Predictive Modeling Lowers Pre-Term Delivery Rates, Increases Prenatal Visits

The Challenge:

A high-risk pregnancy involves increased health risks for the mother, the baby, or both. Several factors contribute to what makes a pregnancy high risk. Certain health conditions, like high blood pressure or diabetes, age of the mother, and previous complicated births can make a pregnancy high risk, for instance.

In the U.S., race and ethnicity also contribute to high-risk pregnancies. Women of racial and ethnic minorities and their newborns are more likely to experience morbidity or mortality. In New York, Black and Hispanic women have higher rates of severe maternal morbidity (life-threatening complications during or after childbirth) than white women — 2.3 times and 1.7 times higher, respectively. And Black infants face a much higher rate of infant mortality than white infants nationwide.

Beyond health conditions and race/ethnicity, inability to access transportation to and from appointments or healthy food can also exacerbate a complicated pregnancy.

Early identification of a high-risk pregnancy, close monitoring and personalized care before and after birth can help avoid maternal and infant morbidity and mortality.

Our Goal:

To identify high-risk pregnancies earlier in order to monitor the pregnant person more closely to reduce maternal morbidity and mortality and the number of infants experiencing complications after birth.

The Healthfirst Solution:

The Healthfirst Cares Maternity Program deployed predictive models, developed by Healthfirst, to identify members at high risk of a complicated pregnancy and birth earlier in their pregnancy.

Highest risk members are referred to a care manager, who works with the member throughout pregnancy and after birth. The care managers worked to improve overall maternal-fetal health outcomes by providing active care management, encouraging adherence to prenatal visits, and increasing referrals to nurse home visit programs for assistance after birth. Care managers also worked to address maternal postpartum depression by enhancing workflows between medical and behavioral health and connecting members to postpartum care.

Population Health Improvements:

On average, the predictive model identifies high-risk members 52 days (more than seven weeks) earlier than traditional identification methods, which allows care managers to intervene much earlier and make a larger health impact.

The program led to:

• 1.4 additional prenatal visits per member engaged
• An 11.6% decrease in pre-term delivery rates for the engaged population
• A $5,000+ decrease in overall delivery costs per engaged member

Takeaway:

Harnessing predictive models to identify women with a high-risk pregnancy earlier and providing personalized support leads to lower complication rates and healthier moms and infants.