Research Review by Dr. Shawn Thistle©


Dec. 2006

Study Title:

Surgical vs. nonoperative treatment for lumbar disc herniation: The Spine Patient Outcomes Research Trial (SPORT): Observational cohort


Weinstein JN et al.

Publication Information:

JAMA 2006; 296: 2451-2459.


Many studies have been published comparing surgical versus non-surgical treatments for lumbar disc herniation. These studies have numerous shortcomings which limit the applicability of their results including: small sample sizes, baseline differences between treatment groups, and lack of validated outcome measures to name a few.

This portion of the SPORT project involved patients who declined randomization, and hence formed the observational cohort. 743 patients participated in this study, which had the same 2 year follow-up period as the randomized trial.

Inclusion and exclusion criteria were identical to the randomized trial. It has been suggested that differences may exist between patients willing to undergo randomization and those not willing. This was the reasoning behind adding an observational cohort to this study, and is a positive step towards resolving this potential discrepancy.

All patients in this observational cohort were considered surgical candidates. They had endured persistent symptoms for at least six weeks, and had confirmatory signs of lumbar disc herniation. The outcome measures utilized in this portion of the study were identical to the randomized trial, and included: Medical Outcomes Study 36-item Short-Form Health Survey (SF-36) bodily pain and physical function scales, the American Orthopaedic Surgeons MODEMS version of the Oswestry Disability Index (ODI).

Outcomes were assessed at baseline, 6 weeks, 3 months, 6 months, and 1 and 2 years from enrollment. Secondary outcomes included patient-reported self-improvement, work status, and satisfaction with current symptoms and with care.

In this study, 528 patients received surgery and 191 received usual nonoperative care. For statistical purposes, all improvement noted before surgery was attributed to the nonoperative care and improvement noted after surgery was attributed to the surgical treatment.

Pertinent Rindings:

  • the study population had a mean age of 41.4, with the majority being male, white, employed, and having at least some college education
  • 98% of patients had a classic dermatomal pain radiation pattern
  • patients who chose surgery were younger, heavier, less likely to be working, more likely to be receiving disability compensation, had lower physical function, worse sciatica, reported fewer comorbid joint problems, had more disc extrusions, more often a positive crossed-SLR, and more neurological deficits
  • nonoperative treatments in the usual care group included: education/counseling (92% of patients), anti-inflammatory medicine (58%), narcotic analgesic medication (35%), epidural injections (38%), and nondescript physical therapy (43%)
  • both treatment groups showed improvements at the follow-up times, with a statistically significant advantage showing for surgery at the 3 months mark
  • the benefit of surgery was seen as early as six weeks, and was maintained at the 2 year point
  • the nonoperative group also improved, with nearly equal scores on the main outcome measures by 2 years
  • patient perception that the problem was getting worse was the most common factor that could predict participation in the observational cohort versus consenting to randomization

Conclusions & Practical Application:

This study had similar results to the randomized portion of the SPORT project. The main difference is slightly better surgical outcomes at three months. This could indicate that patients who refused randomization (and in most cases perceived that their condition was worsening) could have been better surgical candidates to begin with.

The authors mention that outcomes from trials involving self-reported outcomes of non-randomized treatments should be interpreted with caution. The reason for this is that patients who choose a particular treatment may be more likely to report a positive outcome (most people don’t like to be 'wrong', do they?).

The take home message from these two trials for those of us who practice manual medicine is twofold. First, there is still a need to high quality trials involving manual therapies that all of us use. I thought the nonoperative care groups in this project received mainly medical management, and some nondescript physical therapy. I think we have a lot to offer patients with radicular pain and this needs to be further studied.

Second, long-term (in this study 2 years) outcomes in those who have surgery and those who do not seem to be equivalent. It is our responsibility to be aware of this information so we can effectively discuss various treatment options for lumbar disc herniation with our patients.