Hiatal hernia repair during laparoscopic sleeve gastrectomy: systematic review and meta-analysis on gastroesophageal reflux disease symptoms changes

Abstract

Initial reference search yielded 831 articles. After removing duplicates, 601 articles were evaluated through titles and abstracts. This produced 32 papers suitable for full-text review. Four articles were excluded due to wrong study design, 3 because of the wrong type of intervention, 2 – the lack of primary outcome, 7 were duplicates and 11 were conference abstracts only. Finally, we enrolled 1 RCT and 4 comparative studies with a total of 838 patients, 307 in the LSG + HHR group and 531 in the LSG group (Tab. I.) [5, 10–13]. The flowchart of literature search and study selection is summarized in Fig. 1. </br> </br> GERD improvement was reported by 5 authors. Fig. 2. shows analysis based on non-event analysis. There were no significant differences between the analyzed studies, 107/171 (62.6%) in HHR+ vs. 78/131 (59.5%) in HHR-, P = 0.74. Heterogeneity between the analyzed studies was moderate, I2 = 66% (Fig. 2.). The “de-novo” GERD was reported in 5 studies. There were no statistically significant differences between the analyzed groups, 26/139 (18.7%) in HHR+ vs. 101/403 (25.1%) in HHR-, P = 0.77. Heterogeneity between the analyzed studies was moderate, I2 = 73% (Fig. 3.). null Operative time was reported by 3 authors. In all studies, LSG with HHR was significantly longer than LSG alone. When the random effects model was used, the difference between groups was statistically insignificant with heterogeneity of I2 = 98%. Sensitivity analysis showed that the study by Lasnibat is responsible for all heterogeneity. After excluding this study the difference was statistically significant (I2 = 0%, P < 0.01). Since all studies showed operative time in favor of LSG, despite high heterogeneity, a fixed effects model was used. </br> </br> Morbidity was reported in 3 studies. There were no statistically significant differences in any study as well as when the data was pooled, 9/132(6.8%) in LSG +HHR vs. 11/148(7.4%); P = 0.97. The heterogeneity between the studies was low, I2 = 8% (Fig. 5.).