Primary Anastomosis for Perforated Diverticulitis with Peritonitis: Post-hoc Pooled Analysis of Prospective Randomized Trials
Résumé
Background: The impact of specific interventions at resection with primary anastomosis (PRA) for perforated diverticulitis with peritonitis is controversial. The aim of this pooled analysis was to determine whether any specific interventions performed at resection with primary anastomosis in patients with perforated diverticulitis with peritonitis influenced the outcomes. Methods: A post-hoc analysis of pooled data for 254 patients enrolled in three randomized trials that compared PRA to nonrestorative resection (NRR) (NCT01239927; NCT01233713; NCT00692393) was carried out. The primary endpoint was the postoperative complication rate. All patients were adults with perforated diverticulitis with purulent or fecal peritonitis. Specific interventions performed within PRA included intraoperative on-table colonic lavage, anastomosing technique (hand-sewn or stapled), diverting ostomy, and different types of ostomy (colostomy or ileostomy). Results: A total of 116 PRA patients and 138 NRR patients were included in the randomized trials, but only patients undergoing PRA (n=116) were included in the pooled analysis. A negative correlation was found between colonic lavage and postoperative complication rates (rs=-0.482; p=0.011). Positive correlations were found between postoperative complication rates and both stapled anastomoses (rs=0.224; p=0.019) and the creation of an ostomy (rs=0.327; p<0.001). The type of ostomy was not correlated with postoperative complication rates. Conclusion: Intraoperative colonic lavage and hand-sewn PRA were associated with decreased complication rates. The data seem to suggest that the addition of an ostomy may be correlated with increased rates of complications specifically related to ostomy creation.