data science

Teams at NYU and University of Minnesota to Identify Child Maltreatment Causes Using Causal Data Science

NYU Grossman, Silver, and McSilver, with University of Minnesota, received a $7.6M P50 Grant for the NICHD-funded CHAMP Center

A new, federally-funded research center collaboration between New York University’s Grossman School of Medicine, Silver School of Social Work, McSilver Institute for Poverty Policy and Research, along with the University of Minnesota, are investigating the use of causal modeling to identify the factors that lead to child maltreatment and its consequences—such as depression, suicidal behavior, or posttraumatic stress—and to identify targeted interventions that can reduce children’s risk for these devastating outcomes.

The Center on Causal Data Science for Child and Adolescent Maltreatment Prevention (CHAMP Center) has received a five-year, $7.6 million P50 grant from the Eunice Kennedy Shriver National Institute of Child Health & Human Development (NICHD) at the National Institutes of Health. The CHAMP Center is now the fourth project in the nation to join NICHD’s CAPSTONE Centers for Multidisciplinary Research in Child Abuse and Neglect, which are funded to find solutions to ongoing public health concerns caused by the well-documented impact of childhood trauma. It is the first and only center to use causal data science in its approach.

Under the leadership of the project’s principal investigator, Glenn Saxe, MD, NYU Grossman will oversee the collaborative science required for the CHAMP Center to achieve its ambitious goals.

“In 2019, over 650,000 cases of maltreatment were substantiated in the U.S., but this is the tip of the iceberg of children who are maltreated, as so many cases go unreported,” said Saxe. “The CHAMP Center will advance the scientific knowledge required to help professionals to better know how to reduce a child’s risk and will offer them tools to implement personalized interventions to address and prevent harm.”

Causal data science differs from other data analytic approaches by its methodological ability to search through data on the many thousands of factors that might be associated with a child’s risk, to identify the specific ones that can be shown to be the direct cause. Since interventions can only work if they are directed to change an underlying cause, causal data science sees the first step in reducing a child’s risk as knowing its true root origin. Causal data science also aims to offer the additional, crucial benefit of reducing biases related to factors such as race, culture, and income level that can occur in research on child maltreatment. While the threat of bias is always present, using causal data science methods can reduce biases by drilling down to only factors that directly influence the child’s risk.

Sisi Ma
Sisi Ma, PhD

The CHAMP Center involves dedicated collaboration between teams at NYU and the University of Minnesota. The NYU Grossman team, led by Saxe, provides overall leadership and oversight of the center so that its discoveries are translated into the knowledge and tools that will result in a large-scale impact on the considerable problem of child maltreatment. The team at University of Minnesota’s Institute for Health Informatics, led by Sisi Ma, PhD, and Constantin Aliferis, PhD, will apply advanced data science methods to build and test models revealing the children who are most at risk and the causes that most contribute to those risks. A multidisciplinary team across institutions will transform these models into tools to support clinicians caring for at-risk children and families. These tools will then be tested at NYU Langone Brooklyn’s Family Health Centers in a clinical trial led by Kate Sullivan, PhD, of the NYU Silver School.

“The CHAMP Center offers an incredible opportunity to bring cutting-edge data science techniques to bear on some of the most pressing problems facing children and families and the systems that serve them,” said Sullivan. “It is particularly exciting that translation into actual clinical practice is already embedded in this effort, consonant with the priorities of NYU Silver.”

The CHAMP Center will solve the hard problem that Ma identifies by considering the discovery of causes using state-of-the-art machine learning techniques as only the first step in its work. Then, a broad range of stakeholders must be involved in transforming CHAMP Center discoveries into tools that can be used with children and families to reduce their risks. This effort will be led by Andrew Cleek, PsyD, deputy executive director of the NYU McSilver Institute, who will lead the center’s stakeholder engagement process, consulting with key participants to ensure the knowledge and tools produced by the center are safe, trustworthy, and practical. Additionally, Cleek’s team is responsible for disseminating the scientific knowledge and tools produced by the CHAMP Center so they may achieve their widest reach and greatest impact.

“The CHAMP Center provides a unique opportunity to bring a substantial level of inquiry into an array of issues that will lead to the real-world improvement in the lives of children and their caregivers. The McSilver Institute is thrilled and honored to be part of this remarkable collaboration with our colleagues at NYU and the University of Minnesota,” NYU McSilver’s Cleek said.

Other News

Nurse supporting patient
The Minnesota Geriatrics ECHO has evolved to serve a diverse mix of clinical and operational leadership from Minnesota’s aging services…
Christina Daniel-McKee, Ana Garza, Jamie Riskedahl, Shawn Savolainen, Dan Wilson, Fatuma Youb
Six community leaders from rural Minnesota have been selected as the 2024-2025 cohort for Project REACH.
Dr. Weaver teaching a Stop the Bleed Station
Lauren Weaver, MD, is M Simulation’s newest surgical simulation fellow.