Clinical outcomes of esophagogastroduodenoscopy in critically ill patients using high-dose proton pump inhibitor for suspected bleeding: A retrospective cohort study

Esophagogastroduodenoscopy (EGD) is a useful procedure performed for gastrointestinal (GI) bleeding. No definite clinical guidelines recommend EGD implementation in intensive care unit (ICU) patients with suspected GI bleeding. The objective of this study was to compare the clinical effectiveness of EGD in critically ill patients who are using high-dose proton pump inhibitor (PPI) for suspected GI bleeding. We retrospectively analyzed ICU patients using high-dose PPI for suspected GI bleeding from January 2012 to September 2020. Major cases of GI bleeding, such as those with hematemesis and hematochezia, were excluded, and 1:1 propensity score matching was performed. The change in hemoglobin level, requirement of red blood cell transfusion, re-suspected bleeding event, length of ICU stay, and ICU mortality were compared between the EGD and non-EGD groups. Of the 174 subjects included, 52 patients underwent EGD within 24 hours of PPI administration. In the EGD group, 22 (42.3%) patients showed normal findings, while esophagitis and gastritis were most common abnormal finding (n = 11, 21.2%), and 14 patients (26.9%) underwent a hemostatic procedure. While comparing the 2 groups, the EGD group required a higher amount of red blood cell transfusion (packs) than the non-EGD group for a week (3.04 ± 0.44 vs 2.07 ± 0.25, P = .01). There was no significant difference in the change in hemoglobin level after 1 week (P = .15). After propensity score matching, the EGD group showed similar the requirement of red blood cell transfusion and change in hemoglobin level for a week (P = .52, P = .97, respectively). In analyses for all patients and propensity score matched patients, there was no statistically significant difference in term of re-suspected bleeding event rate, duration of ICU stay, and ICU mortality. However, re-suspected bleeding event rate and ICU mortality were lower trend in the EGD group than the non-EGD group. This study showed that EGD had no definite clinical benefit in ICU patients using high-dose PPI for suspected GI bleeding and aggressive EGD is not necessarily recommended. However, it is necessary to consider EGD in patients who are tolerant.