Acupuncture as a Complex Intervention: A Holistic Model

Objectives: Our understanding of acupuncture and Chinese medicine is limited by a lack of inquiry into the dynamics of the process. We used a longitudinal research design to investigate how the experience, and the effects, of a course of acupuncture evolved over time. Design and outcome measures: This was a longitudinal qualitative study, using a constant comparative method, informed by grounded theory. Each person was interviewed three times over 6 months. Semistructured interviews explored people's experiences of illness and treatment. Across-case and within-case analysis resulted in themes and individual vignettes. Subjects and settings: Eight (8) professional acupuncturists in seven different settings informed their patients about the study. We interviewed a consecutive sample of 23 people with chronic illness, who were having acupuncture for the first time. Results: People described their experience of acupuncture in terms of the acupuncturist's diagnostic and needling skills; the therapeutic relationship; and a new understanding of the body and self as a whole being. All three of these components were imbued with holistic ideology. Treatment effects were perceived as changes in symptoms, changes in energy, and changes in personal and social identity. The vignettes showed the complexity and the individuality of the experience of acupuncture treatment. The process and outcome components were distinct but not divisible, because they were linked by complex connections. The paper depicts these results as a diagrammatic model that illustrates the components and their interconnections and the cyclical reinforcement, both positive and negative, that can occur over time. Conclusions: The holistic model of acupuncture treatment, in which "the whole being greater than the sum of the parts," has implications for service provision and for research trial design. Research trials that evaluate the needling technique, isolated from other aspects of process, will interfere with treatment outcomes. The model requires testing in different service and research settings.