Definitely one to follow and also a nice description of mediation analysis:
Understanding how pain education causes changes in pain and disability: Protocol for a causal mediation analysis of the PREVENT trial.
“Pain education (also referred to as pain biology education, pain physiology education, Explain Pain, and pain neuroscience education) is an educational approach that was developed to decrease pain and disability associated with LBP. It is proposed that these effects are achieved by reducing a patient’s conviction that pain is an accurate signal of the extent of tissue damage, and by increasing the conviction that pain is influenced by one’s beliefs and thoughts; and indeed, ‘any credible evidence of danger to body tissue’. The educational material covers the physiology of pain and incorporates biological, physical, cognitive, and social factors to explain how pain is experienced after the integration and evaluation of all these factors.
A two-arm, participant and outcome assessor blinded, NHMRC-funded randomised controlled trial (RCT) (PREVENT) will be conducted to investigate the effectiveness of a pain education intervention for preventing chronic symptoms in patients with an acute episode of LBP. The primary analysis of this study will estimate the effect of pain education on pain intensity 3 months after the onset of LBP.
Complex interventions, such as pain education, are usually evaluated by their effects on outcomes such as pain or disability. However, these interventions do not directly target these outcomes; they target intermediate factors that are presumed to be associated with the outcome. This implies that the effect of an intervention on an outcome is often not a straightforward cause-effect relationship, and hypothesised mechanisms that underlie an intervention need to be evaluated. A method for quantifying the intermediate effects of interventions is mediation analysis.”
Publishing protocols is also just good science