Policy & Practice | Fall 2023

technology speaks

By Jim Daugherty and Julie Barbosa

Innovative Solutions in Illinois: Natural Language Processing Platform to Save Significant Administrative Time

L ike many child welfare organiza tions across the United States, the Illinois Department of Children and Family Services (DCFS) is undertaking a large-scale system modernization to streamline operations and better support the important work of their child welfare teams. Encouraged by the directive from DCFS Director Marc D. Smith to “embrace bold thinking,” both tech and agency leadership committed to finding innovative solutions to transform the state’s child welfare tech nology in a way that would positively impact agency staff, the agency overall, and most important, family outcomes. Accessing the unstructured data in case notes became key to achieving that goal. Agency leadership knew that case workers spend too much time combing through narrative data, reading typed or handwritten case notes, to under stand the story of the family and think through potential courses of action. DCFS leadership knew that the ability to read, summarize, and present the information in the notes would save child welfare teams valuable time and give them a complete view of the family’s story, such as their risks and strengths, their family support struc ture, and the social determinants of health that affect the family’s ability to thrive. Also, the time traditionally spent clicking through and reviewing case notes—note after note—could be better spent in any number of ways, such as spending more time with families. The DCFS chose Natural Language Processing (NLP) platform Augintel to

or animals in the home—information contained in case notes. Augintel is being rolled out to nearly all staff functions and offices across DCFS as well as the private agencies who support the child welfare system in the state to ensure a complete view of the case across the life cycle of that case. This puts Illinois DCFS on the leading edge of human services agencies using NLP to uncover infor mation buried in unstructured data. Roll-out began in February 2023 and occurred over a six-month period, and the early results are very encouraging. Uptake has been good and reports from regular users are impressive.

solve the unstructured data problem. Augintel quickly searches through electronic case information and finds pertinent case notes, data, key words, and family members, eliminating the need for caseworkers to mine the data themselves. The agency benefits from the ability to query case notes across all cases to identify best practices and trends across the agency or to monitor for early warning signs of issues to be addressed. DCFS leadership also recognized the ability to mine data across cases offered an opportunity to help keep caseworkers safe by pin pointing cases that might pose a safety risk to caseworkers due to weapons

Image via Shutterstock

Policy & Practice Fall 2023 38