Surgical versus non-operative treatment for lumbar disc herniation: a systematic review and meta-analysis

Bing-Lin Chen,Jia-Bao Guo,Hong-Wei Zhang,Ya-Jun Zhang,Yi Zhu,Juan Zhang,Hao-Yu Hu,Yi-Li Zheng,Xue-Qiang Wang
DOI: https://doi.org/10.1177/0269215517719952
2017-07-17
Clinical Rehabilitation
Abstract:Objective: To investigate the effects of surgical versus non-operative treatment on the physical function and safety of patients with lumbar disc herniation. Data sources: PubMed, Cochrane Library, Embase, EBSCO, Web of Science, China National Knowledge Infrastructure and Chinese Biomedical Literature Database were searched from initiation to 15 May 2017. Methods: Randomized controlled trials that evaluated surgical versus non-operative treatment for patients with lumbar disc herniation were selected. The primary outcomes were pain and side-effects. Secondary outcomes were function and health-related quality of life. A random effects model was used to calculate the pooled mean difference with 95% confidence interval. Results: A total of 19 articles that involved 2272 participants met the inclusion criteria. Compared with non-operative treatment, surgical treatment was more effective in lowering pain (short term: mean difference = −0.94, 95% confidence interval = −1.87 to −0.00; midterm: mean difference = −1.59, 95% confidence interval = −2.24 to −9.94), improving function (midterm: mean difference = −7.84, 95% confidence interval = −14.00 to −1.68; long term: mean difference = −12.21, 95% confidence interval = −23.90 to −0.52) and quality of life. The 36-item Short-Form Health Survey for physical functions (short term: mean difference = 6.25, 95% confidence interval = 0.43 to 12.08) and bodily pain (short term: mean difference = 5.42, 95% confidence interval = 0.40 to 10.45) was also utilized. No significant difference was observed in adverse events (mean difference = 0.82, 95% confidence interval = 0.28 to 2.38). Conclusion: Low-quality evidence suggested that surgical treatment is more effective than non-operative treatment in improving physical functions; no significant difference was observed in adverse events. No firm recommendation can be made due to instability of the summarized data.
rehabilitation
What problem does this paper attempt to address?