Comparative Effectiveness of Treatments for Chronic Low Back Pain: A Multiple Treatment Comparison Analysis.

STUDY DESIGN: A systematic review and network meta-analysis.

OBJECTIVE: To determine current treatment options of chronic low back pain (LBP) as defined by randomized controlled trials (RCTs) and to compare effectiveness of those treatments using a mixed-treatment comparison (MTC).

SUMMARY OF BACKGROUND DATA: It is important to provide an evidence-based assessment of the treatment options that exist for LBP.

METHODS: A systematic search of RCTs was conducted in MEDLINE and the Cochrane Collaboration Library from 1990 to 2014. From the selected studies, we extracted preoperative and postoperative ODI and VAS back pain scores, additional surgeries, and complications. Standard and network meta-analytic techniques were used.

RESULTS: Twelve RCTs were included in the analysis: 5 total disk replacement (TDR) versus fusion; 1 TDR versus exercise and cognitive behavioral therapy (CBT); 5 fusion versus exercise and CBT; and 1 fusion versus physical therapy (PT). On the basis of MTC, with respect to ODI change scores, the pooled mean difference favoring fusion over exercise and CBT was 2.0 points (95% CI, -1.2 to 4.8). The pooled mean difference favoring TDR over exercise and CBT was 6.4 points (95% CI, 3.2-9.3). The pooled mean difference favoring fusion over PT was 8.8 points (95% CI, 4.1-13.6). The pooled mean differences favoring TDR over fusion was 4.4 points (95% CI, 2.37-6.63). For PT versus structured exercise with CBT, the pooled mean difference favoring exercise with CBT over PT was 6.8 points (95% CI, 1.5-12.8). For TDR versus PT, the pooled mean difference favoring TDR over PT was 13.2 points (95% CI, 8.0-18.4). Additional surgery rates were similar between treatment options.

CONCLUSIONS: All 4 treatments provided some benefit to patients with chronic LBP. According to the MTC analysis, TDR may be the most effective treatment and PT the least effective treatment for chronic LBP. This review is based on a limited number of RCT studies and does not support any 1 treatment modality for all patients.