GI Bleeds Tied to Worse Prognosis for COVID-19 Inpatients

Gastrointestinal (GI) bleeding was observed in 3% of hospitalized COVID-19 patients, and a bleed developing during hospitalization was tied to greater mortality, a large New York cohort study found.

Anticoagulation or antiplatelet agents were not risk factors for GI bleeding but neither did they protect against them, reported Arvind J. Trindade, MD, of the Feinstein Institutes for Medical Research at Northwell Health in Manhasset, New York, and colleagues.

The nested propensity score-matched case-control study, published online in the Journal of Internal Medicine, involved patients treated at a large health system in the metropolitan New York area (including Manhattan, Long Island, Queens, and Staten Island) from March 1 to April 27, 2020.

Of 11,158 polymerase chain reaction-positive COVID-19 inpatients, 314 were identified as having GI bleeding, for a rate of 3%. A GI bleed during hospitalization was associated with an increased mortality risk (odds ratio 1.58, P=0.02).

“There were no identifiable risk factors for GI bleeding,” the researchers wrote. “Use of anticoagulation medication or antiplatelet agents were not associated with increased risk of GI bleeding in COVID-19 patients.”

Study Details

Participants were stratified into two groups:

• Those with GI bleeding on admission within 6 hours of hospitalization (104 patients)

• Those with GI bleeding during hospitalization (210 patients)

Those with GI bleeding on admission were matched 1:1 to patients with no GI bleed on admission. The propensity score model included demographics (age, sex, race, ethnicity) and general comorbidities (by Charlson Comorbidity Index), as well as GI-specific comorbidities and known GI bleed risk factors (end-stage renal or peptic ulcer disease, diverticulitis, inflammatory bowel disease, and a previous GI bleed).

Patients with GI bleeding during hospitalization were matched according to similar parameters.

The mean age of the COVID-19 patients was about 70, with approximately equal numbers of men and women. In 68.15% of patients, bleeding was located in the upper GI tract, and the majority of patients given anticoagulation were on prophylactic rather than therapeutic doses.

The mean hospital stay for on-admission bleeders was 8.41 days vs 15.41 for inpatient bleeders, and 45.71% of inpatient bleeders required intensive care.

There was no difference in mortality (OR 0.62, 95% CI 0.31-1.24, P=0.17) or need for mechanical ventilation (OR 0.50, 95% CI 0.21-1.17, P=0.10) in patients admitted with a GI bleed.

Increased mortality was observed, however, in patients who developed a bleed during hospitalization (OR 1.58, 95% CI 1.06-2.34, P=0.02), and there was no statistical difference in the need for mechanical ventilation (OR 1.34, 95% CI 0.88-2.04, P=0.17).

The researchers noted that GI bleeding has previously been observed in critically ill inpatients at comparable prevalence rates, ranging from 1.5% to 5.5%. A recent study reported a rate of severe GI bleeding in 1.1% of COVID-19 patients receiving intensive care.

While anticoagulation therapy for virus-related coagulopathy has been shown to decrease mortality, it is unclear whether this therapy contributes to increased GI bleeding, and the risks are not well understood, Trindade told MedPage Today, adding that trials are underway to shed light on this.

“COVID-19 causes arterial and venous thrombosis,” he said. “As a result, many patients with COVID-19 receive anticoagulation medication that is already known to predispose to GI bleeding, but it is unknown if putting these patients on anticoagulation increases the risk for GI bleeding.”

The most surprising result, he said, was that GI prophylaxis with proton pump inhibitors (PPIs) or histamine receptor blockers did not have a protective effect against GI bleeds in COVID-19 patients, while a recent network meta-analysis of 57 studies showed a benefit of PPIs in protection against stress ulcer bleeding.

“We would have thought PPI medications would be protective against a GI bleed, but this was not the case,” Trindade said. “So the mechanism of SARS-CoV-2–induced GI bleeding is likely not acid-mediated.”

He said that while it is unclear why no protective effect was seen in the New York COVID-19 cohort, he and his colleagues surmised that the systemic inflammatory response may be 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, is occurring in which antacid therapy will not prevent against a bleed,” the team wrote. But since the vast majority of patients in this series did not undergo endoscopy, which is generally done only after other interventions have failed, the number of patients with stress-induced ulcers could not be determined.

Study limitations, the researchers noted, included that it was retrospective, and that it was not possible to accurately differentiate between therapeutic dose and prophylactic dose anticoagulation due to dose changes, kidney function, and switching between prophylactic and therapeutic doses in the cohort. In addition, risk factors for patients admitted with a GI bleed were not analyzed since current home medications and adherence could not be verified.