Laparoscopic versus open feeding jejunostomy in patients with upper gastrointestinal conditions: a systematic review, meta-analysis with meta-regression and GRADE assessment

Postoperative nutritional support via a feeding jejunostomy (FJ) is vital for patients undergoing upper gastrointestinal surgery. This study evaluates the comparative safety and efficacy of laparoscopic feeding jejunostomy (LFJ) versus open feeding jejunostomy (OFJ). Systematic review and meta-analysis of RCTs and comparative cohort studies comparing LFJ and OFJ, identified from major databases up to May 5, 2026. Primary outcomes: operative time and early mortality. Secondary outcomes: hospital stay, minor/major complications, time to enteral feeding, and day 1 pain scores. Random-effects models estimated RRs, MDs, or SMDs. Robustness was examined with leave-one-out sensitivity analyses; Galbraith plots were used to explore between-study heterogeneity rather than to test robustness. Meta-regression analyzed age and BMI as covariates. The certainty of evidence was appraised using the GRADE framework. Fifteen studies (n = 2,300; 1,290 LFJ, 1,010 OFJ) were included. LFJ significantly reduced minor postoperative complications (RR = 0.47, 95% CI: 0.32–0.67; I² = 0%), shortened hospital stay (MD = -1.09 days, 95% CI: -1.42 to -0.75; I² = 60.8%), accelerated enteral feeding (MD = -0.24 days, 95% CI: -0.34 to -0.13; I² = 0%), and lowered day 1 pain scores (SMD = -0.56, 95% CI: -0.87 to -0.24; I² = 0%). No significant differences were found in early mortality or major complications. Operative time showed high heterogeneity (I² = 98%), rendering evidence inconclusive. Sensitivity analyses indicated that some pooled estimates were sensitive to the exclusion of individual studies. In this predominantly non-randomized evidence base, LFJ was associated with fewer minor complications, shorter hospital stay, and earlier enteral feeding, with no clear difference in early mortality or major complications. Because most data came from retrospective cohorts subject to confounding by indication, and the certainty of evidence ranged from very low to moderate across outcomes, these associations should be interpreted with caution and regarded as supportive rather than confirmatory. Nevertheless, further high-quality randomized controlled trials are warranted to strengthen the evidence base and better define the optimal surgical approach across different patient populations.