Maternal-Infant Health Integration & Dyad Dashboard

Part A — Maternal-Infant Data Standards Framework

Mapped 180 critical data elements for substance-exposed mother–infant dyads to national health data standards, exposing interoperability gaps.
Demonstrated that over 91% of required dyad data can already flow through USCDI v4 and FHIR R4, enabling near-term implementation.
Published the framework in JAMIA (2025), influencing federal interoperability guidance for maternal-infant care.

Supplementary mappings detailing social and supportive care elements.

Built a SMART-on-FHIR dashboard that unifies mother and infant EHR data into a single clinical workspace.
Surfaced synchronized maternal medications, toxicology, and infant withdrawal scores to streamline care coordination.
Demonstrated the prototype with realistic synthetic data, securing stakeholder buy-in for a hospital pilot.

Separate maternal and infant EHR systems prevent clinicians from seeing the full dyad picture when treating substance-exposed families.
Critical dyad-specific details—maternal treatment plans, infant NAS scores, social services involvement—lack consistent standardized fields.
Privacy regulations and lagging claims data slow information exchange, leaving care teams without timely, shareable context.

Catalogued 180 clinical and social data elements essential to dyad care and mapped each to USCDI and FHIR definitions to quantify coverage.
Surveyed HIEs and hospitals to understand current mother–infant linkage practices and adoption of national interoperability standards.
Authored recommendations to close standard gaps and accelerate adoption, guiding agencies on data elements to prioritize for dyad coordination.

Combined manual expert review with Python-assisted lookups to align each dyad element to standard vocabularies without missing edge cases.
Designed and analyzed a structured HIE survey, blending quantitative aggregations with qualitative insights on linkage maturity.
Visualized coverage and gaps through Sankey charts and coverage dashboards to clearly communicate where standards succeed.
Validated mappings with neonatologists, addiction specialists, and informaticians to ensure semantic accuracy before publication.
Documented integration pathways so future tools can rely on the mapped standards, laying the groundwork for operational dashboards.

Proved that a majority of dyad data needs are already covered by national standards, reducing perceived technical barriers for health systems.
Published findings in JAMIA, giving implementers a vetted reference for configuring systems that track maternal-infant outcomes.
Informed ONC equity guidance and an HHS “Dyad Data Tracker” initiative that uses the mapped elements to monitor outcomes.

Technique/Skill	Tools/Implementation
Healthcare Data Standards	Deep expertise in USCDI v4 and FHIR R4; mapped custom dyad requirements to standard fields to ensure semantic alignment.
Data Mapping & Scripting	Combined spreadsheets with Python automation to catalog 180 elements and track coverage versus gaps.
Survey Design & Analysis	Crafted targeted HIE surveys and analyzed responses to understand real-world linkage capabilities.
Stakeholder Collaboration	Coordinated with ONC, HHS, and clinical partners, translating technical findings into actionable policy guidance.
Data Visualization	Built Sankey diagrams and coverage dashboards to communicate mapping outcomes to decision-makers.
Domain & Regulatory Knowledge	Navigated NAS scoring, maternal OUD care, HIPAA, and 42 CFR Part 2 to ensure recommendations fit real-world constraints.

Demonstrated standards-first solution design, a practice applied across subsequent interoperability and surveillance projects.
Bridged policy and engineering, ensuring technical solutions align with national health IT initiatives.
Modeled a rigorous discovery approach that informs later ML deployments, from ICU analytics to equity dashboards.

Journal of the American Medical Informatics Association (2025): detailed mapping of dyad data elements to USCDI/FHIR.Paper
Public supplementary spreadsheet cataloging all 180 elements and their standard codes for implementers.

Mother and infant records live in different systems, forcing clinicians to toggle across interfaces and risking missed context during fast-moving care decisions.
Key maternal information is often buried in long notes, making it difficult to correlate with infant withdrawal trends.
Cross-disciplinary teams rely on manual updates, so actionable dyad intelligence rarely reaches the bedside in time.

Designed a SMART-on-FHIR application that pulls authenticated data for linked mother–infant pairs into a unified, parallel layout.
Collaborated with OB, NICU, and social work stakeholders to prioritize content and interaction patterns that match dyad workflows.
Populated the prototype with realistic synthetic data to demonstrate core use cases—e.g., comparing maternal toxicology results with infant NAS trajectories.

Iterated low- to high-fidelity designs in Balsamiq and Figma, grounding interface choices in clinician journey mapping.
Implemented the front end with HTML, CSS, and JavaScript.
Queried FHIR resources—medications, labs, vitals—for both mother and infant, using shared identifiers to pair records in real time.
Generated synthetic dyad datasets with GPT-4 and custom scripts to test the interface under realistic clinical scenarios.
Ran usability walkthroughs with clinicians, adjusting layout and highlighting logic based on observation and feedback.

Clinicians immediately saw the benefit of seeing maternal context alongside infant NAS scores, validating the design direction.
Stakeholders across HHS and hospital leadership endorsed advancing to a pilot, citing the prototype as proof of feasibility.

Technique/Skill	Tools/Implementation
UI/UX Design	Led user-centered design with wireframes and high-fidelity prototypes tailored to dual-patient views.
Web Development	Built the SMART-on-FHIR front end with HTML, CSS, and JavaScript for secure deployment inside the EHR.
FHIR API Integration	Queried medications, labs, vitals, and notes for linked patients using OAuth2-authenticated FHIR calls.
Synthetic Data Generation	Created realistic dyad datasets with GPT-4 and custom scripts to test workflows without exposing PHI.
Usability Testing	Conducted iterative clinician walkthroughs, capturing insights that shaped layout and alert logic.
Product & Project Leadership	Orchestrated multi-phase development, aligning designers, engineers, and clinical advisors under agile cadences.

Reinforced ability to synthesize disparate data streams into unified experiences, echoed in biosurveillance and ICU projects.
Showcased the fusion of AI outputs with clinician-friendly UX, a throughline for bedside decision-support work.
Highlighted cross-disciplinary leadership, guiding teams from concept through demonstrable prototypes.

Internal report and demo video distributed to HHS opioid policy leaders outlining architecture and usability findings.
Planned AMIA/HIMSS submissions describing the standards-driven design approach and early clinician feedback.
Open-source prototype code (with synthetic FHIR data) released to encourage adoption and iteration by other health systems.
Featured in national health IT briefings and newsletters as a forward-looking example of dyad-centered care.