The ideal Ileal-Pouch design : a long-term randomized control trial of J- vs W-pouch construction

<b>BACKGROUND:</b> The IPAA has become established as the preferred technique for restoring intestinal continuity postproctocolectomy. The ideal pouch design has not been established. W-pouches may give better functional results owing to increased volume, whereas the J-pouch’s advantage is its straightforward construction. We report short- and long-term results of an randomized control trial designed to establish the ideal pouch.<br /> <br /><b>DESIGN: </b>Ninety-four patients were randomly assigned to J- and W-pouches (49:45) and assessed at 1 and 8.7 years postoperatively. Assessment was questionnaire based and designed to assess pouch function and patient quality of life.<br /> <br /><b>RESULTS:</b> Eighty-five percent of patients were followed up at 1 year, and 68% were followed up at 8.7 years. At 1 year, there was a significant difference in 24-hour bowel movement frequency J- vs W-pouches 7 vs 5(p < 0.001) and in daytime frequency J- vs W-pouches 6 vs 4 (p < 0.001), with no difference in nocturnal function. At 9-year follow-up, function had equilibrated between the 2 groups: 24-hour bowel movement frequency J- vs W-pouches 6.5 vs 6 (p = 0.36), daytime frequency 5.5 vs 5 (p = 0.233), and nocturnal function 1 vs 1 (p = 0.987). Mean operating time of J- and W-pouches was 195 and 215 minutes (p < 0.05). All other parameters, pad usage, urgency, incontinence, and quality of life, did not differ significantly between groups.<br /> <br /><b>CONCLUSION:</b> These data demonstrate that the theoretical functional advantage conferred on the W-pouch by its greater volume exists only in the short term and is of little consequence to patients’ long-term quality of life. This advantage is attenuated as the pouches mature, resulting in no disparity in pouch function. This, combined with the more consistent, efficient, and easily taught construction of the J-pouch, should conclusively establish it as the optimum ileal-pouch design.