Webinar 01 Recap

AI Improves Outcomes for Youth in Child Welfare & Juvenile Justice Systems

April 9 to 10, 2021

10:30 PM – 12:30 AM EST

Virtual Session

Session Recording


What’s the session all about?


This webinar will introduce participants to the FirstMatch™ model.

FirstMatch™ is a predictive analytics software solution that will match youth with the right treatment program the first time, significantly increasing the chances of a successful outcome. FirstMatch™ is the only tool of its kind that will predict the likelihood of a match between child and treatment program based upon a youth’s specific predictive factors and a provider’s historical clinical and outcomes data. When systems of care use the FirstMatch™ model in tandem with high quality and high fidelity programs, program completion rates, remain out of care rates, recidivism rates, hospitalization rates are improved.

Participants will be introduced to the model, experience a demonstration of a live tool, and learn about how FirstMatch™ is implemented.

Speaker Information

Shawn Peck - Vice President, FirstMatch

Shawn Peck, Vice President

Adelphoi Innovative Solutions

Shawn oversees administration of all FirstMatch services. Shawn previously served as SPEP Project Manager at the Evidence-Based Prevention and Intervention Support (EPIS) Center at Penn State University, and was responsible for engaging probation departments across the state to apply the findings of the Standardized Program Evaluation Protocol (SPEP™), a scoring system used to estimate the impact of juvenile justice programs on reducing recidivism. Shawn holds a Master of Public Administration from Penn State University and a Bachelor of Science in Biblical Studies from East Coast Bible College.

Mark Mortimer, Chief Operation Officer, Adelphoi / FirstMatch

Mark Mortimer, President / Chief Operating Officer


Mark oversees planning, design, and implementation of business operations at FirstMatch. Mark began his career at Adelphoi 20 years ago as a counselor, and subsequently became Supervisor of Adelphoi’s Middle Creek Secure Group Home, which was named PA Juvenile Court Judges’ Commission Program of the Year in 2005. He has also served as Director of Residential Services and as Vice President at Adelphoi. Mark is a graduate of Indiana University of Pennsylvania where he earned a Bachelor of Science in Criminology and an MBA.


FirstMatch is a state of the art admissions software that uses predictive analytics to help decision makers select the right treatment program for a child, increasing likelihood of success.

FirstMatch was developed as a better way to recommend a treatment for a child in the child welfare, juvenile justice or behavioral health systems.  By using data to make treatment decisions, youth spend less time cycling in an out of programs and more time with their families, leading to reduced trauma, and better outcomes.

Right now, youth are matched with a treatment program based upon a combination of factors such as cost, space in the program, proximity to home, past experiences, relationships with referral source, macro level outcomes, contract status, and anecdotal information.

When a child’s clinical information is entered into FirstMatch, the tool uses a provider’s historical outcomes to recommend the best program for that particular child. Counties and providers use the data to make an informed choice.

Currently, 66% of youth in congregate placement have already been in one or more placements and 44% have been in two or more previous placements.  It’s not uncommon for youth to be in 10 or more prior residential placements, with the average being 4.4.  In addition, 88% of these youth have been in one or more previous in-home programs.

This problem is not unique to youth in residential placements. Right now, 64% of youth in in-home programs have been enrolled in at least on other in-home program.

Machine learning  is a type of artificial intelligence that allows software applications to become more accurate at predicting outcomes without being explicitly programmed to do so. Machine learning algorithms use historical data as input to predict new output values.

FirstMatch use machine learning to comparespecificpredictive factors of a youth referral to the historical outcomes achieved by unique programs for other youth that presented with the same or similar predictive factors (and combinations of these various factors).  The program then uses this data to recommend the most appropriate treatment program for the youth.

The tool increases its effectiveness by using an agency’s unique outputs and results to continuouslylearn what to look for in terms of patterns and trends that are undetectable by human analysis.

Once a youth’s data is entered into FirstMatch, the tool makes a program recommendation for the child.  FirstMatch also produces a report that reveals the likelihood that a youth will achieve the desired outcome within a particular program, including recidivism, re-entry, remain-out-of-care and program completion rates, among others.

FirstMatch leads to improved outcomes for systems, organizations, and youth.  So far, we have seen a decrease in negative incidents, higher program completion rates, and cost savings from youth being placed in fewer programs.

Through comparative graphs, the dashboard displays the tool’s predictions to actual outcomes, as well as referral source trends and referral status. The dashboard can be customizable to meet an agency’s specific needs.

FirstMatch offers many benefits to organizations.  Agencies can have confidence that by using data to drive decisions, they are either accepting the right youth into their programs. When kids are in the right program for them, programs have less negative incidents associated with inappropriate placements, reduced staff turnover, and better outcomes for kids and families.  The FirstMatch tool also provides a dashboard that allows for real-time admissions referral management. 

Matching a youth to program based upon data means youth receive the help they need from one right program, rather than cycling through four or five wrong ones.  Less cycling means less trauma, better outcomes, and more success for kids.

By using FirstMatch, systems can make a more informed decision for a child’s treatment, significantly increasing the chance of a successful outcome.  This tool also provides another layer of quality assurance within agencies by helping to identify anychanges in programming that may lead to a decrease in desired outcomes.  FirstMatch also offers transparency with placing agencies in the acceptance and rejection of youth.

The amount of time necessary for implementation varies, depending on an organization’s capacity.  The average implementation takes about 4-6 months to build the tool and about another 8-12 months to train the tool.

FirstMatch is HIPAA compliant and the data is considered to be Protected Health Information (PHI).  Organization’s control access to their data through a cloud based server that they own.

Please visit the website at firstmatch.com. We’ll be happy to provide a demo upon request.  You can also contact us by calling 724-331-1767 or emailing shawn.peck@adelphoi.org.