For CareSource to provide the best outcomes for our members, we rely on accurate historical and situational data that we receive throughout their care. Data can help predict possible health obstacles before they arise, enabling us to address them proactively and provide better health care. In the case of expectant mothers, good data can even guide life-altering care decisions.
CareSource currently uses a number of third‐party high‐risk pregnancy tools like notifications from state health agencies to flag high‐risk pregnancies. But by using our own data, we have a better opportunity to study how we identify and prioritize high risk pregnancies to design effective targeted outreach programs.
In collaboration with CareSource Care Management leadership, our Predictive Analytics & Data Science team has developed a machine learning algorithm that looks for patterns in high‐risk pregnancies within our membership. This helps identify clinical, behavioral and social drivers that enable the prediction of potential adverse outcomes for current pregnancies.
Our model leverages historical and current clinical factors and is enhanced by demographics, behavioral attributes and social determinants of health to predict risk of premature birth, neonatal abstinence syndrome indicator, failure to thrive, low birth weight, still birth, NICU and sick newborns. By helping to identify challenges as early as the first trimester of pregnancy, we’re using our internal data resources to improve the odds for better lives for babies and their families.
“The opportunity to leverage predictive analytics allows for a more targeted approach to identifying members at risk. This improved model is important to my team from a workflow process, but what really matters is that this will save lives and improve care.”
– Shannon Steele, CareSource Vice President, Care Management