The short-term and long-term cost-effectiveness of a pedometer-based intervention in primary care: a within trial analysis and beyond-trial modelling

Abstract

Background There is little evidence of the cost-effectiveness of pedometer-based interventions. We examined the short-term and long-term cost-effectiveness of a pedometer-based walking intervention in inactive adults. Methods Data were collected as part of a three-arm cluster-randomised trial conducted (2012–14) in seven primary care practices in London to assess the effectiveness of pedometer-based walking interventions (PACE-UP trial). Eligible participants were inactive adults aged 45–75 years, without contraindications to increasing moderate-tovigorous physical activity. 11 015 people were mailed an invitation. 6399 were non-responders, and 548 individuals who self-reported as being active were excluded. 1023 people were randomised to usual care (control, n=338), postal pedometer (339), and nurse-supported pedometer interventions (346). 956 participants (93%) provided outcome data. Intervention groups received pedometers, 12 week walking programmes, and diaries to record physical activity. The nurse group was also offered three physical activity consultations. A within trial cost-effectiveness analysis was done at 1 year. Additionally, a Markov model, using routine data obtained via reviews of epidemiological and economic literature, was used to extend trial results to a life-time horizon. Cost per change in physical activity (step count, and moderate-to-vigorous physical activity in ≥10 minute bouts) and quality-adjusted life-years (QALYs) for interventions were assessed. Costs (in 2013 prices) are presented from a health-care provider’s perspective and uncertainty as a costeff-ectiveness acceptability curve. Ethics approval was provided by London Research Ethics Committee (Hampstead). PACE-UP is registered with Current Controlled Trials, ISRCTN98538934. Findings In the short term, incremental cost per step and cost per min in a 10 min or more bout of moderate-to-vigorous physical activity were £0·19 and £3·61, respectively, for nurse-support. The postal group took a greater number of steps and cost less. In the long term, the postal group dominated both control and nurse groups in that QALY gains (759, 95% CI 400 to 1247) added to increased cost savings (–£11 million, 95% CI –12 to –10), resulting in an incremental net monetary benefit of £26 million (based on a hypothetical cohort of 100 000 people). The postal group had a 50% chance of being cost-effective in terms of QALYs at 1 year and, at a £20 000 per QALY threshold, robustly dominated both nurse and control groups in the long term. Interpretation A pedometer-based intervention delivered by post, compared with current practice, would deliver cost savings in the short term and the most quality of life benefits in the long term.