Explaining How Cognitive Behavioural Approaches Work for Low Back Pain: Mediation Analysis of the Back Skills Training Trial.
Fordham B., Ji C., Hansen Z., Lall R., Lamb SE.
STUDY DESIGN: .: This is secondary research examining the longitudinal mediation effect within a structural equation model. OBJECTIVE: .: To identify possible mechanisms which mediate the effects of a cognitive behavioural approach upon disability and pain in low back pain patients. SUMMARY OF BACKGROUND DATA: .: Cognitive behavioural interventions (CBIs) can improve pain and disability in low back pain (LBP) but the mechanisms of action are unclear. We used data from a large randomised controlled trial to investigate mediators of the treatment effect of CBI. METHODS: .: Pain self-efficacy, fear-avoidance, physical and mental functioning were selected as candidate mediators based on the theoretical rationale of the intervention. The primary treatment outcomes were the Roland Morris Questionnaire (RMDQ) and the modified-Von Korff scale (MVK pain and disability) at 12-months. We used structural equation models to estimate the contribution of mediators. All models were tested for goodness-of-fit using chi-square, Root Mean Square Error of Approximation, Adjusted Goodness of Fit Index and Bentler Comparative Fit Index. RESULTS: .: We included 701 adults with LBP. The average RMDQ score at baseline for those on the intervention arm was 8.8 (Standard Deviation 5.0). The intervention was effective in reducing disability and pain at 12-months. Change in mental functioning was not a significant mediator. Changes to pain self-efficacy, fear-avoidance and physical functioning were causal mediators of the treatment effect at 12-months (RMDQ b= -0.149, p < 0.001; MVK-pain b = -0.181, p < 0.001 and MVK-disability b = -0.180, p < 0.001). Overall, the SEM model exceeded the threshold for acceptable goodness-of-fit. CONCLUSIONS: .: Fear-avoidance and self-efficacy were important causal mediators of the cognitive behavioural treatment effect. Self-assessed change in physical function was a causal mediator but mental functioning was not. This suggests people need to experience meaningful change in physical function and beliefs but not in mental functioning associated with LBP, to achieve a treatment benefit. LEVEL OF EVIDENCE: 2.