Primary anastomosis and nonrestorative resection for perforated diverticulitis with peritonitis: meta-analysis of randomized trials

Aim The aim of this meta-analysis was to comparatively evaluate the outcomes of primary anastomosis (PRA) and nonrestorative resection (NRR) as emergency surgery and ostomy reversal in patients with perforated diverticulitis and peritonitis. Methods PubMed, MEDLINE via Ovid, Embase, CINAHL, Cochrane Library and Web of Science databases were systematically searched. Postoperative morbidity following emergency resection was the primary end-point. Quality assessment of the included studies was performed using the Cochrane Quality Assessment Tool including recruitment bias and crossover with intention-to-treat analysis. The Haenszel-Mantel method with odds ratios (OR, 95% CI) and the inverse variance method with mean difference (MD, 95% CI) as effect measures were utilized for dichotomous and continuous outcomes, respectively. Results Four randomized controlled trials totaling 382 patients (180 PRAvs204 NRR) were included. Morbidity rates following emergency resection did not differ (OR = 0.99, 95% CI 0.65, 1.51;P = 0.95; number needed to treat/harm (NNT) 96). Organ/space surgical site infection rates were 3.3% in PRAvs11.3% in NRR (OR = 0.29, 95% CI 0.12, 0.74;P = 0.009; NNT = 13). Postoperative morbidity rates following ostomy reversal were significantly lower in PRA (OR = 0.31, 95% CI 0.15, 0.64;P = 0.001; NNT = 7). Pooled ostomy non-reversal rates were 16% in PRAvs35.5% in NRR (OR = 0.37, 95% CI 0.22, 0.62;P = 0.0001; NNT = 6) with high heterogeneity (I-2 = 63%;tau(2) = 8.17). Meta-regression analysis revealed significant negative correlation between the PRA-to-NRR crossover rate and the ostomy non-reversal rate (P = 0.029). Conclusion This meta-analysis found that PRA was associated with better short- and long-term outcomes at the cost of significantly longer operating time at emergency surgery.