The « DEBE » score predicts the risk of delayed bleeding after colorectal endoscopic dissection in a large European multicenter cohort

Aims Clinically significant delayed bleeding (DB) is an important adverse event after endoscopic submucosal dissection (ESD) of large non-pedunculated colorectal polyps. The Limoges scores¹ estimated an individualized DB risk after ESD. The aim of this study is to evaluate the incidence and risk factors of DB after ESD in a large European cohort, to validate the Limoges score and to create a new updated one.

Methods Procedural data and complications occurring within 30 days after ESD were prospectively recorded in French, Belgian, and Spanish centers and analyzed retrospectively. The Limoges score was tested in the European cohort. Multivariate logistic regression analysis on risk factors allowed the development of a prediction model using weighted points in a risk score. Model accuracy was determined by discrimination and calibration. Internal validation was performed by bootstrapping and leave-one-out cross-validation. The score was studied in two geographical areas (Spain, France/Belgium).

Results 4771 ESD procedures were included : 23.7% rectal ESD ; median lesion size 50 mm ; mean age 68.2 years-old ; 72.9% ASA score I-II. Anticoagulants use in 10.9% and antiplatelet in 17%. DB was observed in 6.8% of cases. The discrimination of “Limoges score” was modest (AUC 0.682) A new European model called DEBE (DElayed Bleeding Esd) score was developed with the main risk factors: age≥75 years (2 points), lesion size≥50 mm (5 points), ASA classification III-IV (4 points), location in the rectum (2 points) or in the proximal colon (1 point), anticoagulants (7 points) or antiplatelets (3 points). The DEBE score ranges from 0 (lowest risk) to 23 points (highest) and divided patients into a low-risk group (DB risk 3.9%) and an intermediate-high-risk group (14.2%). The DEBE score shows better discrimination (AUC 0.712 (95% CI 0.682 – 0.743). Calibration and Hosmer-Lemeshow test (chi-square 4.322, p=0.827) were satisfactory in the derivation cohort. Cross-validation also showed an adequate model performance of 72.6% and demonstrated adequate performance in both datasets [1].

Conclusions -A new developed European model called the DEBE score has an improved and validated performance to identify the individual risk of bleeding after colorectal ESD.

-The model is based on 5 pre-procedural variables and allows a personalized assessment of the bleeding risk. It permits the clinician to inform the patient about the DB risk, to select the patients most likely to benefit from prophylactic treatments and to define those requiring strict monitoring after the procedure.

Publication History

Article published online:
27 March 2025

• References

• 1 Albouys J, Montori Pina S, Boukechiche S. et al. Risk of delayed bleeding after colorectal endoscopic submucosal dissection: the Limoges bleeding score. Endoscopy 2024; 56: 110-118