Perioperative mortality in bariatric surgery: meta-analysis

BACKGROUND: Bariatric surgery is an established treatment for severe obesity; however, fewer than 1 per cent of eligible patients undergo surgery. The perceived risk of surgery may contribute to the low uptake. The aim of this study was to determine perioperative mortality associated with bariatric surgery, comparing different operation types and data sources. METHODS: A literature search of Ovid MEDLINE, Embase, and the Cochrane Central Register of Controlled Trials was conducted to identify studies published between 1 January 2014 and 31 July 2020. Inclusion criteria were studies of at least 1000 patients reporting short-term mortality after bariatric surgery. Data were collected on RCTs. Meta-analysis was performed to establish overall mortality rates across different study types. The primary outcome measure was perioperative mortality. Different operation types were compared, along with study type, in subgroup analyses. The study was registered at PROSPERO (2019: CRD 42019131632). RESULTS: Some 4356 articles were identified and 58 met the inclusion criteria. Data were available on over 3.6 million patients. There were 4707 deaths. Pooled analysis showed an overall mortality rate of 0.08 (95 per cent c.i. 0.06 to 0.10; 95 per cent prediction interval 0 to 0.21) per cent. In subgroup analysis, there was no statistically significant difference between overall, 30-day, 90-day or in-hospital mortality (P = 0.29). There was no significant difference in reported mortality for RCTs, large studies, national databases or registries (P = 0.60). The pooled mortality rates by procedure type in ascending order were: 0.03 per cent for gastric band, 0.05 per cent for sleeve gastrectomy, 0.09 per cent for one-anastomosis gastric bypass, 0.09 per cent for Roux-en-Y gastric bypass, and 0.41 per cent for duodenal switch (P < 0.001 between operations). CONCLUSION: Bariatric surgery is safe, with low reported perioperative mortality rates.