Twenty-five studies (n=11,429 patients) were included in the review; four were RCTs (n=358 patients) and 21 were non-randomised studies (n=11,071 patients). All RCTs used an intention-to-treat analysis and made an attempt at allocation concealment. Blinding was not possible due to the nature of the intervention.
Meta-analysis of RCTs: The pooled analysis showed a statistically significantly lower rate of clinical anastomotic leak in groups with a stoma (RR 0.39, 95% CI 0.23 to 0.66; four RCTs) and also a statistically significantly lower rate of reoperation for leak (RR 0.29, 95% CI 0.16 to 0.53; four RCTs). There was no statistically significant difference between the groups in mortality related to leak. There was no evidence of statistically significant heterogeneity between the trials for any outcome.
Meta-analysis of non-randomised studies: As with the RCTs, pooled analyses showed a statistically significantly lower rate of clinical anastomotic leak in groups with a stoma (RR 0.74, 95% CI 0.67 to 0.83; 21 studies) and also a statistically significantly lower rate of reoperation for leak (RR 0.28, 95% CI 0.23 to 0.35; 15 studies). These studies also showed a statistically significant lower rate of mortality in patients without a stoma (RR 0.42, 95% CI: 0.28 to 0.61; 19 studies). Statistically significant heterogeneity was detected only for the risk of leak (I2=75%).
Morbidity and mortality after routine closure of defunctioning stoma (10 studies including four RCTs): Routine stoma closure was associated with morbidity rates ranging from 0 to 19.9% and mortality rates ranging from 0 to 1.4%.