Inequalities and Inclusion in Exercise Referral Schemes: A Mixed-Method Multi-Scheme Analysis

Physical activity prescription, commonly through exercise referral schemes, is an established disease prevention and management pathway. There is considerable heterogeneity in terms of uptake, adherence, and outcomes, but because within-scheme analyses dominate previous research, there is limited contextual understanding of this variance. Both the impact of schemes on health inequalities and best practices for inclusion of at-risk groups are unclear. To address this, we modelled secondary data from the multi-scheme National Referral Database, comprising 23,782 individuals across 14 referral schemes, using a multilevel Bayesian inference approach. Scheme-level local demographics identified over-sampling in uptake; on the basis of uptake and completion data, more inclusive schemes ( n = 4) were identified. Scheme coordinators were interviewed, and data were analyzed using a grounded theory approach. Inequalities presented in a nuanced way. Schemes showed promise for engaging populations at greater risk of poor health (e.g., those from more deprived areas or of an ethnic minority background). However, the completion odds were lower for those with a range of complex circumstances (e.g., a mental health-related referral). We identified creative best practices for widening access (e.g., partnership building), maintaining engagement (e.g., workforce diversity), and tailoring support, but recommend changes to wider operational contexts to ensure such approaches are viable.