The rapid growth of social prescribing in England: analyses of primary care medical records using data from the Clinical Practice Research Datalink (CPRD)

BackgroundSocial prescribing (SP) is growing rapidly in England and across the world. However, whom it is reaching and how effectively it is being implemented remains unclear. This study aimed to assess longitudinal trends in SP in Englands primary care system, including growth trajectories and target alignment, sociodemographic profiles of referred patients, and predictors of service refusal over time. MethodsThis study analysed primary care records from 1.2 million patients from 1,736 practices in the Clinical Practice Research Datalink in England. We estimated SP trends between 2019 and 2023 using growth curve modelling on SP numbers at practice level. Descriptive analyses were used to show changes in sociodemographic profiles of SP patients over time. To assess sociodemographic disparities in service refusal (defined as having a medical code of social prescribing declined), we used multilevel logistic regression models stratified by year, accounting for nested data structure where patients were nested within practices. FindingsAs of the end of 2023, an estimated 9.4m GP consultations in England have involved SP codes, and 5.5m consultations have specifically led to SP referrals. In 2023, females constituted 60% of SP patients and ethnic minority groups represented 23%. Representation from patients living in more deprived areas increased from 23% to 42% between 2017-2023. Service refusal declined from 22% to 12% between 2019-2023. Age, sex and ethnicity were associated with service refusal across multiple years. In 2023, notably, all age groups had higher odds of refusal compared to the youngest age group. Females had 21% lower odds of refusal than males (95% CI=0.77-0.82), and patients from white ethnic backgrounds had 32% higher odds of refusal than ethnic minority patients (95% CI=1.26-1.39). InterpretationSP has expanded rapidly in England, far exceeding initial targets of 900,000 patients by 2023/24 and suggesting broad service acceptability. Progress is being made in reaching certain target groups such as more deprived communities. However, there are still disparities in accessibility and uptake, calling for targeted strategies to address underlying inequalities. FundingMQ Transforming Mental Health, Rosetrees-Stoneygate Trust Fellowship, National Academy for Social Prescribing Research in contextO_ST_ABSEvidence before this studyC_ST_ABSWe systematically searched PubMed, PsycINFO, Cochrane Library, Web of Science, and OpenGrey for studies (including grey literature) published in English between January 1980 and March 2025, using search terms such as social prescribing, non-medical referral/intervention, non-clinical referral/intervention, community referral, referral scheme. Prior evidence on rates and patterns of referrals was limited to small-scale evaluations, cross-sectional data, or regional analyses, with a lack of nationally representative longitudinal studies examining implementation trends or equity. Systematic reviews highlighted gaps in understanding disparities in service access and uptake. While the National Health Service (NHS) Long Term Plan (2019) set ambitious targets for SP, no studies had quantified progress toward these goals since the national rollout. Added value of this studyUsing primary care medical records from the Clinical Practice Research Datalink (CPRD), this study provides the first longitudinal, national analysis of SP implementation across 1.2 million patients in England. We demonstrate that SP referrals exceeded NHS targets by 27-51% in 2023, with at least 1.1-1.4 million patients receiving referrals. The analysis reveals significant progress in reaching deprived populations (representation increased from 23% to 42% between 2017-2023) but identifies persistent disparities in service uptake across age, sex, and deprivation groups. Implications of all the available evidenceThe rapid expansion of SP reflects its growing integration into primary care. However, persistent sociodemographic disparities highlight the need for targeted interventions to ensure equitable service access and uptake. This study provides policymakers with evidence to standardise referral protocols and allocate resources to underserved areas. Future research should rigorously track the implementation of SP, evaluate its long-term health outcomes and cost-effectiveness to fulfil its potentials as a key component of universal personalised care.