Aim: A meta-analysis of randomised controlled trials was performed to investigate the effectiveness of surgical fusion for chronic low back pain compared to non-surgical intervention.

Methods: Several electronic databases (MEDLINE, EMBASE, CINAHL and Science Citation Index) were searched from 1966 to October 2005. Two authors independently extracted data. The meta-analysis comparison was based on mean difference in Oswestry disability index (ODI) change from baseline to follow up of patients undergoing surgical versus non-surgical treatment.

Results: Four studies of 58 articles identified in the search were eligible with a total of 740 patients. One of the studies recruited patients with adult isthmic spondylolisthesis, whereas the other studies recruited patients with a history of chronic low back pain of at least 1 year duration. Surgical treatment involved pos-terolateral fusion with or without instrumentation and flexible stabilisation. Non-surgical treatment involved exercise programs with or without cognitive therapy. The follow-up period ranged from 1 to 2 years. The mean overall difference in ODI between the surgical and non-surgical groups was statistically in favour of surgery (mean difference of ODI: 3.90; 95% confidence interval: 0.17–7.62; p=0.04; I²=21.4%). Surgical treatment was associated with a 13% pooled rate of early complications (95% confidence interval: 6–20%).

Conclusion: Surgical fusion for chronic low back pain favoured an improvement in the ODI compared to non-surgical intervention; this difference in ODI is of minimal clinical importance. Furthermore, surgery is associated with a significant risk of complications. Therefore, the cumulative evidence at present does not support routine surgical fusion for the treatment of chronic low back pain.