Comparison of interventions for lumbar disc herniation: a systematic review with network meta-analysis.

Background: Implants for use in disc herniation surgery have been commercially available for some time. Several clinical trials have shown promising results. There are now a wide variety of surgical methods for treating lumbar disc herniation.

Purpose: The objective of this systematic review was to compare all current surgical methods for disc herniation, including newer methods with implants for annulus repair and dynamic stabilization.

Study Design: Systematic review and network meta-analysis.

Methods: PRISMA-P guidelines were followed in this review. Literature search in PubMed, Embase, and Cochrane library databases identified eligible randomized controlled trials (RCT) studies comparing interventions for lumbar disc surgery. The investigated outcomes were: changes in pain score, disability score and reoperation rate with a minimum follow-up of 1 year. Risk of bias was assessed in concordance with Cochrane Neck and Back Review Group recommendation. A network meta-analysis was performed using gemtc and BUGSnet software, and each outcome evaluated using Confidence in Network Meta-Analysis (CINeMA).

Results: Thirty-two RCT studies, with 4,877 participants, and eight different interventions were identified. A significant difference was seen in change of pain score, as all treatments were superior to conservative treatment and percutaneous discectomy. This difference was only found to be of clinically importance when comparing conservative treatment and dynamic stabilization. There was no significant difference in reoperation rates or change in disability score, regardless of treatment. However, SUCRA plots showed a trend in ranking annulus repair and dynamic stabilization highest. Risk of bias assessment showed that 15 studies had a high overall risk of bias. Meta-regression with risk of bias as covariate did not indicate any influence in risk of bias on the model. Confidence in Network Meta-Analysis evaluation showed a high level of confidence for all treatment comparisons.

Conclusions: With this network meta-analysis, we have aimed to compare all treatments for herniated lumbar disc in one large comprehensive systematic review and network meta-analysis. We have compared across the three main outcomes: disability score, pain score and reoperation rate. We were not able to rank one single treatment as the best. Most of the treatment performed at the same level. However percutaneous discectomy and conservative treatment consistently performed worse than the other treatments. In general, the CINeMA evaluation according to the GRADE recommendations gave a high level of confidence for the study comparisons.