Using Secondary Data to Find and Track Vulnerable Mothers’ in the Hospital Haystack

• In Event: 2-011 - Interactive Poster Session and Reception

Abstract

Each year almost 4 million babies are born in the United States, making it difficult to identify struggling mothers who are eligible for government funded intervention programs aimed at improving maternal-child health. As psychologists, we are in a unique position to harness the power of secondary data and statistical modeling to answer developmental questions and create tools that help find vulnerable families. We partnered with a government agency to develop an algorithm that can identify and connect vulnerable mothers to free home-based intervention programs. These services are designed to enhance positive parenting and ensure that children meet key developmental milestones.

Sociodemographic data was collected from mothers when admitted to the hospital and hospital staff interviewed mothers to assess their eligibility for services. By linking these data, our algorithm is able to predict, with 30% accuracy, which mothers are eligible for services from sociodemographic data alone. Ideally we would link hospital data to data collected by home-based intervention programs and track the outcomes of families after program completion. This would allow us to both assess if mothers identified by our algorithm finish the intervention and ask developmental questions such as, do these children have better heath in early childhood or perform better in school?

We are interested in learning how hospital, government agency, and school data could be linked. Our ultimate goal is to create a data pipeline that will allow academics to work with key stake-holders to design smarter more effective interventions to improve developmental outcomes in at-risk families.

Authors

• Ashley A Ricker, Chapman University
  Presenting Author

• Jennifer Hahn-Holbrook, Chapman University

• Ilia Rolon, CFCOC

• Margarita McCullough, CFCOC

• Christina Luarca, HASC