Primary: to describe the incidence of delayed complications after CR-ESD (bleeding, perforation, and post-electrocoagulation syndrome [PECS]). Secondary: to characterise their timing and clinical impact; to identify associated risk factors; and to evaluate the effect of prophylactic defect closure.

This retrospective cohort study used data from a prospective national registry of CR-ESDs performed in 28 Spanish centres (January 2013-September 2025) by 40 endoscopists. Eligible patients were aged >18 years and underwent CR-ESD or attempted ESD for colorectal neoplasia. Exclusion criteria were incomplete records, technical variants (hybrid ESD, endoscopic intermuscular dissection [EID] or per-anal endoscopic myotomy [PAEM]), aborted procedures, immediate surgery after ESD, failed ESDs, and low-volume centres.

Delayed complications were defined as clinically significant bleeding, PECS (pain/fever without intraprocedural perforation requiring >1-day stay), or perforation confirmed radiologically occurring after completion of the procedure. Events not prolonging hospitalisation were not considered complications. Main data source was the Spanish National ESD Registry. Follow-up followed routine practice.

Univariable logistic regression assessed crude associations, and all clinically relevant variables entered multivariable model. The effect of prophylactic closure was evaluated using crude comparisons and propensity-score matching (PSM) to control for confounding. Analyses were two-sided with significance set at p<0,05. Deterministic logical rules were used for imputation when values could be unequivocally deduced; otherwise, missing data were retained. No sample-size calculation was performed.

A total of 2.618 CR-ESDs were performed in 2.556 patients. After exclusions (179 hybrid ESDs, 28 EIDs, 38 PAEMs, 101 aborted procedures, 21 immediate surgeries, 133 incomplete records, 6 failed ESDs), 2.112 procedures were analysed. Intraprocedural perforations (267; 12,6%) were excluded from PECS and delayed perforation analyses.

The cohort had a mean age of 67,7 years; 59% were male; 88% were ASA II–III; 14% received antiplatelets and 10% anticoagulation. Median lesion size was 39 mm and fibrosis was present in 60%. Technical success was 95,5%, en-bloc resection 90,3%, and R0 resection 78,7%. Median hospital stay was 1 day.

Overall delayed complications occurred in 258 out of 2.112 cases (12,2%; CI95% 10,1-13,7). PECS occurred in 7,0% (CI95% 5,9-8,9) and delayed bleeding in 5,6% (CI95% 4,7-6,7), whereas delayed perforation was rare (1,4%; CI95% 0,9-2,0). The median (IQR) length of hospital stay was 3 (2–4) days for any delayed complication, 3 (2–5) days for PECS, 3 (1–5) days for bleeding, and 7 (6–11) days for perforation. Median time to bleeding was 3 days and was longer in anticoagulated patients (6 vs 2 days). All PECS and bleeding cases were managed non-surgically; 48% of delayed perforations required surgery.

Independent risk factors for overall delayed complications were anticoagulation, larger dissected area, and poor manoeuvrability. PECS was associated with larger area while left-sided location showed a protective effect. Delayed bleeding was strongly associated with anticoagulation and rectal location. Delayed perforation was linked to right-sided and flexural lesions.

Complete prophylactic closure was associated with a lower risk of overall delayed complications and bleeding in crude analyses (OR 0,6; CI95% 0,4-0,8 vs 0,4; CI95% 0,2-0,6). After adjustment using PSM, complete closure reduced the absolute risk of any delayed complication by 8,6% and of delayed bleeding by 5,7%. In anticoagulated patients, complete closure markedly reduced bleeding risk (absolute reduction ~17%; NNT≈6).

In this large multicentre cohort, delayed complications occurred in 12,2% of CR-ESDs, with PECS and bleeding being the most frequent, assuming a median increase of 2 days in the length of hospitalisation. Anticoagulation, lesion complexity (larger area or poor manoeuvrability), and specific anatomical locations emerged as key risk factors. Complete prophylactic closure was associated with a reduction in overall delayed complications and delayed bleeding, with a pronounced benefit in anticoagulated patients.