Gastrogastric fistula following Roux-en-Y bypass is attributed to both surgical technique and experience.

BACKGROUND: The stomach can either be divided or undivided in performing Roux-en-Y gastric bypass (RGB) for morbid obesity. We evaluated whether surgical technique is the sole contributing factor to the formation of gastrogastric fistula (GGF). METHODS: A retrospective analysis of 1,036 consecutive patients was evaluated. RGB was performed as open undivided, open divided, and laparoscopic (divided). Incidence of GGF was identified for each technique and its relationship to surgical experience was assessed. RESULTS: Overall incidence of GGF was 1.3%. All fistulae occurred in patients who received undivided open RGB. There was a significant difference between the undivided open group and the divided open+laparoscopic groups (2.1% vs 0%, P<.01). Incidence of GGF decreased over time with increasing open undivided RGB volume. CONCLUSIONS: GGF was only identified in undivided RGB. The occurrence decreased with increasing surgical experience. Together, overall surgical technique in addition to gastric division must play a role in fistula formation.

Should bariatric revisional surgery be avoided secondary to increased morbidity and mortality?

BACKGROUND: Revisional bariatric surgery may be necessary due to inadequate weight loss or postoperative complications of the primary operation. We sought to identify the reasons for revision, characteristics of the surgery, and outcomes. We hypothesize that revisional surgery, although technically challenging, can produce desirable outcomes. METHODS: Patients undergoing bariatric surgery at our institution between 1998 and 2007 were reviewed from a prospective database. Patients who had revisional surgery were compared to those who had primary surgery. RESULTS: We have identified 46 of 1,038 patients who underwent revisional surgery. Twenty of 46 had a primary Roux-en-Y gastric bypass. The most common indication for revisions is inadequate weight loss secondary to gastrogastric fistula (15/20). Leaks occurred more frequently following revisional surgeries (11% vs 1.2%), but intensive care unit (ICU) utilization was less (11% vs 4.4%) and mortality was lower (0% vs .3%) with bariatric revision surgery. CONCLUSIONS: Although we saw a 9-fold increase in leaks, a 2-5 fold increase in ICU utilization, and 1.5-fold increase in length of stay, our mortality rate was zero. In experienced hands, bariatric revision surgery can be performed to produce desirable outcomes.