The FAMCare Blog

Predictive Risk Modeling

Posted by George Ritacco on Apr 19, 2018 9:00:00 AM

predictive modeling

In a recent issue of Social Work Today, Kate Jackson reported this troubling statistic:

“Each day, 4-8 children die from abuse or neglect at the hands of their parents or caretakers…Social workers and other child welfare workers are largely tasked with the prevention of these tragedies…They staff child protection hotlines all across the country and struggle to field calls alerting them to children at risk of neglect or abuse and to determine whether suspicions or allegations warrant further investigation. They must determine quickly and accurately, and typically based on limited information, whether the assertions are valid—which calls can be screened out, requiring no further action, and which must be screened in, generating an investigation. In every instance, they're aware that the cost of an error could be a child's life.”

Marc Cherna, MSW, director of the Allegheny County Department of Human Services adds these insights:

"What happens in call screening in child protection services is that you get an allegation and you have to do your due diligence on that call—speaking to collaterals, consulting with [the] source of the referral, determining whether the family is known in your system, and creating a risk score that influences whether you're going to give the case to an investigator or screen it out."

A high score doesn't automatically indicate that a child needs to be removed from a home, but it demands an investigation, he says.

This process can be both time consuming and inefficient. Caseworkers must cull information from disparate databases that may or may not be predictive. Children at risk can easily be overlooked, and the time it takes to access and analyze the data may render the decision to investigate mute.

Advances in Technology

There is a new technological application being tested in child welfare that can help. Predictive risk modeling (PRM) is an emerging technology that utilizes the foundation logic of an artificial intelligence application. PRM draws on vast data mines to identify patterns of family characteristics or behaviors and their associated outcomes then compares the information about an individual or a family to patterns revealed in historical data to help predict certain outcomes. From the model that emerges, child welfare workers might better infer which referrals require investigation and which do not. Kate Jackson says,

 “In addition to potentially increasing the accuracy of the child worker’s decision-making processes, predictive analytics tools can reduce the time it takes to determine the need for investigation – a potentially lifesaving combination.”

Cautions in Child Care

Although PRM tools have been employed for decades in insurance, finance, and medicine, for the following reasons their use in child welfare has evolved cautiously:

  • There are concerns about the ways sensitive personal data might be used.
  • PRM scores are not intended to overrule human judgment.
  • PRM could be used disproportionally in a manner that could heighten racial bias and stigmatize or marginalize parents of color.
  • Where is the consent point when using background family information to formulate the risk score?
  • What’s to prevent false positives?
  • Where does professional competence enter the picture? In other words, will high scores intimidate other judgment factors?
  • The field of PRM for social services currently lacks agreed “best practice” principles for the development, implementation, and governance of PRM tools.
  • Social workers tend to be reluctant to use technology to improve practice because they fear anything they think will impede or limit their discretion.

The Adoption of Analytic Tools in Child Welfare

Rhema Vaithianathan, a PhD professor of economics and co-director of the Center for Social Data Analytics at Auckland University doesn’t see the widespread adoption of PRM that’s occurred in other fields.

“Child welfare is an area where we cannot afford to get things wrong, so we are naturally cautious. However, comprehensive guidelines around PRM development, implementation, and governance will make a big difference to this process in the future, removing the need for every project to ‘reinvent the wheel’ and reducing the risk of failure.”

At Global Vision Technologies, our R&D team has been actively working through our library of assessments, driving data through our KPI analysis tools to start to determine trends that could possibly lead to predictors and future modeling of cases and types.  Predictive modeling and risk stratification are significant tools to have as a caseworker.  This is an important project and we will be sharing more information in the future.  Stay tuned.

Topics: Child Welfare, predictive modeling, risk stratification