Timing of emergency gastroscopy (door to endoscopy time) in variceal and non-variceal upper gastrointestinal bleeding: a retrospective cohort analysis in a tertiary center

Aims International guidelines recommend performing emergency gastroscopy within 12 hours for variceal (V) bleeding and within 24 hours for non-variceal (NV) bleeding in cases of upper gastrointestinal bleeding (UGIB). This study aimed to analyze the time elapsed between emergency department admission and time of endoscopy, referred to as 'door-to-endoscopy time' (DET), and to identify factors influencing DET across different UGIB subgroups.

Methods We conducted a retrospective cohort analysis of patients who underwent emergency gastroscopy for UGIB between January 1, 2022, and December 31, 2023, at the Department of Emergency Medicine at Semmelweis University. Data were collected on clinical parameters and 30-day mortality. NV patients were categorized into high-risk (HNV) and low-risk (LNV) groups based on endoscopic findings (Forrest classification). Welch’s t-test was used for comparing mean DET among the subgroups, whereas for hemoglobin linear regression was applied. The significance level was set at 0.05 for all analyses.

Results Among 654 patients, we identified 126 V and 528 NV UGIB cases, with the latter comprising 99 HNV and 429 LNV cases. Rebleeding occurred in 5.4% of patients (V: 12.7%, HNV: 10.1%, LNV: 2.1%), and 30-day mortality was 19.7% in the cohort (V: 24.8%, HNV: 22.6%, LNV: 17.6%). The mean DET was 7.28 hours (SD=±6.04), with subgroup averages as follows: V: 5.87±3.77 hours, HNV: 6.33±4.04 hours, and LNV: 7.92±6.83 hours, with significant differences between LNV-HNV (p=0.003) and LNV-V (p<0.0001) groups; no significant difference was observed between HNV-V (p=0.37). In the entire cohort, shorter DET was significantly associated with tachycardia (6.61 vs. 7.73 hours, 95% confidence interval (CI) -1.99 to -0.25, p=0.01), hematemesis (5.82 vs. 8.09 hours, 95% CI -3.11 to -1.43, p<0.0001), syncope (6.25 vs. 7.57 hours; 95% CI -2.30 to -0.33, p=0.009), and lower hemoglobin levels (1 g/L decrease in hemoglobin reduced DET by an average of 1 minute, p=0.029). DET was not associated with 30-day mortality (p=0.501), vasopressor use (p=0.857), shock index (p=0.091), coffee-ground emesis (p=0.192), hematochezia (p=0.564), intubation (p=0.566), rebleeding rate (p=0.462), or melena (p=0.961). In the V group, DET was significantly shorter in case of hematemesis (5.20 vs. 7.25 hours, 95% CI -3.61 to -0.49, p=0.011) and tachycardia (5.19 vs. 6.64 hours, 95% CI -2.84 to -0.08, p=0.039); whereas in HNV patients with coffee-ground emesis, DET was significantly reduced (4.94 vs. 6.58 hours, 95% CI -3.08 to -0.20, p=0.026); in LNV patients, DET was shorter in patients with hematemesis (6.42 vs. 8.38 hours, 95% CI -3.20 to -0.71, p=0.002) and syncope (6.56 vs. 8.25 hours, 95% CI -3.10 to -0.29, p=0.019).

Conclusions In our retrospective cohort study, DET significantly differed between LNV-V and LNV-HNV subgroups. Shorter DET correlated with tachycardia, hematemesis, syncope, and lower hemoglobin levels. DET showed no association with 30-day mortality or rebleeding rate.

 

MissingLinkText

Conflicts of Interest

Authors do not have any conflict of interest to disclose.

Publication History

Article published online:
27 March 2025