Comparison of Different Operative Approaches for Lumbar Disc Herniation: A Network Meta-Analysis and Systematic Review

Background: New approaches and technologies can be beneficial for patients but also bring corresponding complications. Traditional pairwise meta-analyses cannot be used to comprehensively rank all surgical approaches. Objectives: The purpose of this systematic review and network meta-analysis (NMA) was to compare the outcomes of different surgical approaches for lumbar disc herniation (LDH). Study Design: NMA of randomized controlled trials (RCTs) for multiple treatment comparisons of LDH. Methods: The PubMed, Embase, MEDLINE, Ovid, and Cochrane Library databases were searched for RCTs comparing different surgical approaches for patients with LDH from inception to February 10, 2020. The Markov chain Monte Carlo methods were used to perform a hierarchical Bayesian NMA in WinBUGS version 1.4.3 using a random effects consistency model. The primary outcomes were disability and pain intensity. The secondary outcomes were complications and reoperation. The PROSPERO number was CRD42020179406. Results: A total of 22 trials including 2529 patients and all 5 different approaches (open discectomy or microdiscectomy [OD/MD], microendoscopic discectomy [MED], percutaneous endoscopic discectomy [PED], percutaneous discectomy [PD], and tubular discectomy [TD]) were retrospectively retrieved. PED had the best efficacy in improving patients' dysfunction with no statistical significance (probability = 50%). PD was significantly worse than OD/MD, MED, and PED in relieving patients' pain (standardized mean differences: 0.87 [0.03, 1.76], 0.94 [0.06, 1.88], and 1.02 [0.13, 1.94], respectively). There was no statistically significant difference between any 2 surgical approaches in dural tear; intraoperative, postoperative, and overall complications; or reoperation rate. PED had the lowest dural tear rate and the lowest intraoperative and overall complication rates (probability = 51%, 67%, and 33%, respectively). TD had the lowest postoperative complication and reoperation rates (probability = 35% and 39%, respectively). Limitations: The limitations of this NMA include the inconsistent follow-up times, the criteria for complications, and the reasons for reoperation. Conclusions: Compared with other approaches used to treat LDH, PED had the best safety and efficacy in general, and TD had the lowest reoperation rate. Finally, we recommended PED for LDH.