Network meta-analysis of urinary retention and mortality after Lichtenstein repair of inguinal hernia under local, regional or general anaesthesia

Background: Urinary retention and mortality after open repair of inguinal hernia may depend on the type of anaesthesia. The aim of this study was to investigate possible differences in urinary retention and mortality in adults after Lichtenstein repair under different types of anaesthesia. Methods: Systematic searches were conducted in the Cochrane, PubMed and Embase databases, with the last search on 1 August 2018. Eligible studies included adult patients having elective unilateral inguinal hernia repair by the Lichtenstein technique under local, regional or general anaesthesia. Outcomes were urinary retention and mortality, which were compared between the three types of anaesthesia using meta-analyses and a network meta-analysis. Results: In total, 53 studies covering 11 683 patients were included. Crude rates of urinary retention were 0·1 (95 per cent c.i. 0 to 0·2) per cent for local anaesthesia, 8·6 (6·6 to 10·5) per cent for regional anaesthesia and 1·4 (0·6 to 2·2) per cent for general anaesthesia. No death related to the type of anaesthesia was reported. The network meta-analysis showed a higher risk of urinary retention after both regional (odds ratio (OR) 15·73, 95 per cent c.i. 5·85 to 42·32; P < 0·001) and general (OR 4·07, 1·07 to 15·48; P = 0·040) anaesthesia compared with local anaesthesia, and a higher risk after regional compared with general anaesthesia (OR 3·87, 1·10 to 13·60; P = 0·035). Meta-analyses showed a higher risk of urinary retention after regional compared with local anaesthesia (P < 0·001), but no difference between general and local anaesthesia (P = 0·08). Conclusion: Local or general anaesthesia had significantly lower risks of urinary retention than regional anaesthesia. Differences in mortality could not be assessed as there were no deaths after elective Lichtenstein repair. Registration number: CRD42018087115 (https://www.crd.york.ac.uk/prospero).