Critical Care and Shock

Corticosteroids and gastrointestinal bleeding in critical care: a systematic review and meta-analysis

, ,


Background: Current belief suggests that patients receiving corticosteroids have an increased risk of bleeding from stress ulceration and that these patients should receive stress ulcer prophylaxis. This issue is important as many ICU patients receive corticosteroids and the pharmacologic agents used for stress ulcer prophylaxis are associated with adverse events.

Aim: The goal of this systematic review and meta-analysis was to compare the rate of clinically significant (overt) gastrointestinal (GI) bleeding in critically ill patients receiving corticosteroids versus placebo.

Methods: We searched PubMed, Embase, and the Cochrane database from inception through December 2018. In addition, the bibliographies of selected articles were reviewed for relevant studies and included if inclusion criteria were met. Included studies were randomized, placebo-controlled and blinded studies that compared treatment with corticosteroids for any indication in a patient population that included only ICU patients. Primary outcome of interest was rate of clinically significant GI bleeding in patients treated with corticosteroids versus placebo. Results were expressed as risk ratio (RR) with accompanying 95% confidence interval (CI). Heterogeneity, sensitivity analysis, and risk of bias were explored. In addition, we did a subgroup analysis according to the use of “low-dose” (<400 mg hydrocortisone or equivalent/day) versus “high-dose” corticosteroid.

Results: Thirty-five studies, which enrolled 16,659 patients, met inclusion criteria and were analyzed. Significant GI bleeding was recorded for 355 patients (overall rate of 2.1%). Summary data demonstrated no difference in the risk of GI bleeding between those treated with corticosteroids versus placebo (RR 1.08; 95% CI 0.88-1.33; p=0.46) with minimal heterogeneity between studies (Q statistic p=0.86, I2=0%). Similarly, there was no significant difference in the risk of GI bleeding in either the low (RR 1.04; 95% CI 0.78-1.38) or the high dose groups (RR 1.13; 95% CI 0.84-1.53) and in those studies at low risk of bias (RR 1.16; 95% CI 0.91-1.49) and those at a high risk of bias (RR 0.88; 95% CI 0.6-1.28).

Conclusion: This meta-analysis did not identify a clinically significant difference in the rate of overt GI bleeding in critically ill patients receiving corticosteroids as compared to placebo. The role of stress ulcer prophylaxis in these patients remains uncertain.