https://gastroenterology.acponline.org/archives/2020/08/28/7.htm

In high-risk patients with acute upper GI bleeding, urgent vs. early endoscopy did not differ for 30-day mortality

While a randomized trial found no better outcomes with endoscopy within six hours of gastroenterology consult, “middle-of-the-night” endoscopy should still be considered for certain higher-risk patients who were excluded from the trial, an ACP Journal Club commentary noted.


In a randomized controlled trial, performing urgent endoscopy for acute upper GI bleeding (UGIB) did not improve outcomes compared to completing the procedure within 24 hours. Researchers in China randomized 258 patients to receive endoscopy within six hours and 258 to receive endoscopy six to 24 hours after gastroenterology consult. Patients were ages 18 years and older and had acute UGIB and a Glasgow-Blatchford score (GBS) of 12 or more out of 23. The primary outcome of 30-day mortality occurred in 8.9% of the urgent-endoscopy group and in 6.6% of the early-endoscopy group (difference, 2.3 percentage points; 95% CI, −2.3 to 6.9 percentage points). Rates of further bleeding within 30 days were also not significantly different between groups.

The study was published on April 2 by The New England Journal of Medicine and was summarized in the April 24 ACP Gastroenterology Monthly. The following commentary, by Heidi Ahmed, MD, and David R. Lichtenstein, MD, appeared in the ACP Journal Club section of the Aug. 18 Annals of Internal Medicine.

Endoscopy is a time-sensitive procedure that allows for diagnosis and treatment of UGIB. Endoscopic treatment of selected lesions provides immediate hemostasis and reduces recurrent bleeding, blood transfusions, and need for surgery. Current U.S. and international guidelines recommend endoscopy within 24 hours of presentation for most patients with nonvariceal UGIB, placing the initial focus on early resuscitation to optimize hemodynamic parameters and comorbid conditions

Decisions about endoscopy timing depend on risk stratification of patients. The GBS integrates nonendoscopic parameters into a score for predicting recurrent bleeding and mortality. Low-risk patients without hemodynamic compromise or serious comorbid conditions who have endoscopy within the first few hours of admission can be more expeditiously managed, leading to early hospital discharge and reduced resource utilization, with no effect on mortality. The lowest-risk patients (GBS ≤1) may be discharged from the emergency department with subsequent outpatient endoscopic evaluation.

The benefit of urgent endoscopy in the predicted high-risk group remained unknown as studies yielded conflicting results or focused on low-risk patients. Lau and colleagues provide the first randomized trial assessing endoscopy timing in this subset of high-risk patients (GBS ≥12) with UGIB. They found no difference between urgent (≤6 h after consultation) and early (6 to 24 h after consultation) endoscopy for 30-day all-cause mortality, recurrent bleeding, transfusion requirement, or length of hospital stay. Endoscopic therapy was done more frequently in the urgent endoscopy group, probably because longer proton-pump inhibitor therapy reduced endoscopic stigmata of hemorrhage in the early endoscopy group.

This important and well-designed trial indicates that urgent endoscopy does not improve clinical outcomes in high-risk patients with UGIB, and we can continue to perform endoscopy within 24 hours of presentation in almost all patients. “Middle-of-the-night” endoscopy should still be considered for patients with suspected variceal bleeding or those with persistent hypotensive shock despite volume resuscitation, who were excluded from the [acute] UGIB trial.