The common measurement outcome employed by four randomized trials was the Oswestry Disability Index (ODI), a validated measure specific for lumbar degenerative disorders. This index is measured from 1 to 100 with higher scores indicating a higher level of disability. Developers of the ODI indicate that a clinically relevant change is 4 points, whereas other studies have suggested thresholds of up to 18 points are required for clinical relevance.1

All four of the randomized trials compared surgical treatment of mechanical LBP with nonoperative treatment. Of the four randomized trials reviewed, three of them had structured nonoperative regimens.2–4 The Fritzell et al. 2002 study did not have a structured regimen of physical therapy. Instead it used any kind of physical therapy as the main component, which could be supplemented with “information and education, TENS (transcutaneous electrical nerve stimulation), acupuncture, injections, cognitive and functional training, and coping strategies.”5

All four studies showed a similar improvement in the surgical arm of patients, with improvement from baseline ranging from 8.9 to 15.6 points (percent improvement 18.9 to 37.1%).6 In the nonoperative arm, improvements ranged from 2.8 to 12.8 (percent improvements from baseline were 5.8 to 30.1%). Only in the Fritzell et al. study was the improvement in nonoperative treatment below the clinically relevant threshold of 4 points on the ODI,5 whereas in the other three studies the improvement seen in the nonoperative patients was similar to that of the operative patients. These three studies used a structured nonoperative treatment regimen incorporating cognitive behavioral therapy, whereas the Fritzell et al. study did not.

The greatest improvement in surgical patients, when compared to their nonsurgical counterparts, was seen in the Fritzell et al. study (ΔODI surgery group − ΔODI nonsurgical group) at 8.8.5 The Fairbank et al. study showed improvement in the surgical group of patients as well, with a minor improvement at 4.1 (95% confidence interval [CI]: 0.1–8.1). This value was not considered statistically significant.2 Neither Brox et al. study showed a statistically significant difference between surgical and nonsurgical intervention for the ODI. The first study, looking at patients without prior surgery, the improvement seen with surgery for this study was 2.3 (95% CI: −6.8 to 11.4).3 In the second Brox et al. study, looking at patients with a prior discectomy, greater improvements in the ODI were seen with nonsurgical treatment. When adjusted for gender and treatment expectations, this value was −9.7 (95% CI: −21.7 to 1.7).4 Mannion et al. in their study consisting of three multicenter RCTs of surgery versus multidisciplinary cognitive‐behavioral and exercise rehabilitation found no difference in patient self‐ reported outcomes between fusion and multidisciplinary cognitive‐behavioral and exercise rehabilitation for chronic LBP.7