Research Review By Dr. Kent Stuber©

Date Posted:

Feb. 2009

Study Title:

Surgical versus nonsurgical therapy for lumbar spinal stenosis


Weinstein JN et al.

Author's Affiliations:

Investigators for the Spine Patient Outcomes Research Trial (SPORT), Dartmouth Institute for Health Policy and Clinical Practice, Department of Orthopedics, Dartmouth Medical School, New Hampshire, USA.

Publication Information:

New England Journal of Medicine 2008; 358(8): 794-810.

Background Information:

Spinal stenosis is a relatively common condition – it is the most common reason for lumbar spinal surgery in patients over the age of 65 years. Optimal management strategies have still not been determined. There is limited evidence comparing surgical and non-surgical interventions for lumbar spinal stenosis to date.

A previous study (1) on stenosis patients provided some interesting background information: the average age of the 363 spinal stenosis patients was 64, 40% were female, 20% had college degrees, and 8% were receiving Worker’s Compensation. The most common co-morbidities in this group were joint disease, hypertension, heart disease, digestive problems, diabetes, and osteoporosis. 33% of these patients had their symptoms for greater than one year.

The average Oswestry Disability Index (ODI) scores were 42.3 (equivalent to severe disability due to pain). 33% had received chiropractic treatment, 7% had been to the Emergency Room, 29% had used opiate medications, and 19% had used antidepressants and muscle relaxants. Overall, these patients are older, heavy users of the medical system, and in relatively severe pain.

The aim of this study was to compare the efficacy of surgical versus non-surgical therapy for spinal stenosis as part of the SPORT Trial (Spine Patient Outcomes Research Trial). This study consisted of two components: a randomized cohort study and an observational cohort study (for those who declined randomization), and included patients who were surgical candidates with at least 12 weeks of pain due to spinal stenosis.

Pertinent Results:

654 patients enrolled in the study (289 in the randomized cohort, 365 in the observational cohort). The cohorts were similar at baseline, with only negligible, unimportant differences. The average age of the patients was 65 years, 80% had neurogenic claudication, 79% had dermatomal pain radiation, 91% had stenosis at L4 or L5, while 61% had multi-level stenosis (note: patients did NOT have spondylolisthesis as confirmed via imaging). Throughout the 2 year study period, the following findings were noted:
  • 42% of the non-surgical RCT group underwent surgery at 1 year, 43% at 2 years
  • 22% of the non-surgical group from the observational cohort underwent surgery within 2 years
  • those who crossed over to receive surgery generally had severe symptoms, and a preference for surgery
  • at 2 years, the results of the intention-to-treat analysis favored surgery to a significant degree for bodily pain on the SF-36, but not to a significant degree on the ODI or in terms of physical function on the SF-36
  • the as-treated analysis (for both cohorts combined with adjustments made for confounding baseline factors) favored surgery significantly with respect to improvements in pain, function, satisfaction, and self-perceived progress - the treatment effect was noted as early as 6 weeks, reached a maximum at 6 months, and lasted for the full two years
  • there were moderate improvements in the non-surgical group during the 2 year study period
  • there was a pattern of non-adherence in the surgical group as well as delays or refusal of surgery - 9% of surgical patients experienced dural tear, and re-operation was required in 8% of surgical patients
  • 13 patients died over the course of the study due to non-study related factors

Clinical Application & Conclusions:

The authors concluded that patients who underwent surgery had significantly more improvement in all primary outcomes (pain, function, satisfaction, subjective progress) than non-surgical patients. While this is a logical conclusion based on the results of this study, the authors do not temper this conclusion with a statement along the lines that a trial of conservative therapy (chiropractic, physiotherapy, etc) is warranted before surgical consultation.

In general, studies comparing surgical vs. non-surgical treatments for lumbar spinal stenosis have favored surgery, but many of those authors have still maintained that a non-surgical trial of care is desirable instead of going straight to surgery. The present authors do not make such a statement. Certainly there are studies in the literature that have shown very good results using non-surgical treatments for stenosis, including one study using manual therapy (2). Even in the present study, moderate improvements were noted for the non-surgical group. The ideal non-surgical means of managing lumbar spinal stenosis patients have not been identified and further research into this issue is necessary.

It is prudent for clinicians to inform stenosis patients of the state of the literature. It may be advisable for patients to seek a surgical consultation while undergoing initial conservative therapy in an attempt to ameliorate symptoms. The conservative treatment regimen employed in this study helped produce improvements in this study group and may be a useful starting point for therapy for lumbar spinal stenosis patients.

Study Methods:

This study was part of the SPORT project (Spine Patient Outcomes Research Trial – see below for links to 2 additional reviews on other SPORT publications). SPORT took place at 13 centers in 11 American states including 8 academic centres and 5 private practices, with 115 participating physicians across 6 medical specialties and 2 surgical disciplines. It is the largest cooperative multi-centre, multi-disciplinary study ever undertaken for spinal stenosis. Enrollment took place over a 5 year period. SPORT consisted of two components:
  1. a randomized cohort study
  2. an observational cohort study (for those who declined randomization)
Treatments were either surgical decompression or nonsurgical care. The surgery employed was a standard posterior decompressive laminectomy. Nonsurgical care included active physical therapy, patient education, home exercises, and NSAIDs (if tolerated). Outcome measures included SF-36 scores and modified Oswestry Disability Index scores at 6 weeks, 3 months, 6 months, 24 months, and 2 years. Additional outcomes included subjective improvement and satisfaction, and the level of bother from the stenosis and back pain. Inclusion criteria consisted of patients with neurogenic claudication or radicular symptoms for a minimum of 12 weeks with confirmatory imaging (CT or MRI) showing stenosis at a minimum of one level. Nonsurgical care before enrollment was unspecified. T-tests were used to compare groups with significance set at 0.05 and 85% power levels.

Comparisons within and between cohorts were made. Predictors of crossover into surgical groups were determined via stepwise proportional-hazards regression modeling. The randomized cohort was analyzed initially on an intention-to-treat basis, but crossovers made as-treated analysis necessary. This was done separately for both the randomized and observational cohorts.

Study Strengths / Weaknesses:

From an ethics standpoint this study is extremely strong as subjects could opt to be in either the randomized cohort where they did not choose their treatment or they could be in the observational cohort where they could select their treatment. Crossover from non-surgical to surgical groups in both cohorts was allowed. Ethical approval was obtained in each centre where the study was conducted.

Another significant strength of this study was the number of subjects, reflecting the sheer scope of the project.

This study represents the largest trial comparing surgical and non-surgical intervention for lumbar spinal stenosis. It was conducted in multiple centres including multiple different types of practice (academic centres, private practice, different specialties involved). The size and varying centres makes the results from this study much more conducive to generalization to others with lumbar spinal stenosis.

There were two significant weaknesses to this study. The first was that the forms of non-surgical care were not set strongly; patients in these groups could receive a variety of different therapies. If a stronger non-surgical program was laid out, that would allow more exact comparison of surgery with the program in place. As it stands, it is hard to say which of the non-surgical interventions (NSAIDs, exercise, physical therapy [again broadly defined], or patient education) were more successful and which were less successful. Another weakness to the study was the high degree of non-adherence to treatment groups, particularly by those in the randomized cohort. This reduced the power and ability of the intention-to-treat analysis to show treatment effects.

Additional References:

  1. Cummins J, Lurie JD, Tosteson TD, Hanscom B, Abdu WA, Birkmeyer NJO, Herkowitz H, Weinstein J. Descriptive epidemiology and prior healthcare utilization of patients in the spine patient outcomes research trial’s (SPORT) three observational cohorts. Spine 2006; 31(7): 806-814.
  2. Murphy DR, Hurwitz EL, Gregory AA, Clary R. A non-surgical approach to the management of lumbar spinal stenosis: a prospective observational cohort study. BMC Musculoskeletal Disorders 2006; 7:16: doi: 10.1186/1471-2474-7-16.