Abstract
BackgroundSocial prescribing (SP) is increasingly being adopted internationally to address non-medical, health-related social needs through referral to community-based supports and services. Although economic arguments have been central to the policy development of SP, the health economic evidence base remains fragmented. We systematically reviewed peer-reviewed and grey literature on the economic impacts of SP globally.MethodsFollowing our PROSPERO registered record (CRD420251235365) and PRISMA and ISPOR guidance, we searched twelve bibliographic databases and 34 grey literature sources from inception to March 30, 2026. Studies of any design reporting economic outcomes or economic evaluations of SP were eligible. Risk of bias was assessed using design-specific appraisal tools. Owing to methodological heterogeneity, findings were narratively synthesised.FindingsSeventy-three sources met inclusion criteria (39 peer-reviewed and 34 grey literature publications), yielding 76 studies and 83 economic analyses. Most studies were conducted in the UK (67 [88%]) and involved adults (74 [97%]). Analyses comprised 17 partial economic evaluations, 15 full economic evaluations, 47 return-on-investment evaluations, and four decision-modelling studies. Overall, 61 (77%) analyses reported economically favourable outcomes, six (8%) reported mixed findings, and 12 (15%) reported unfavourable outcomes. SP delivery costs were generally low, with annual savings ranging from £78·20-£670 per participant. Cost-per-quality-adjusted life-year estimates at 12 months ranged from £2137-£12111 and were generally below accepted willingness-to-pay thresholds. Social return on investment estimates were typically positive, ranging from £0·96-£7·08 returned per £1 invested over 12 months. Variation in estimated value was driven primarily by analytical perspective and outcomes monetised rather than intervention type, follow-up duration, or study quality.InterpretationThe economic evidence base for SP is substantially larger than previously recognised but narrow healthcare-demand models underestimate its wider cross-sectoral value. Future evaluations must broaden valuation boundaries and develop financing models that reflect how costs and returns are distributed across sectors; absence of demonstrated sector-specific savings should not be interpreted as absence of value.