Changes in healthcare use and cost associated with referrals to social prescribing link workers in England: a matched case-control cohort study.

Background: Social prescribing primarily aims to address patients' non-clinical needs and increase the appropriateness of healthcare use. A major new national programme in England, introduced in 2019, has led to substantial increases in referrals to social prescribing link workers. Impacts on healthcare use and cost have not been examined. We aimed to address this. Methods: This case-control cohort study used anonymised electronic healthcare records from the Clinical Practice Research Datalink (CPRD) Aurum dataset, which contains around 25% of the English population to examine healthcare use and costs up to 1 year post-referral. Calliper matching was used to identify five controls for each case based on age and sex, with one control patient selected within the same practice. We matched 168,699 patients with a referral to social prescribing between July 2019 and April 2022 to 1,302,443 non-referred patients. We used multivariable linear regression models to estimate effects on subsequent healthcare use and costs at three timepoints before referral (2 years, 1 year, 6 months) and two timepoints after referral (6 months, 1 year). Findings: Patients referred to social prescribing had 2.252 more contacts (95% Confidence Interval (CI): 2.089-2.414, 24% relative to mean) with primary and/or secondary care compared to non-referred patients in the first 6 months. This was likely driven by primary care and elective outpatient appointments. By 1 year, this increase was smaller (+0.306, 95% CI: 0.147-0.464) and referred patients had 0.048 (95% CI: -0.084 to -0.012) fewer appointments with general practitioners, compared to non-referred patients. Healthcare costs were <0.1% higher for referred patients (clinically insignificant), due to costs of additional secondary care admissions over both time periods. Interpretation: The initial increase in use of elective healthcare suggests social prescribing may encourage health-seeking behaviour. The reduction in general practitioner appointments within 1 year suggests social prescribing may result in a shift toward more suitable forms of care. Better integration between social prescribing data platforms and electronic healthcare records would enable future research to estimate the effects of uptake, fidelity and completion of social prescribing link worker referrals on healthcare use and costs. Funding: National Institute for Health and Care Research.