Nebulous Controls, Noncompliance Mar JAMA Low Back Pain Study
OrthoBuzz occasionally receives posts from guest bloggers. This guest post comes from Brett A. Freedman, MD, in response to a recent JAMA study on treatment for chronic low back pain.
Chronic low back pain (CLBP) is truly a bio-psycho-social disease. Cherkin et al. in the March 22/29, 2016 issue of JAMA published a randomized clinical trial comparing the performance of two psychologically focused interventions for CLBP with usual care.
The authors randomly assigned 342 subjects solicited from an integrated health plan in the state of Washington who had at least 3 months (average 7.3 years) of nonspecific CLBP to one of three cohorts: mindfulness-based stress reduction (MBSR), cognitive behavioral therapy (CBT), or usual care. MBSR is a pain self-management program that incorporates yoga and focuses on “increasing awareness and acceptance of moment-to-moment experiences.” The two therapy arms included eight 2-hour sessions.
The primary and secondary outcomes were computed values on patient-reported outcome (PRO) instruments compared from baseline out to one year. According to intention-to-treat analysis, MBSR and CBT resulted in a significantly higher chance of patients obtaining a clinically meaningful response, which equated to a >30% improvement in scores on the modified Roland Disability Questionnaire at 26 weeks after enrollment (61% for MBSR vs. 58% for CBT vs. 44% for usual care).
While these findings are interesting and support the notion of more research into non-pharmacological and non-interventional CLBP treatment, the impact of this study is limited by inherent flaws. The investigators’ intent was to have the usual care group represent a control group. However, the usual care cohort was far from controlled. At the time of enrollment, those randomized to the usual care cohort were each given $50 and were set free to “seek whatever treatment [for their CLBP], if any, they desired.” The resultant placebo effect of receiving active treatment (i.e. MBSR, CBT) versus no prescribed treatment (i.e. usual care cohort) is substantial.
Also, aside from reporting their collected PRO data, the authors say little about what happened to the usual care group during this trial, further making this cohort too nebulous to serve as a meaningful comparator. If this cohort is excluded from the analysis, this becomes a negative-findings study, since there were no significant differences in any measured outcome between the MBSR and CBT cohorts, aside from mental health measures, which were significantly improved following CBT.
Another major flaw is the very high rates of patient noncompliance with treatment. Only 51% of the subjects in the 2 therapy arms attended at least 6 of the sessions, and 13 subjects (11%) in each of the active-therapy groups attended no sessions. A substantial minority of patients failed to meaningfully participate in their prescribed intervention, yet their improved outcomes are attributed to the impact of these programs. If the same lack of adherence to protocol occurred in a pharmacological or surgical study, the results would be ignored and the article would likely go unpublished, or at least not published in a high-impact journal such as JAMA.
In conclusion, the greatest merit of this study is the research question it poses. We certainly need more work on this subject, but unfortunately this particular study does little to further advance our understanding of the best practices for approaching the bio-psycho-social disease we call CLBP.
Brett A. Freedman, MD is an orthopaedic surgeon specializing in spine trauma and degenerative spinal diseases at the Mayo Clinic in Rochester, MN.