The Boey score was a significant predictor of mortality and  leak rate in our patients with PPU. Adding age as a variable may improve the ability to predict mortality in our setting, while the impact of  gender and ethnicity needs further investigation.  


Journal Identifiers


eISSN: 2078-5151
print ISSN: 0038-2361