Guidance on artificial intelligence use for rapid evidence mapping of early childhood intervention outcome measures

AIM: To synthesize research evidence on the psychometric properties of outcome measures alongside practical implementation considerations relevant to children, families, practitioners, and services, and produce evidence-based practice resources to support clinical decision-making in outcome measurement for the early childhood intervention (ECI) context. METHOD: We conducted a mapping review that incorporated rapid review principles and was augmented by generative artificial intelligence tools, including Perplexity, Consensus, and Claude. Outcome measures used in ECI contexts were identified from multiple sources and screened against predefined criteria. An iterative pilot informed the development of an 11-step standard operating procedure. Data were extracted and synthesized using structured templates, with expert verification integrated at key stages. RESULTS: Within a time-limited 11-week project, 128 outcome measures were identified and examined. Of these, 26 met the criteria for full synthesis and practice resource development (child, n = 15; parents, carers, and family, n = 8; service, n = 3). A total of 2201 records were retrieved and 469 were included in the evidence synthesis, with additional measures and literature screened or in progress at project completion. INTERPRETATION: This study demonstrates how a transparent mapping workflow, augmented by artificial intelligence tools, can be implemented to support the identification and synthesis of information on outcome measures in ECI. The methodology offers a feasible approach for addressing time-sensitive evidence needs and may be transferable to other applied contexts.