Global Tech UpliftLinking intelligence

Interoperability · Decision support

At Boston Children’s Hospital

Maternal–infant health integration and dyad dashboard

Mapped national standards for substance-exposed mother–infant care, then prototyped an interoperable dyad dashboard to use them at the point of care.

  1. Map
  2. Standardize
  3. Build
  4. Validate
Role
Principal Data Scientist and Product Lead
Outcome
Data standards framework published in JAMIA (2025); dashboard prototype endorsed for a hospital pilot by HHS stakeholders. Read the paper
Focus
FHIR · USCDI · SMART on FHIR · coordinated care · synthetic data · generative AI
Dyad dashboard prototype showing synchronized maternal and infant information.
Part A

A data standards framework for mother and infant care

The problem

  • Separate maternal and infant EHR systems keep clinicians from seeing the whole dyad when treating substance-exposed families.
  • Maternal treatment plans, infant NAS scores and social services involvement lack consistent standard fields.
  • Privacy rules and lagging claims data slow exchange, leaving care teams without timely, shareable context.

The solution

  • Catalogued 180 clinical and social data elements and mapped each to USCDI and FHIR to quantify coverage.
  • Surveyed HIEs and hospitals on current mother–infant linkage and adoption of national standards.
  • Wrote recommendations that tell agencies which elements to prioritize for dyad coordination.

Architecture

  • Expert review plus Python-assisted lookups to align each element to standard vocabularies.
  • A structured HIE survey, analysed quantitatively and qualitatively.
  • Sankey charts and coverage dashboards to show where standards succeed.
  • Mappings validated with neonatologists, addiction specialists and informaticians.

Results and impact

  • Showed most dyad data needs are already covered by national standards.
  • Gave implementers a vetted reference in JAMIA.
  • Informed ONC equity guidance and an HHS “Dyad Data Tracker” initiative.
Skills and tools
Health data standardsUSCDI v4 and FHIR R4; custom dyad requirements mapped to standard fields.
Data mapping and scriptingSpreadsheets with Python automation to track coverage and gaps across 180 elements.
Survey design and analysisTargeted HIE surveys on real-world linkage capability.
Stakeholder collaborationONC, HHS and clinical partners; findings turned into policy guidance.
Domain and regulationNAS scoring, maternal OUD care, HIPAA and 42 CFR Part 2.
Part B

A dyad care dashboard prototype

The problem

  • Mother and infant records live in different systems, so clinicians toggle between screens during fast decisions.
  • Key maternal information is buried in long notes, hard to correlate with infant withdrawal trends.
  • Teams rely on manual updates, so dyad insight rarely reaches the bedside in time.

The solution

  • A SMART-on-FHIR app that pulls linked mother–infant data into one parallel layout.
  • Content and interactions prioritized with OB, NICU and social work teams.
  • Realistic synthetic data to show core cases, such as maternal toxicology beside infant NAS trajectories.

Architecture

  • Low- to high-fidelity designs in Balsamiq and Figma, grounded in clinician journey maps.
  • Front end in HTML, CSS and JavaScript.
  • FHIR queries for medications, labs and vitals, pairing records by shared identifiers.
  • Synthetic dyad datasets generated with GPT-4 and custom scripts.
  • Usability walkthroughs with clinicians shaped layout and highlighting.

Results and impact

  • Clinicians saw the value of maternal context beside infant NAS scores right away.
  • HHS and hospital leadership endorsed moving to a pilot.
  • Prototype code with synthetic FHIR data released for other health systems.

Next step

Records in different systems? Let’s talk.

A free 30-minute discovery call for your team.

Book the call More for healthcare