The Invisible Burden: How Measurement Gaps shape the insight for U.S. Maternal Mental

• In Event: Health Financing and Spending

Abstract

Maternal mental health is increasingly recognized as a critical policy priority in the United States, given its far-reaching implications for maternal well-being, infant outcomes, and healthcare spending. Yet the way maternal mental health is measured remains fundamentally misaligned with clinical reality. Most policy analysis relies on administrative claims data, which capture only conditions that have been formally diagnosed and coded for billing. As a result, a substantial share of individuals experiencing clinically significant symptoms never appear in the data that inform funding decisions, program evaluation, and policy design.
This paper examines how measurement choices shape what policymakers can see—and what they systematically miss—in maternal mental health. Drawing on linked screening and administrative data from the Health Care Financing Initiative (HCFI), we compare symptom-based identification using the Edinburgh Postnatal Depression Scale (EPDS) with diagnosis-based identification using ICD-10 codes for a sample of 3,144 birthing individuals between 2020 and 2022.
We document a clear “diagnostic drop-off” between screening and diagnosis. While 22.3% of individuals screened positive for maternal mental health symptoms, only 18.2% received a corresponding diagnosis. Those identified through screening but not diagnosis are disproportionately more likely to be Black, publicly insured, and unmarried—populations already facing structural barriers to care. Importantly, this “invisible” group is also associated with worse outcomes, including higher rates of preterm birth and greater healthcare costs, suggesting that diagnosis-based measures may substantially underestimate both need and economic burden.
These findings highlight that measurement is not merely a technical detail but a consequential policy choice. Reliance on diagnosis-based data risks distorting estimates of prevalence, obscuring inequities, and undervaluing early-stage interventions that operate before formal diagnosis.
We argue that improving maternal mental health policy requires rethinking measurement itself. Integrating screening-based indicators into policy-relevant data systems can better capture unmet need, strengthen program evaluation, and support more equitable and effective resource allocation.

Authors

• Wenjie Deng, Georgetown University
  Presenting Author

• Aditi Bhardwaj, Georgetown University
  Non-Presenting Co-Author

• Carol Barnett Davis, Georgetown University
  Non-Presenting Co-Author