Big Data/Machine Learning/AI
Governance-First AI Research Platform Accelerates Epidemiologic Studies: Multi-Site Implementation Allison Curry* Allison Curry Curry Curry Curry Curry Children’s Hospital of Philadelphia
Epidemiologic research infrastructure remains fragmented across design, recruitment, engagement, data collection, and analysis—creating operational barriers. Rising personnel costs, compliance requirements, and funding instability mean fewer studies, longer timelines, and staff effort diverted from science to documentation—inefficiencies that compound in under-resourced settings where disease burden is highest.
We developed a HIPAA-compliant Research Management System (RMS) to accelerate participant-based epidemiologic studies while meeting institutional governance requirements. The RMS consolidates operations across the study pipeline with role-based access, immutable audit trails, and IRB-ready documentation. Protocol-grounded AI capabilities enable: multilingual participant recruitment; automated consent with comprehension validation; and real-time tracking of progress, participant engagement, and metrics. Security architecture and study-level isolation were designed for academic implementation. We evaluated feasibility and preliminary performance, with ongoing validation comparing response rates, representativeness, and data quality to traditional methods.
Five institutions—Harvard, Johns Hopkins, UWash, UIowa, Nationwide Children’s—completed security/IRB reviews and initiated projects. The UIowa feasibility study (n=20 survey) showed clear AI interactions, 2-day completion vs. 4-week conventional estimate, and staff effort reduced from 50 to 4 hours. Other validations will be completed in Spring 2026.
A governance-first AI platform successfully implemented across diverse academic settings demonstrated feasibility for accelerating epidemiologic operations. Preliminary findings suggest substantial efficiency gains; larger studies are evaluating methodological validity, representativeness, and equity implications. The RMS enables epidemiologic research in resource-constrained settings and accelerates evidence generation for time-sensitive public health decisions.
