Randomised controlled trials in the evaluation of non-biomedical therapeutic interventions for pain: The gold standard?

Abstract

This paper offers a critical review of important issues in the implementation and interpretation of randomised controlled trials (RCTs) used to evaluate non-biomedical therapeutic interventions for pain. The analysis of methodological issues focuses on the validity, relevance and clinical importance of outcome measures; sampling and randomization bias; blinding procedures used to control for the placebo response and therapist influences, and treatment standardisation. A key issue is that pain management is inextricably linked to the establishment of a therapeutic relationship. In such circumstances, it is argued that the 'gold standard' double-blind RCT is impracticable. Those relying on RCTs as sources of evidence need to be aware that an 'unblinded' RCT can enhance placebo responses to the intervention and introduce important sources of bias. The challenge faced by healthcare researchers is to identify alternative research designs able to provide valid and reliable evidence that the proposed therapeutic intervention is capable of achieving cost-effective, clinically important and personally relevant outcomes in naturalistic settings.