Background: Lumbar disk surgery is performed disproportionately in different regions of the United States, and its effectiveness compared with nonoperative management for patients with herniated disks is unclear. The Spine Patient Outcomes Research Trial (SPORT) followed two cohorts—one randomized and one observational—to compare outcomes for surgical and nonoperative management of herniated disks and other nerve-compressing conditions.

The Study: Patients who had received non-operative treatment for disk herniation but whose symptoms had persisted for at least six weeks were enrolled. Eligible participants also had to have physical examination evidence of nerve-root impingement (e.g., a positive result on a straight-leg raise) and advanced vertebral imaging showing a herniated disk.

Participants were randomized to receive surgery or nonoperative management. Those in the surgery arm had diskectomy with release of the affected nerve root. The usual-care arm involved a recommended minimum of physical therapy, exercise counseling, and nonsteroidal anti-inflammatory drugs. However, patients could receive other modalities, such as epidural injection, chiropractic therapy, or opioids.

Outcome measures were performance on the Medical Outcomes Study 36-Item Short-Form Health Survey bodily pain and function scales and on the American Academy of Orthopaedic Surgeons' version of a disability index. Patients were followed for two years and surveyed at baseline, six weeks, three and six months, and one and two years. Secondary outcomes included patients' accounts of their improvement, work status, and satisfaction, and a measure of symptom severity.

Results: A total of 501 patients were randomized; 472 (94 percent) completed at least one follow-up and were included in the analysis. The mean age was 42 years. During the study, some patients from the surgical arm opted for nonsurgical treatment, and vice versa.

Both groups improved at each follow-up, with a treatment effect favoring surgery. There were no statistically significant differences in the primary outcomes. In the secondary outcomes, there were no statistically significant differences in patient satisfaction or employment status. On the Sciatica Bothersomeness Index, a treatment effect favored surgery, and there was a small, statistically significant difference favoring surgery in patients' perception of their progress. When crossover treatment was taken into account with an as-treated rather than an intention-to-treat analysis, the results strongly favored surgery at all follow-up times.

Conclusion: Patients with persistent symptoms attributable to lumbar disk herniation improved substantially whether they received nonoperative or surgical treatment. In the intent-to-treat analysis, there was no statistically significant difference between the two groups in the global primary outcomes measures, but there were some differences in secondary outcomes favoring surgery. However, an as-treated analysis showed stronger improvement from surgery, suggesting that the study methods led to an underestimation of the surgical benefits.

Many problems limit the interpretation of the study results, including the fact that 43 percent of the nonoperative group crossed over to surgery. In addition, the choice of nonoperative treatments was determined by the patient and the treating physician, and a large proportion of patients in the non-operative group had interventions such as epidural injections.

The authors indicate that many of their findings point to greater improvements from surgery; however, based on the intent-to-treat analysis, no conclusions about the best approach can be made.