Patients | Comparison | Number of trials (n); follow-up | Findings† |
Non-radicular back pain with common degenerative changes | Fusion v non-surgical therapy | 4 (767); 1–2 years | Fusion was better than non-intensive supervised physical therapy for pain and function (1 RCT) but not clinically (1 RCT) or statistically (2 RCTs) better than intensive rehabilitation with cognitive-behavioural therapy. |
Fusion v artificial disc replacement | 2 (596); 2 years | CHARITÉ artificial disc was non-inferior to anterior lumbar interbody fusion (1 RCT), and Prodisc II was better than instrumented circumferential fusion (1 lower-quality RCT) for composite outcomes; results for individual outcomes were inconsistent. | |
Radiculopathy with prolapsed lumbar disc | Discectomy v non-surgical therapy | 4 (968); 2–10 years | Open discectomy was better at {1 year}‡ but not 4 or 10 years (1 lower-quality RCT); microdiscectomy was better at 8 weeks (1 RCT) but not 2 years (2 RCTs); open discectomy or microdiscectomy was better for function and disability at 3 months but not 2 years (1 RCT). |
Symptomatic spinal stenosis | Laminectomy v non-surgical therapy | 4 (718); 2–10 years | Laminectomy was better for some pain measures up to 2 years (2 RCTs) and at 4 years but not 1 or 10 years (1 RCT); treatments did not differ in 1 RCT. |
Interspinous spacer v non-surgical therapy | 2 (275); 2 years | Interspinous spacer was better for overall success at 2 years (2 RCTs, 1 lower-quality). Results for subsequent laminectomy were inconsistent (6% v 22% in 1 lower-quality RCT, 12% v 12% in 1 high-quality RCT). |
*RCT, randomised controlled trial.
†Trials were rated high quality (> 4 of 9 or > 5 of 10 criteria) unless otherwise noted.
‡Data confirmed by author.