A Matter of Time: Does Early vs Late Failure Impact Redo Pouch Survival?
Related Experiment Video
Updated: Nov 4, 2026
Arterial Pouch Microsurgical Bifurcation Aneurysm Model in the Rabbit
Published on: May 14, 2020
Diana C Jimenez1, Olga Lavryk1, Ali Alipouriani1
1Dept. of Colorectal Surgery, Digestive Diseases Institute, Cleveland Clinic, Cleveland OH.
Background:
Anastomotic leaks, pelvic sepsis, and related complications are established predictors of pouch failure after ileal pouch-anal anastomosis (IPAA). However, the effect of the timing of index pouch failure on outcomes after redo pouch surgery remains poorly understood.
Objective:
To evaluate whether the timing of index pouch failure predicts redo pouch survival.
Design:
Retrospective cohort study.
Setting:
Single quaternary referral center.
Patients:
Patients with prior IPAA who underwent redo pouch surgery between 2000 and 2024.
Main Outcome Measures:
The primary outcome was redo pouch failure, defined as a composite of pouch excision, permanent rediversion after redo pouch creation, or failure to achieve ileostomy closure. The primary exposure was the timing of index pouch failure, categorized as early failure (rediversion ≤ 6 months after index IPAA) or late failure (rediversion > 6 months).
Results:
A total of 320 patients met inclusion criteria, including 57 with early index pouch failure and 263 with late failure. Index pouch leak (50.9% vs. 33.5%; p=0.01) and pelvic abscess (36.8% vs. 14.1%; p<0.001) were more common in the early group. Ileostomy closure rates after redo pouch surgery were similar (96.5% vs. 88.6%; p=0.07). On multivariable Cox proportional hazards regression, late index failure (HR 3.09; 95% CI 1.07-8.87; p=0.04) and inflammatory indications for redo (HR 2.40; 95% CI 1.15-5.0; p=0.02) were independent predictors of redo pouch failure. Secondary pouch failure occurred in 10.7% of the early group vs. 22.8% of the late group (p=0.04). Sensitivity analyses using bootstrap internal validation confirmed the model's reliability, with an optimism-corrected C-index of 0.64; the low model optimism (3.9%) suggests the findings are robust and likely to generalize to other patient populations.
Limitations:
Retrospective design and single-center setting may limit generalizability.
Conclusions:
Late failure after index IPAA was associated with increased risk of subsequent pouch failure after redo IPAA. See Video Abstract.
