Over the past two decades, maternal mortality in the U.S. has risen sharply, and the U.S. now has the highest maternal death rate among high-income, developed countries. Preeclampsia remains a major driver of severe maternal complications and death.
To counterbalance this, AI-driven predictive technology is helping clinicians to identify risk early enough for closer monitoring and timely intervention. Instead of waiting for clinical decline, data captured in routine care can spot patterns sooner, leading to informed decisions about surveillance, escalation, and resource planning.
To discuss the shift toward predictive clinical decision support in maternal health, Digital Journal sat down with Lia Butler, Director of Sales, Marketing, and Business Development at NeoPredics, which has developed a predictive platform that uses non-invasive clinical data and biomarker inputs to improve maternal and newborn outcomes. Butler is also a preeclampsia survivor and an advocate for maternal and women’s health.
Digital Journal: How is AI-driven predictive technology beginning to shift the way we approach maternal care?
Lia Butler: The current one-size-fits-all approach to maternal care often drives over-management for some, delayed escalation for others, and a large group whose risk is rising before it is obvious. We wait for blood pressure to cross a line, symptoms to intensify, labs to come back worse… then we act. AI-driven prediction flips the timeline. By learning from patterns across routine noninvasive clinical signals, plus biomarker inputs when available, technology can flag risk while it is forming, not after it has arrived. When risk becomes visible earlier, care becomes more intentional. DJ: What data sources are most valuable for developing accurate risk models?
Butler: The work of developing accurate risk models comes in two steps. First is identifying which data elements carry predictive signal, meaning which variables and trajectories consistently show up early as disease progresses. Second is making sure those signals can be captured in real time.
From that lens, the most valuable data sources are ones that are already captured for patients in routine care and can be collected consistently across health systems. Clinical records are foundational for that reason: demographics, obstetric history, comorbidities, medications, ultrasound context, clinician documented symptoms, and longitudinal vital sign trends such as blood pressure. Biomarkers can then add specificity. In preeclampsia, for example, angiogenic markers like sFlt 1 and PlGF, including their ratio, can strengthen risk assessment. Patient apps and remote monitoring can add signal as well, but they are most valuable when inputs are structured, consistent, and digitally integrated into care.
DJ: What are the main barriers to adopting predictive analytics in everyday obstetric care in the U.S.?
Butler: The main barriers are trust, funding, and the lack of a clear pathway to adoption.
Maternal care is high stakes, so clinicians are understandably cautious about new technology. But we are turning a corner as more digital tools and predictive systems become routine and reliable.
However, building a predictive tool is a long, expensive process.
Teams must build the infrastructure and the pathway while building the product, and many technologies never reach the bedside.
Even when a tool works, additional barriers include complex EMR integration, confusing reimbursement, budget cycles, competing priorities, risk concerns, and internal politics. The good news is, these pathways are forming, and adoption is happening.
DJ: You recently launched a prediction tool for adverse outcomes related to preeclampsia in the EU called PreFree. Tell us about that platform.
Butler: PreFree is a clinical decision support tool built to predict the risk of severe features of preeclampsia. While many patients flagged as high risk never develop severe outcomes, without a reliable personalized risk score, care is often over- or under-prescribed. PreFree helps clinicians identify which patients are trending toward severe preeclampsia-related adverse outcomes requiring intense monitoring or inpatient observation, and which are low-risk and can be safely monitored on an outpatient basis.
PreFree combines routinely captured clinical information with key lab parameters and biomarkers to generate an individualized risk profile for augmented clinical decision making. At scale, the goal is sharper targeting of care: fewer unnecessary tests, treatments, and procedures, more timely escalation for those most likely to develop complications, less strain on clinicians and hospitals, and better maternal and newborn outcomes.
DJ: Is there a plan to launch PreFree in the U.S.?
Butler: Yes, PreFree is already clinically validated, CE-certified, and used with patients in Europe. We are actively pursuing the FDA pathway to meet U.S. regulatory requirements. Our goal is a late 2026 U.S. launch.
DJ: Where does pregnancy care break down outside the clinic, and how can remote monitoring and patient-facing tools close those gaps?
Butler: Pregnancy care often breaks down in the space between visits. Symptoms evolve at home, blood pressure can rise silently, and patients are left deciding what is normal versus urgent with limited context. That gap is growing in the U.S. because we have an epidemic of declining maternal care resources with more maternal care deserts, fewer OB units, and stretched clinical teams.
Remote monitoring and patient-facing tools can close those gaps by making the home the first line of early detection and connection. Home blood pressure and symptom tracking can surface change sooner, while structured check-ins help patients report the right information at the right time. When those inputs flow back to the care team in a usable way, they enable earlier outreach, more targeted follow-up, and clearer triage, including when to reassure, when to bring someone in, and when to escalate immediately.
DJ: What emerging data sources and technologies will most improve maternal risk prediction over the next few years?
Butler: The biggest near term unlock is democratizing data sharing, especially cross-border. During a recent trip to Barcelona, it was hard to miss how parts of Europe are already operationalizing this through EU level infrastructure, including MyHealth@EU for cross border exchange and the European Health Data Space framework for broader interoperability and reuse. Scaling that kind of connectivity matters because maternal risk models get dramatically better when they are trained and validated on larger, more diverse populations rather than a single health system’s slice of reality.
On the data side, the most impactful technologies are reliable home blood pressure monitors, wearable tech that captures trend shifts between visits, and structured patient reported symptoms that are low friction and clinically actionable. The combination of more continuous non-invasive signals plus truly shareable, interoperable clinical data is what will push maternal risk prediction forward fastest over the next few years.
And once clinicians and innovators can access that data at a reasonable cost, everything changes. The barrier drops, the iteration cycle accelerates, and innovation becomes unstoppable.
