Predicting Risk

Predictive analytics is a current hot topic.  In simplest terms, it’s a method for predicting the future based on past data.  Predictive analytics is used in insurance, banking, marketing, telecommunications, travel, healthcare, pharmaceuticals, and, increasingly, social services as large amounts of data are becoming more available.

The most challenging use of predictive analytics is predicting risk of child abuse and neglect.

Over 1500 children die each year from child abuse and neglect.   About 70% are children under the age of 3 with almost three-quarters of the deaths resulting from neglect.  Identified risk factors include substance abuse by parents and prior contact with child welfare system. More details are available in the report Child Maltreatment 2014.  The report also includes a helpful chart showing the disposition of the 3.6 million annual referrals for child abuse and neglect.

U.S. Department of Health & Human Services, Administration for Children and Families, Administration on Children, Youth and Families, Children’s Bureau. (2016). Child maltreatment 2014. Available from http://www.acf.hhs.gov/programs/cb/research-data-technology/statistics-research/child-maltreatment

The Chronicle of Social Change is a top source of information on predictive analytics in the child welfare system.  The Chronicle has reported on the following tools that are used to predict risk with child abuse and neglect cases:

In Florida, Eckerd’s Rapid Safety Feedback (ERSF) tool was developed as a result of child deaths in open child welfare cases.  Cases often involved small children and substance abuse. Analytics now identify a high risk case which is reviewed by a person not assigned to the case.  Critical protective factors include information sharing across systems, supervisory reviews, safety plans, and quality of contacts with families.  There have been no child deaths since implementation three years ago while there were 9 deaths in the three years before.  Use of the tool is being expanded to 5 other states.

Los Angeles County has the largest child welfare system in the world with 220,000 complaints annually.  The county contracted with SAS to develop risk modeling. Dubbed AURA, or Approach to Understanding Risk Assessment, the model tracked child deaths, near fatalities, and “critical incidents” in 2011 and 2012.  Risk scores were assigned from 1-1,000.  AURA could identify child deaths and near child deaths that had occurred but also identified false positives in which there were no subsequent critical events.  Use of the model has become very controversial.  In addition to false positives, opponents also base objections on concerns of profiling and disproportionality.

Allegheny County, PA, has one of the most effective child welfare systems in the nation (see http://www.nfpn.org/articles/monograph-an-effective-child-welfare-system) and is also one of the most advanced in terms of predictive analytics.  The county integrates data from other systems along with the child welfare system   based on a tool developed in New Zealand.  The findings are not the sole determinant of risk but one of the tools used by social workers in decision-making.

For more information on the development and use of predictive analytics in the child welfare system, see a sampling of articles at: https://chronicleofsocialchange.org/?s=predictive+analytics.

Over on the Preserving Families Blog the conversation is about safety.  Visit: https://preservingfamiliesblog.wordpress.com/

NFPN’s online courses for in-home services will include training on risk and safety.  The three courses are scheduled for March 10, 17, and 24 at 10:30 a.m. Central Time. The cost is $50 per course or all three courses for $100. You can register by contacting NFPN’s Executive Director, Priscilla Martens, director@nfpn.org. One prerequisite is current use (or purchase) of the NCFAS-G or NCFAS-G+R assessment tool. Participants will receive a Certificate of Completion.  

Priscilla Martens, NFPN Executive Director

 

Leave a Comment

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: