Study: Groundbreaking Tool is Keeping Kids Safer From Abuse
Allegheny County (PA) program being adopted by other child protection agencies
Imagine this scenario:
You're a child abuse hotline worker. A woman calls to complain that her neighbor's kids have been standing at the school bus stop in sub-freezing weather with no coats on all week. Is it child abuse? Maybe the kids just forgot their coats. Or maybe their mother is a strung-out addict who spent the winter-coat-money -- and the money for food, heat, rent and everything else -- on drugs.
Allegheny County's groundbreaking Allegheny Family Screening Tool (AFST) is supposed to help those call-takers make better decisions on whether that call -- or any other -- needs more investigation. And the first results-based study of AFST shows it's working .
- Researchers from Stanford University found the tool increased the accuracy of identifying children who ended up needing some kind of intervention , without increasing the workload on investigators. That's important because, let's face it, while they want to protect kids, the county doesn't have an unlimited number of caseworkers. They want to send those workers on high-risk cases and avoid taking their time investigating low-risk ones.
- Researchers also found that AFST helped even out racial disparities . Before, more black families than would be expected were targeted for abuse investigations. The study says under the AFST, there have been increases in cases opened for white children, and declines in the rate at which black families were screened-in for investigation.
Fewer black families are being "screened-in" for child abuse investigation.
Department of Human Services Deputy Director Erin Dalton says it's all about doing the best job protecting kids. "We're most pleased that the accuracy of decisions has improved especially for kids who go on to need further services," she said.
AFST was designed by Allegheny County in 2016. It uses something called a mathematical "algorithm" to combine all of the information the county has on a child and family to predict the level of risk in a situation reported to the county's hotline.
It looks at 100 other variables, like previous abuse, housing instability, addiction and mental health problems and criminal records in the family. Instead of the call-taker having to manually look up all that information, the AFST grabs it, crunches it, and assigns the case a number from 1-20, one being low risk, and 20 being high risk, which would prompt the call-taker to recommend an immediate home visit.
"(Call-takers) have immediate access to our databases," explained Dalton. "The worker can weigh the current crisis against that information. They can make decisions and predict the likelihood of longer term adverse effects," if no action is taken.
"We are encouraged that the AFST has shown positive results by increasing accuracy, while preserving clinical judgement." Marc Cherna, DHS Director DHS
In case you think a computer alone is making decisions on the safety of children, Dalton says that couldn't be further from the truth. "The call takers are using the score to make decisions but they don't feel bound by it. Let's imagine the score is very low but there's something in the person's voice, say a grandma who is concerned about a new boyfriend in the home we don't know about." In that case Dalton says, a call-taker would likely recommend additional investigation. "(AFST) is not intended to take away that kind of discretion."
Trained hotline call-takers make the final call on how an abuse complaint is handled.
AFST has already been revised and is in its second "model" which will also be evaluated by the Stanford researchers. It has proven to be so effective so far that other child welfare agencies across the country are looking at adopting it. Click here to read more about it .