Building Executive Dashboards to Support Child Welfare Leaders in Michigan

Project Context: 

• Child welfare leaders in the Michigan Department of Health and Human Services (MDHHS) wanted to use data more effectively to shape decisions that could help make progress towards agency goals of effectively managing system capacity, strengthening the quality of service delivery, and improving child and family outcomes.  
• Senior leaders sought to restructure their existing executive performance dashboards in a way that could better show early indicators of longer-term system outcomes as well as trends over time to track progress being made and know whether and when to intervene early.  

How the GPL Supported: 

• Brainstormed with leaders to identify gaps in existing dashboards where leaders lacked a clear sense of performance and dashboards were not helpful. 
• Organized those performance questions into three key categories — capacity, quality, and outcomes — and carefully selected a high-priority set of 42 key metrics for monitoring performance. Examples include: 
  • Capacity:  Share of staff with 11 or fewer investigations, 12 investigations, 13-14 investigations, 15+ investigations 
  • Quality:  Share of primary caregivers / parents with goal of reunification with face-to-face visit in last 30 days 
  • Outcomes: Among investigations that did not open to CPS, share with subsequent contact to centralized intake / screen-in within 3 months 
• Helped create new visualizations for these key metrics that depicted results over time and incorporated suggested guidance for interpreting trends. 

Results:

• With new dashboards, the leadership team in Michigan was able to more effectively review system performance data and act quickly to address concerns.
• For one example, during a dashboard review session, leaders noticed a significant spike in the rate of abandoned calls to the hotline. They worried people who were hanging up before they could report suspected abuse or neglect might not call back later. Leaders identified several potential steps to adjust staffing levels and reduce call processing times.
• Over the course of eight months as these changes were implemented, the rate of abandoned hotline calls dropped from 35% to near 5%, falling back in line with historical trends.