Surgery for low back pain

Surgery for low back pain works for some people but not for others, and the evidence is mixed on which patients benefit most.

The retrieved studies focus on specific surgical techniques for specific diagnoses—disc herniation and spinal stenosis—not on low back pain in general. For a herniated disc causing leg pain, discectomy (removing the disc fragment) produces substantial improvement in leg pain and function for most patients, with about 85% reporting good or excellent outcomes at one year in one study of 163 patients [2]. A long-term follow-up of 72 patients over 10–22 years found that while leg pain improved, 74% still had residual low back pain and 12.5% needed a second surgery [3].

For spinal stenosis (narrowing of the spinal canal), decompressive surgery provides lasting improvements in pain and walking ability up to 10 years, with about 68% of patients satisfied at 5–10 years, but roughly a quarter require reoperation [4]. A randomized trial of 79 patients found that a minimally invasive technique (ULBD) reduced hospital stay from about 4 days to 2 days and cut morphine use by 80% compared to open surgery, with similar complication rates [1].

The evidence does not support surgery for chronic low back pain without a clear structural cause like nerve compression. A 2011 evidence synthesis explicitly states that in the absence of deformity or symptomatic neural compression, MRI findings should not determine surgical treatment, and routine MRI for chronic low back pain is not recommended [5]. Fusion surgery for chronic low back pain has variable and weak evidence, with patient selection and comorbidities like smoking and depression modifying outcomes [5].

Complications are real but uncommon. In the endoscopic discectomy study of 163 patients, there were 6 transient motor deficits (3.7%), 6 dural tears (3.7%), and 2 wound infections (1.2%) [2]. The long-term stenosis study reported a 25% reoperation rate over 10 years [4].

My call: For a herniated disc with clear nerve compression causing leg pain, surgery is reasonably effective. For chronic low back pain without nerve compression, the evidence does not support surgery. Confidence: moderate—the evidence is strongest for disc herniation with radiculopathy, weaker for spinal stenosis, and absent for nonspecific low back pain.