The conclusion of the Weinstein et al. article is that surgical decompression offers greater improvement in pain and physical function than nonsurgical therapy. Have the authors, however, identified those individuals who are good surgical candidates and have they demonstrated a “good” outcome for these spinal stenosis sufferers? The study had 654 individuals: 400 surgical and 254 nonsurgical participants. The individuals eligible to participate in this trial had neurogenic claudication for a minimum of three months and radiographic evidence of spinal stenosis at one or more levels without instability; they were also judged to be surgical candidates although this criterion was never fully defined.
The authors describe the group who underwent surgery as younger, working, with more pain, lower level of function, psychological distress, more severe stenosis, and with worsening symptoms. A slim majority of participants had symptom duration of more than six months, but the scope of the persistence of radicular pain was not mentioned. An incongruity arises when matching duration of symptoms with the radiographic grading of stenotic severity. A majority of individuals are described with evidence of severe stenosis at one or more levels. I would have expected these individuals to have been symptomatic for a longer period of time than six months.
The most frequent comorbidity in these individuals with an average age of 65 was the presence of other joint disease. The impact of this finding on physical function is not mentioned. Nonsurgical therapy was left to the discretion of the practitioner who was most likely an orthopedic surgeon. The nonsurgical candidates may not have been individuals who failed medical therapy, but those who never received a full complement of educational, pharmaceutical, physical, and injection interventions.
The study design included two groups: 1) a randomized cohort assigned to one therapy (n=289) or 2) an observational cohort who self-selected therapy (n=365). A major complicating factor of this study was that 33% of surgical patients who chose medical therapy, and 43% of nonsurgical patients who underwent laminectomy. The authors completed both an intention-to-treat and as-treated analysis because of the frequent crossovers. A partial summary is included in Table 1 (below).
At two years, in the intention-to-treat analysis, surgical patients had a significant improvement in pain but not in physical function or disability. In the as-treated analysis, however, the superiority of surgical decompression over nonsurgical therapy was shown in pain and physical improvement. The as-treated group included individuals who chose the therapy that they thought would help them the most in the setting of their pain severity and physical dysfunction. I believe that the general applicability of the conclusions of the study is diluted both by the very specific clinical characteristics of the study subjects and the absence of randomization for a significant proportion of patients. The Cochrane Review group will probably say that more randomized studies are needed, despite the addition of this clinical trial to the medical literature.