Screening: inclusion and exclusion criteria

The inclusion and exclusion criteria were developed using PICOC and SPICE frameworks (25) to ensure selection of relevant studies in the search. Studies meeting the following criteria were included: Participants/Population (P): articles with study populations of any age, from all countries and from those consisting of patients/service users and healthcare practitioners; Intervention (I): (i) specified referral routes in order to meet the definition of ‘arts on prescription’; (ii) participants were referred to community arts activities or interventions delivered by artists or other facilitators (e.g., museum education officers); (iii) Interventions were group arts activities, but all arts disciplines were included (for example visual arts, literary arts and performing arts); Comparison (C): Studies with and without comparison groups were included; Outcome (O): Reporting measures of impact (use of a validated tool to measure mental health symptoms, wellbeing, mental health, physical health outcomes, social isolation and/or loneliness) and/or articles with qualitative accounts of patients/service users and healthcare practitioners’ experiences; Context (C): (i) Community based or primary care-based studies; (ii) publications from 1994 (this corresponds to the first reported Arts on Prescription programme in the literature); (iii) studies written in English; Evaluation (E): include empirical data (quantitative, qualitative, or mixed methods studies).

Studies meeting the following criteria were excluded: Participants (P): No exclusion criteria; Intervention (I): (i) reporting on the expressive arts therapies (art, music, drama, dance) delivered as a psychotherapeutic intervention; (ii) reporting on community arts programmes without a referral process; Comparison (C): No exclusion criteria; Outcomes (O): No exclusion criteria; Context (C): based in inpatient/hospital-based and residential care home settings; Evaluation (E): case reports; opinion pieces and editorials, review papers and essays.

The protocol for screening was developed in March and April 2023, piloting tools and methods for sharing and saving data (e.g., Zotero and Mendeley). Literature search results were downloaded and shared in an Excel file, with a record of the screening process kept for each database. All titles and abstracts for each database search were saved on a separate Excel sheet, and the screening outcome for each study was recorded (whether included or excluded and a category for the reason for exclusion). All articles were screened for inclusion and exclusion criteria by the authors (AJ, HB, NH, SK) between July and September 2023, screening both abstracts and titles, and, where required, full texts of articles. All full text articles were available to authors. All study titles and abstracts identified by the searches were screened independently for inclusion in the review by two researchers using the study inclusion criteria. For studies that met the inclusion criteria, full text articles were independently screened and assessed for eligibility by the same two researchers. A third reviewer resolved any discrepancies. The final selection of studies was assessed and approved by all authors.