IDDF2020-ABS-0194 Timing of endoscopy for acute upper gastrointestinal bleeding: a territory-wide cohort study

Cosmos Liutao Guo, Sunny H. Wong, Louis Ho Shing Lau, Rashid Nok Shun Lui, Joyce Wing Yan Mak, Raymond Shing Yan Tang, William Ka Kei Wu, Grace Lai–Hung Wong, Francis K.l. Chan, James Y. Lau, Joseph J.�y. SungAbstracts 2020 · DOI

Introduction

Acute upper gastrointestinal bleeding (AUGIB) remains a significant medical emergency worldwide, requiring rapid assessment and intervention. While clinical guidelines generally recommend that patients receive an endoscopy—the primary diagnostic and therapeutic tool—within 24 hours of hospital admission, the optimal timing within that window is a subject of ongoing debate.

A common clinical dilemma is whether to perform an "urgent" endoscopy immediately upon presentation or to wait until the patient has been medically stabilized. While rushing to intervene might seem intuitive to stop a hemorrhage, performing procedures on hemodynamically unstable patients carries inherent risks. This study sought to identify the "Goldilocks" zone for endoscopy timing to optimize patient survival and clinical outcomes.

Methodology

Researchers conducted a retrospective, territory-wide cohort study using data from the Hong Kong Hospital Authority’s Clinical Data Analysis and Reporting System (CDARS). This database provides a comprehensive look at electronic health records across all public hospitals in Hong Kong.

The study analyzed 6,474 adult patients (aged 18+) who presented with AUGIB between 2013 and 2019 and received therapeutic endoscopy within 48 hours. To evaluate the impact of timing, patients were categorized into three distinct groups based on the interval between hospital admission and endoscopy:

  1. Urgent: $\le$ 6 hours
  2. Early: 6 to 24 hours
  3. Late: 24 to 48 hours

To account for the non-randomized nature of the data, the team employed Inverse Probability of Treatment Weighting (IPTW). This statistical method balances baseline characteristics—such as age, comorbidities, and initial lab results—across the groups, ensuring that the observed differences in outcomes were primarily attributable to the timing of the procedure rather than the initial severity of the patient's condition.

Key Results

After applying IPTW to balance the cohorts, the study revealed that the "Early" endoscopy group (n = 3,865) achieved significantly better clinical outcomes than both the Urgent and Late groups.

1. Survival and Mortality

Both the Urgent (n = 1,008) and Late (n = 1,601) groups exhibited higher 30-day all-cause mortality compared to the Early group. This suggests a U-shaped risk curve where intervening too early or too late increases the risk of death.

2. Procedural Success

Patients in the Urgent and Late groups had higher rates of repeated therapeutic endoscopy. This indicates that the initial procedure was less likely to be definitive when performed in these windows, possibly due to poor visualization in an unoptimized patient or continued bleeding in a delayed one.

3. Resource Utilization

ICU admission rates were significantly higher for patients who underwent urgent or late endoscopy. The Early group required fewer intensive care resources, suggesting a smoother post-procedural recovery.

Discussion: The Urgency Paradox

The most striking finding is the "Urgency Paradox": performing endoscopy within the first six hours was associated with worse outcomes than waiting until the 6-to-24-hour mark. This highlights the critical importance of the "pre-endoscopic" phase.

The first few hours of admission are often best utilized for aggressive resuscitation—restoring blood volume, stabilizing blood pressure, and administering pharmacotherapy (such as proton pump inhibitors or vasoactive drugs). Rushing a patient to the endoscopy suite before they are hemodynamically stable may increase the risk of procedural complications or sedation-related adverse events.

Conversely, the data confirmed that delaying beyond 24 hours is equally detrimental. The "Late" group likely suffered from prolonged exposure to the effects of uncontrolled bleeding, leading to the observed increases in mortality and ICU stays.

Conclusion

This territory-wide study supports the current recommendation that AUGIB patients should receive endoscopy within 24 hours, but it refines this window by suggesting that the 6-to-24-hour period is optimal. The findings emphasize that while AUGIB is an emergency, medical optimization and stabilization should not be bypassed in a rush to the procedure room.

Future research, particularly prospective randomized controlled trials, is needed to determine if these timing windows should be adjusted based on specific bleeding etiologies (e.g., variceal vs. peptic ulcer bleeding) or specific patient risk scores.

Summary updated

Abstract

<h3>Background</h3> Whilst it is commonly recommended that patients presenting with acute upper gastrointestinal bleeding (AUGIB) should receive endoscopic evaluation within 24 hours, the optimal timing is yet to be established. We aimed to assess, for AUGIB patients, whether endoscopy timing after hospital admission would significantly affect clinical outcomes. <h3>Methods</h3> We conducted a retrospective, territory-wide, cohort study with clinical data extracted from the Hong Kong Hospital Authority (HA) Clinical Data Analysis and Reporting System (CDARS). Patients of age 18 and above that (a) presented with AUGIB to the public hospital between 2013–2019 and (b) received therapeutic endoscopy within 48 hours (n = 6,474) were recruited. Patients were classified into 3 groups based on endoscopic timing after admission: urgent (t ≤ 6), early (6 &lt; t ≤ 24) and late (24 &lt; t ≤ 48). Important baseline characteristics (e.g. blood tests and comorbidities) were balanced with inverse probability of treatment weighting (IPTW). After balancing, the 30-day all-cause mortality, 30-day repeated therapeutic endoscopy rate, 30-day ICU admission rate and other endpoints were compared. We hypothesised that, amongst the three groups, early timing would be associated with the best clinical outcomes. Subgroup analysis was performed based on comorbidity burden and aetiology of bleeding. Sensitivity analysis was performed. <h3>Results</h3> With IPTW, differences in patient characteristics across 3 groups were well balanced. Results showed that urgent (n = 1,008) and late endoscopy (n = 1,601) had worse outcomes compared to early endoscopy (n = 3,865), with higher 30-day mortality, repeat endoscopy rates and ICU admission rates. <h3>Conclusions</h3> Compared to urgent and late endoscopy, early endoscopy timing was significantly associated with superior outcomes. This supports the notion that AUGIB patients should receive endoscopy within 24 hours, but also emphasises the importance of adequate resuscitation, medical optimisation and pharmacotherapy before rushing patients to endoscopy.
BibTeX
@inproceedings{guo2020iddf,
  author = {Cosmos Liutao Guo and Sunny H. Wong and Louis Ho Shing Lau and Rashid Nok Shun Lui and Joyce Wing Yan Mak and Raymond Shing Yan Tang and William Ka Kei Wu and Grace Lai–Hung Wong and Francis K.L. Chan and James Y. Lau and Joseph J.�Y. Sung},
  title = {{IDDF2020-ABS-0194 Timing of endoscopy for acute upper gastrointestinal bleeding: a territory-wide cohort study}},
  booktitle = {Abstracts},
  year = {2020},
  doi = {10.1136/gutjnl-2020-iddf.116},
}
Figure