Social and material deprivation and the cost-effectiveness of an intervention to promote physical activity: cohort study and Markov model

BACKGROUND We developed a method to model the cost-effectiveness at different levels of deprivation of an intervention to promote physical activity. METHODS The cost-effectiveness of a brief intervention in primary care was estimated by means of a Markov model stratified by deprivation quintile. Estimates for disease incidence, mortality, depression prevalence and health service utilization were obtained from 282 887 participants in the UK Clinical Practice Research Datalink with linked deprivation scores. Discounted results were compared for least deprived and most deprived quintiles. RESULTS An effective intervention to promote physical activity continuing for 5 years gave an increase in life years free from disease: least deprived 54.9 (95% interval 17.5-93.5) per 1000 participants entering model; most deprived 74.5 (22.8-128.0) per 1000. The overall incremental quality adjusted life years were: least deprived, 3.7 per 1000 and most deprived, 6.1 per 1000 with probability cost-effective at £30 000 per QALY being 52.5 and 63.3%, respectively. When the intervention was modelled to be 30% less effective in the most deprived than the least deprived quintile, the probability cost-effective was least deprived 52.9% and most deprived 55.9%. CONCLUSION Physical activity interventions may generate greater health benefits in deprived populations. When intervention effectiveness is attenuated in deprived groups, cost-effectiveness may sometimes still be similar to that in the most affluent groups. Even with favourable assumptions, evidence was insufficient to support wider use of presently available brief primary care interventions in a universal strategy for primary prevention.