Mo1731 TREATMENT OF DEFINITIVE DIVERTICULAR HEMORRHAGE: SHORT- AND LONG-TERM OUTCOMES OF COLON SURGERY VS. COLONOSCOPIC HEMOSTASIS

With improvement in endoscopic treatments, the paradigm for treatment of colonic diverticular bleeding has been shifted toward minimally invasive management. However, colon resection is still required in some patients who have severe or uncontrolled bleeding after endoscopic hemostasis. Long-term outcomes comparing surgical management and endoscopic management in diverticular bleed have not been reported. To compare the short and long-term outcomes of patients treated with colon resection (CR) and colonoscopic treatment (EndoRx) for definitive diverticular hemorrhage. This is a retrospective study of prospectively collected data from two teaching hospitals from January 1993 to September 2019. Diagnosis of definitive diverticular (TIC) bleeding was from urgent colonoscopy, red blood cell scans, angiography, or intraoperative findings based on stigmata of recent hemorrhage. Univariate analysis was used to compare demographics and outcome variables between the two groups. The estimate time-to-rebleeding event was performed using Kaplan-Meier analysis. 75 patients were included, 15 with CR and 60 cases with EndoRx. Median (IQR) PRBCs transfused was 7 (2-8) in CR group vs. 2 (1-4) units in EndoRx group (p<0.01). Median length of stay in CR vs. EndoRx was 14 (9-21) vs. 4 (3-5) days, p<0.01. Median ICU stay was 4 (2-5) vs. 0 (0-1) days, p<0.01. Overall post-procedure complication rates were 33.3% vs. 3.3%, p<0.01. Major postoperative complication rates were 6.7% vs. 1.7%, p=0.282 (1 anastomotic leakage in CR group and 1 bleeding from post-coagulation site in EndoRx group). The 30-day rebleeding rate was 0% vs. 3.3%, p=0.677. The 30-day readmission rate was 6.7% vs. 5%, p=0.797. The cumulative incidences of TIC rebleeding for CR vs. EndoRx at 1, 2, 5, and 10 years were 0% vs. 10%, 6.7% vs. 11.7%, 6.7% vs. 23.2% and 6.7% vs. 26.7%. For rebleeding and follow-up outcomes, see Table 1. Median (IQR) follow-up time was 54 (26-95) vs. 52 (26-81) months, p=0.756. On Kaplan-Meier analysis, the TIC rebleeding rate was higher in the EndoRx group, but it was not statistically significant (Figure 1). The all-cause mortality rate was 40% vs. 41.7%, p=0.907. No mortality from diverticular bleeding occurred in either group. 1. Patients in EndoRx group had fewer RBC transfusions, shorter hospitalization, and fewer ICU days. 2. The overall post-procedure complication rate was significantly higher in colon resection group but there was no difference in major post-procedure complications. 3. The rebleeding rate in EndoRx increased every year but was not significantly higher than colon resection group. 4. The long-term all-cause mortality rate was 41.3% but no patient died of TIC bleeding.Figure 1Kaplan-Meier analysis estimates of recurrent diverticular bleed after CR and EndoRx showed that the rebleeding rate of was higher in EndoRx group but was not statistically significant (Log rank test, p=0.113).View Large Image Figure ViewerDownload Hi-res image Download (PPT)