Use of prophylactic mesh during initial stoma creation to prevent parastomal herniation: a systematic review and meta‐analysis of randomised controlled trials
Syed Mohiuddin, William Hollingworth, Niroshini Rajaretnam, Barnaby C Reeves, Neil Smart
Abstract
Abstract Aim Parastomal hernia (PSH) is a common complication following stoma creation. Previous reviews found mesh reinforcement during initial stoma creation beneficial in reducing PSH incidence. Since then, several multicentre randomised controlled trials (RCTs) produced widely ranging results rendering previous findings debatable. This current review assessed whether combining the latest larger multicentre RCTs would alter the previous findings. Methods The Cochrane Library, MEDLINE and Embase were searched from the respective dates of inception until 15 January 2021. RCTs were included if they compared mesh with no mesh during initial stoma creation in adult patients to prevent PSH. Included RCTs were summarised narratively and meta‐analysed to estimate the relative risk (RR) of PSH incidence (primary analysis), peristomal complications and PSH repair (secondary analyses). Several subgroup analyses were performed, including mesh type (synthetic/biologic), surgical technique (open/laparoscopic) and mesh position (sublay/intraperitoneal). Results Thirteen RCTs were included in the primary meta‐analysis (1070 patients); PSH incidence was reduced in patients with mesh compared with patients without mesh at maximal follow‐up (RR = 0.54; 95% CI 0.39–0.77; I 2 = 67%; P < 0.01). The number of PSH repairs was fewer in patients who had mesh (RR = 0.63; 0.35–1.14; I 2 = 6%; P = 0.39), with no difference in peristomal complications (RR = 0.96; 0.55–1.70; I 2 = 0%; P = 0.71), comparing with no mesh. Subgroup analyses suggested that placing synthetic mesh using an open sublay technique might be more beneficial. Conclusions Prophylactic mesh reinforcement during initial stoma creation reduces PSH incidence and potentially its repair, without an increase in peristomal complications. However, substantial heterogeneity among included RCTs limits confidence in the results.