Characteristic and incidental (placebo) effects in complex interventions such as acupuncture

This landmark article by Charlotte Paterson and Paul Dieppe, published in the BMJ in 2005, represents a fundamental contribution to the debate on research methodology in acupuncture and other complex interventions. The work critically questions the applicability of the traditional randomized placebo-controlled trial model, originally developed for testing pharmaceuticals, when applied to complex therapies such as acupuncture. The authors based their analysis on a comprehensive qualitative research program involving 88 interviews with patients and 11 with professional acupuncturists. This qualitative methodology enabled deep understanding of patients' real-world experiences and practitioners' perspectives on how acupuncture actually works in clinical practice.

The study identifies three fundamental assumptions that underpin traditional controlled trials but may not adequately apply to complex interventions. First, that diagnosis occurs before the intervention begins. In acupuncture, however, diagnosis according to traditional Chinese medicine is an emergent and continuous process that develops throughout each treatment session, integrating intimately with the therapy. Second, that incidental factors (placebo effects) are generic and independent of the specific therapeutic theory.

The researchers found that elements such as specialized listening, therapeutic conversations, and the differentiated diagnostic process are specific characteristics of acupuncture, not merely generic placebo effects. Third, that characteristic and incidental effects are distinct and additive. In acupuncture, these elements are intimately interwoven and cannot be easily separated. The findings have profound implications for study design.

When controlled trials with sham acupuncture attempt to isolate only the needling effect, they may inadvertently and dramatically underestimate the intervention's full therapeutic effect. This occurs because other important characteristic elements — such as the unique diagnostic process and the specific therapeutic interactions — are provided to both groups, masking the true differences between real and sham treatment. The authors argue that this can explain recurrent paradoxes in the acupuncture literature, such as the discrepancy between its long history of clinical use and the lack of demonstrated efficacy in traditional randomized trials, as well as the fact that both real and sham acupuncture frequently show good effects. The study suggests that the choice of research methodology should be guided by the therapeutic theory underlying the intervention being tested.

For complex interventions such as acupuncture, pragmatic designs that evaluate the effectiveness of the complete intervention may be more appropriate and rigorous than attempts to isolate specific components. This work has significantly influenced the field of acupuncture and complementary medicine research, promoting important discussions about research methodologies appropriate for different types of health interventions.