The 2022 March of Dimes Report Card gives the U.S. a grade of D+ for preterm birth, with the worst maternal and neonatal outcomes of any industrialized nation. Worse yet, there are vast gaps in health equity, where Black women are three to four times more likely to experience severe morbidity or mortality through childbirth. While conventional wisdom suggests little can be done to stop preterm birth in the highest-risk patients, the advent of new data science techniques offers a real and actionable opportunity to change this maternal health crisis devastating the United States.
Artificial Intelligence (AI) and Machine Learning (ML) are two powerful tools that can help this effort. By using this technology to mine data sets and generate predictions, health plans and providers can identify at-risk mothers early. Typically pregnant women do not have their first prenatal appointment until 10 weeks gestational age, at which point most of the first trimester has passed. For women with pre-existing conditions such as diabetes or substance abuse issues, putting a care plan in place during this critical trimester can help reduce both mother and infant mortality.
Advanced analytics assemble a comprehensive clinical and social profile for each pregnant patient with personalized health improvement plans, making it easy for care managers to complete actions that optimize the health and well-being of women. For example, care management teams can leverage maternity-specific risk-stratification to prioritize their patient outreach. Trained on a database of over half a million mom and baby pairs, machine learning algorithms assess pregnant patients constantly in real-time allowing for timely engagement when conditions change. Pregnancy is finite with lifelong ramifications, so any rising risk is urgent.
This approach comes at a critical time when the U.S. has the highest maternal mortality rate among industrialized countries, especially for women of color. Since 2000, the maternal mortality rate has risen nearly 60%, making it worse now than it was decades earlier, and the financial costs are staggering. In 2019, the maternal and child costs due to maternal morbidity for U.S. births were $32.3 billion, from conception to age five.
Dr. Holly Puritz, a practicing obstetrician/gynecologist with The Group for Women, a division of Mid-Atlantic Women’s Care, vice chairman of Sentara Healthcare’s Quality Care Network, and a member of the advisory board for Lucina Analytics, considers the use of data analytics imperative for effective care. “As an OB/GYN, I see a genuine sense of urgency for digital tools like those from Lucina.” Dr. Puritz adds, “This kind of AI-technology can be harnessed to improve not just perinatal engagement but maternal health equity and postpartum outcomes.”
To fully understand the impact predictive analytics can have on outcomes, Lucina Analytics provided statistics on its maternity platform. Across entire plan populations, their approach resulted in:
64% reduction in racial disparities for maternal outcomes
10.4% reduction in preterm birth
9.4% reduction in NICU utilization
$13,000 savings per NICU admission
To understand more specifically how predictive analytics can impact real lives, consider the following case review from Lucina.
A single mother raising two children under five finds herself pregnant again. She can’t afford to take time off work and childcare is out of her reach. As with her other pregnancies, she plans to wait until well into her third trimester to visit the doctor. The mother-to-be feels increasingly panicked at the thought of having another child when she can barely afford the two she already has. The only thing that helps her feel better is self-medicating.
Fortunately, the Lucina algorithm soon identifies the woman’s pregnancy, and clues in the woman’s health history soon alert the case manager (CM)* that this is potentially a high-risk case. As a result, the CM reaches out at eight weeks’ gestation. Using motivational interviewing techniques, the CM encourages the woman to see a doctor and ultimately is able to successfully schedule an obstetric appointment outside the woman’s work hours. As a result of this outreach, the woman is also placed in outpatient treatment for opioid abuse and receives a referral to a mental health professional for treatment of anxiety.
Continued follow-up from the CM helps ensure the woman is getting proper prenatal care in between office visits. These interventions help the woman deliver a healthy baby at 39 weeks gestation with no NICU stay.
It is important to note that while many health plans offer maternity analytics, they often fall short when it comes to early identification and real-time patient risk levels. Thus, in order to identify at-risk women and create proper treatment plans as early as possible, it is essential to embrace technology that includes predictive data and information like that illustrated above.
Statistics show that health plans do not identify 40% of at-risk mothers until the time of delivery. This means that legacy healthcare systems are allowing almost half of all pregnant women to fall through the cracks. With a predictive, advanced analytics platform using AI and ML, payors’ case managers come closer to ensuring that every woman can access the best healthcare in the world. Such a value-based framework can improve delivery outcomes and, over time, build healthier communities.
*Note: Lucina does not currently provide case management. The case manager in this case is employed by the woman’s health plan, which uses the Lucina platform.
Article written by Dr. Matt Eakins, a physician innovator with a degree in medicine from Northwestern University and a background in genetic research at the National Institutes of Health and the Mayo Clinic. He has deep expertise in innovation in the health technology sector and has founded several companies focused on delivering better patient outcomes. Dr. Eakins has a passion for bridging technology to today’s population healthcare demands. Currently, he is President of Lucina, a sophisticated digital health tool that is closing significant gaps in maternal health disparities. He has led the growth to support over Medicaid 100,000 deliveries annually, significantly improving outcomes for all and reducing health disparities by 64%.