Question: In patients with intervertebral disc herniation and
persistent symptoms, is surgery more effective than nonoperative
treatment?
Design: Randomized (unclear allocation concealment), unblinded,
controlled trial with 2-year follow-up (Spine Patient Outcomes Research Trial
[SPORT]).
Setting: 13 spine practices in 11 U.S. states.
Patients: 501 patients who were =18 years of age and had
intervertebral disc herniation and persistent symptoms despite some form of
nonoperative treatment for =6 weeks. Inclusion criteria were radicular pain
and evidence of nerve-root irritation with a positive nerve-root tension sign
or a corresponding neurologic deficit, and candidacy for surgery with evidence
of disc herniation (protrusion, extrusion, or sequestered fragment) on
advanced vertebral imaging. Exclusion criteria included previous lumbar
surgery, cauda equina syndrome, scoliosis >15°, segmental instability,
vertebral fractures, spine infection or tumor, inflammatory
spondyloarthropathy, pregnancy, comorbid conditions contraindicating surgery,
or inability or unwillingness to have surgery within 6 months. 472 patients
(mean age, 42 y; 59% men) had =1 follow-up visit. 1-year follow-up was 83%;
2-year follow-up was 74%.
Intervention: Patients were allocated to operative (n = 245) or
nonoperative treatment (n = 256). Operative treatment involved standard open
discectomy under general or local anesthesia, with patients in the prone or
knee-chest position. The interlaminar space was entered through a midline
incision reflecting the paraspinous muscles. The fragment of disc was removed
through a small anular incision. The canal was inspected and the foramen
probed for residual disc or osseous pathology, and the nerve root was
decompressed. Nonoperative treatment ("usual care") included at
least active physical therapy, education and/or counseling with home exercise
instruction, and nonsteroidal anti-inflammatory drugs if tolerated.
Nonoperative treatment was individualized to the patient.
Main outcome measures: Change from baseline on the Medical Outcomes
Study 36-Item Short-Form Health Survey (SF-36) bodily pain and physical
function scales (range, 0 [more severe] to 100 [less severe]) and the American
Academy of Orthopaedic Surgeons MODEMS version of the Oswestry Disability
Index (ODI) (range, 0 [less severe] to 100 [more severe]). Secondary outcomes
included symptom severity according to the Sciatica Bothersomeness Index (SBI)
(range, 0 [not bothersome] to 24 [extremely bothersome]) and patient-reported
improvement. With an originally determined sample size of 250 patients in each
group, the study had 85% power to detect a 10-point difference in the SF-36
bodily pain and physical function scales or a similar effect in the ODI.
Main results: Analysis was by intention to treat. 50% of patients in
the operative treatment group declined or delayed surgery and 30% of patients
in the nonoperative treatment group received surgery within 3 months of
enrollment. 1-year outcomes are presented in the table on this page (2-year
follow-up was < 80%). The difference between groups in SF-36 measures and
ODI favored surgery but did not reach significance (Table). Greater
improvements in the SBI were seen in the operative treatment group than the
nonoperative treatment group (Table). Among patient-reported measures, there
was no difference between groups except in self-rated progress, which, at the
time of the 1-year follow-up, showed greater improvement with surgery (76% vs
67%, 95% confidence interval for the 9% difference 0.3 to 18).
Conclusion: In patients with intervertebral disc herniation and
persistent symptoms, patients improved after either operative or nonoperative
treatment.
The Spine Patient Outcomes Research Trial (SPORT) marks a milestone for
evidence-based spine surgery. In this trial, patients showed marked
improvement in symptoms and function with either operative or nonoperative
treatment, but nearly 40% of the operative group and 45% of the nonoperative
group switched groups, diluting any potential real differences in treatment
efficacy. With such high crossover rates, analysis of the results according to
assigned treatment can seem
counterintuitive1.
It is important, however, to remember that randomization equalizes all of the
unknown factors that may be associated with outcome. The factors that lead
patients to deviate from the randomly assigned treatment may be associated
with outcome2. If
study results were analyzed according to administered treatment, we could no
longer assume that these unknown factors were equally distributed in the 2
treatment groups; the magic of randomization would be lost.
The SPORT reflects real-world spine care. It evaluates "use
effectiveness"3
or "programmatic
effectiveness"4
of surgery, as opposed to "method effectiveness" or
"biologic
efficacy,"5
which would require sham
surgery6. The SPORT
evaluates the effectiveness of surgery as the package of care in the context
of routine clinical
practice4. Symptoms
from disc herniation can change, and as symptoms change, patients may rethink
their treatment assignment and opt out. In this dynamic process, results at 3,
6, 12, and 24 months were similar whether patients were initially assigned to
surgery or not. Surgery was relatively safe, with fewer than 5% of patients
having complications. Cauda equina syndrome did not develop in any of the
patients in either treatment group.
The fact that patients enrolled in SPORT exercised their right to change
treatment attests to successful conveyance of equipoise and informed consent
by the research team. It shows respect for the autonomy of patients even in
their role as research subjects. The SPORT has yielded valuable information.
Both the research subjects and the research team can be proud of their
contribution to the betterment of others.
Lee YJ, Ellenberg JH, Hirtz DG, Nelson
KB. Analysis of clinical trials by treatment actually received: is it really
an option? Stat Med.
1991;10:
1595-605.101595
1991
[PubMed][CrossRef]
Heitjan DF. Causal inference in a
clinical trial: a comparative example. Control Clin Trials.
1999;20:
309-18.20309
1999
[PubMed][CrossRef]
Meier P. Comment on "Compliance as
an explanatory variable in clinical trials" by B. Efron and D. Feldman.
J Am Stat Assoc. 1991;86:
19-22.8619
1991
[CrossRef]
Wright CC, Sim J. Intention-to-treat
approach to data from randomized controlled trials: a sensitivity analysis.
J Clin Epidemiol. 2003;56:
833-42.56833
2003
[PubMed][CrossRef]
Sommer A, Zeger SL. On estimating
efficacy from clinical trials. Stat Med.
1991;10:
45-52.1045
1991
[PubMed][CrossRef]
Flum DR. Interpreting surgical trials
with subjective outcomes: avoiding UnSPORTsmanlike conduct.
JAMA. 2006;296:
2483-5.2962483
2006
[PubMed][CrossRef]