Statecraft How to Fix Risk Assessment in Child Welfare
Apr 17, 2025
In this engaging discussion, Alex Jutca, who leads analytics at the Allegheny County Department of Human Services, dives into the innovative use of data in child welfare. He highlights the success of the Allegheny Family Screening Tool in improving decision-making while addressing concerns about algorithmic bias. Jutca also explores the complexities of involuntary commitments in mental health and the pressing need for new approaches to tackle addiction issues. The conversation underscores the importance of integrating data to enhance service delivery and care quality.
AI Snips
Chapters
Transcript
Episode notes
Integrated Data Enables Better Services
- Allegheny County DHS integrates many social services under one department, enabling better coordination and data sharing.
- This structure breaks down silos and improves client service across sectors like child welfare, housing, and behavioral health.
AI Supports Child Welfare Screening
- Allegheny created a machine-learning tool called the Allegheny Family Screening Tool to aid child welfare referral decisions.
- The tool predicts risk of future maltreatment and works alongside human judgment to allocate resources efficiently.
Human Judgment Complements Algorithms
- CPS workers take calls and make initial risk assessments while the algorithm generates a risk score independently.
- Staff can override the algorithm, ensuring human insight complements machine predictions.