More than 1,600 children in the US died from child abuse and neglect in 2015, according to the US Administration for Children and Families. One of those children was named Nora, who died at just 4 months old. Although her case included many risk factors for potential abuse -- siblings who had active cases of abuse in the system already and several adult family members who had suffered abuse as children -- investigators only learned of those risk factors when they were looking into Nora's death. The information was not surfaced for case workers who could have intervened if they had known.
Fixing this kind of problem is why SAS chose to create SAS Analytics for Child Safety as an early proof of concept for what SAS Viya and SAS Visual Investigator can do. Steve Bennett, director of the Global Government Practice at SAS (the sponsor of this site), provided a demo of this implementation at the SAS Global Forum last month and spoke to AllAnalytics.com about the solution.
I saw the demo last month, and it's really a powerful example of how organizations can make data useful for end users -- in this case it is child welfare case workers. The integration and analytics are all under the hood, and the information is delivered to the case workers -- or any non-technical worker -- in a way that's visual and easy to understand. It also makes it easy for these users to navigate through the information.
That's so important for child welfare case workers, who normally have bigger caseloads than they can handle. They may be doing the best they can to prioritize their case load, but typically they have limited visibility into the information that could reveal that a child is at risk and potentially save that child's life.
Case workers can be much more effective if they have risk scoring and alerts at their fingertips and can drill down into why certain children are getting higher risk scores. For instance, adults living at the same address who have been abused as children themselves.
That's something that SAS delievered with this use case of Visual Investigator in North Carolina.
Visual Investigator provided these case workers with more than just greater visibility into factors affecting a particular case. It connected information from multiple data sources at multiple agencies. Using the insights available through this integration it was able to provide a much more accurate risk scoring than was previously available.
For instance, in Nora's case, it could have surfaced the fact that someone who had recently been released from prison listed an address that was the same as Nora's address.
Case workers may have tried to gather this type of information themselves by calling around to various different agencies and leaving voice mail messages and hoping someone would call back with needed information. But that's a time consuming process, and the clock is ticking for children at risk.
The SAS Analytics for Child Safety system creates a "plain English" description of the risk and the factors contributing to the risk score. Case managers are provided with visualizations of potential risk through timelines and other visual representations. For instance, a triangle represents a potential event. Purple triangles are alleged perpetrators, and triangles with exclamation points are events that have been verified, Bennett told me.
Case workers can click on the events for more information. They are also shown a list of all the people associated with the child at risk and can drill down on those names for more information. All the information about Nora, her siblings, the adults in her life, and their histories would be available to these case workers on a single screen. Plus, these case workers would receive alerts for children scoring high on the at-risk scale, providing the information needed to potentially avert a tragedy.
The Child Safety system is one of many proof-of-concept solutions SAS is creating with industry around its new technologies, Bennett told me.
"We wanted to know the places where we could find investigative tasks where managers have to synthesize a lot of information," Bennett told me. Other use cases could include prescription drug monitoring for opioid abuse, insider threats in computer networks, and law enforcement applications.
"The exciting thing about Visual Investigator is that it is geared toward the operational end use," Bennett said. "It's meant to take all that great risk scoring, all that sits under the hood, all the world-class analytics of SAS that is baked in but the investigator doesn't see. It's a different approach."