Algorithmic-Assisted Decision-Making Tools in Child Welfare Practice: A Systematic Review

Purpose: Algorithmic-assisted decision-making tools are increasingly used in child welfare services, yet key factors and challenges for successful and ethical implementation remain underexplored. This review centers on fairness, equity, and ethics in their application. Methods: Using PRISMA guidelines and including gray literature, nine studies were reviewed that described algorithmic tools implementation across various stages of the child welfare system. The review focused on identifying challenges and critical success factors, especially concerning fairness, equity, and ethics. Results: This study used a holistic framework to review critical aspects from tool development to deployment. Additionally, strategies to address fairness and ethical considerations were identified and synthesized. Discussion and Applications to Practice: Algorithmic-assisted tools hold promise for supporting high-stakes decisions in child welfare, but responsible use requires attention to ethical implementation. The review reveals significant methodological and empirical gaps, underscoring the need for future research to ensure equitable and effective deployment in practice.