A Randomised Controlled Trial of Acupuncture Care for Persistent Low Back Pain: Cost Effectiveness Analysis

This randomized controlled trial published in the BMJ in 2006 represents a robust economic analysis of acupuncture for persistent low back pain, a condition that significantly affects quality of life and generates substantial costs for health systems. Low back pain has a prevalence of 40% in the English adult population, with estimated annual costs of £480 million for the NHS. The research was conducted in York, England, involving three private acupuncture clinics and 18 general medical practices. Participants were 241 adults aged 18-65 years with persistent nonspecific low back pain of 4-52 weeks' duration, considered suitable for primary care management.

The design followed 2:1 randomization, with 160 patients allocated to receive up to 10 sessions of traditional Chinese acupuncture over three months, delivered by qualified acupuncturists, while 81 patients received only usual care. Follow-up extended for 24 months to capture long-term effects. The cost-effectiveness analysis used the British NHS perspective as the primary analysis, with a secondary societal analysis including productivity costs. Costs were measured in 2002-3 pound sterling, and effectiveness was measured through quality-adjusted life-years (QALYs) derived from the SF-36 questionnaire converted to SF-6D.

Results showed that the acupuncture group had mean total costs of £460 compared with £345 in the control group during the two years, a difference of £115. Acupuncture treatment accounted for approximately 45% of the total NHS costs in the intervention group. In terms of effectiveness, both groups had improved quality of life, but the acupuncture group showed an incremental gain of 0.027 QALYs at 24 months. The incremental cost-effectiveness ratio was calculated at £4,241 per QALY gained, a figure considered highly favorable.

Probabilistic sensitivity analysis demonstrated more than 90% probability of acupuncture being cost-effective considering the £20,000 per QALY threshold established by NICE. When societal costs were included, especially productivity gains from reduced time off work, acupuncture became dominant (less costly and more effective). The clinical implications are significant, suggesting that commissioners of musculoskeletal services would be justified in making acupuncture available to patients with persistent low back pain in primary care. The study provides solid economic evidence complementing the clinical findings, where a modest effect was observed at 12 months but a more robust benefit at 24 months.

The pragmatic nature of the study allows generalization to real clinical practice, although usual care varied among participants. Limitations include reliance on self-reported questionnaires for resource use, which can introduce recall bias, and a considerable proportion of missing data (35% for the base-case analysis), although characteristics of those excluded were similar to those included. Sensitivity analysis with multiple imputation confirmed the robustness of the results. This work establishes an important precedent for economic evaluations of acupuncture and provides a scientific basis for evidence-based health policy decisions.