Gastrointestinal Bleeding in Hospitalized covid‐19 Patients: a propensity Score Matched Cohort Study

Accepted Article

---

This article is protected by copyright. All rights reserved

antiplatelet, anticoagulants and NSAIDS were not associated with the GI bleeding in patients with 

COVID-19. Nonetheless, the complication of GI bleeding was associated with an increased risk of 

mortality in patients who developed GI bleeding during the hospitalization. 

The reported prevalence of GI bleeding in critically ill patients is variable and has been reported to be 

anywhere from 1.5% to 5.5%[24].  The reported point prevalence of 3% in hospitalized COVID-19 

patients falls within this range and thus there is not expected increased point prevalence.   Perhaps the 

most surprising result was that GI prophylaxis with proton pump inhibitors or histamine receptor 

blockers did not have a protective effect against GI bleeds in COVID-19 patients.  A recent network 

meta-analysis of 57 studies showed a benefit of PPIs in protection against GI bleeding[25].  In our 

study, patients were considered have GI prophylaxis therapy if started within 24 hours of admission. 

Thus it is unclear why a protective effect is not seen in the hospitalized COVID-19 population.  

Perhaps the systemic inflammatory response is too great in this patient population to prevent stress 

ulcer prophylaxis.  It is also possible that a non-acid mediated cause of GI bleed, such as disseminated 

intravascular coagulation induced bleeding[6], is occurring in which antacid therapy will not prevent 

against a bleed. Given the vast majority in this series did not undergo endoscopy, this remains 

unknown how many patients had stress induced ulcers.  The management of GI bleeding in COVID-

19 only involves endoscopy after conservative measures fail[5,7], and thus a low rate of endoscopy is 

not surprising in this cohort. 

It was surprising that anticoagulation was not associated with an increased risk for GI bleeding in the 

hospitalized COVID-19 population. This is a known risk factor for GI bleeding in non-COVID-19 

populations[24].  This study is reassuring given anticoagulation is being prescribed to critically ill 

hospitalized COVID-19 patients. Further prospective research from the randomized trials described 

above could provide more insight regarding this. 

Our study has notable strengths.  The data is from a large health system in which a relatively large 

number of GI bleed events were matched to COVID-19 patients without a GI bleed. The propensity 

scoring system accounted for many variables include demographics, age, known GI bleeding risk