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

Choosing and using the right outcome measures is essential for practice in early childhood intervention, because these tools shape how children's progress and families' experiences are understood and valued. Many measures are available, but information about what they measure, how well they work, and how practical they are to use is scattered across research publications, manuals, and websites. Bringing this information together using traditional review methods can take months or years, making it difficult to respond to current policy and practice needs. We describe a practical approach to mapping evidence on outcome measures within a short time frame, using artificial intelligence (AI) tools alongside expert judgement. The work was done as part of the development of Australia's National Best Practice Framework for Early Childhood Intervention (the Framework), which aims to support better outcomes for children and families. We designed a human-led workflow in which AI tools supported tasks such as searching for and organizing large amounts of relevant information. Experienced researchers reviewed and checked information at key stages to make sure it was accurate, relevant, and clearly documented. This approach allowed us to identify outcome measures, locate evidence about their strengths and limitations, and summarize practical information such as training requirements and accessibility. Using this approach, we reviewed a large body of evidence and developed practice-ready resources for 26 outcome measures in less than 12 weeks. This case study shows that AI can be used responsibly to assist with evidence mapping when it is embedded within a clearly defined process and combined with expert oversight.