Transforming Legacy Clinical Trial Schedules of Activities into Interoperable Digital Formats

Abstract

Objective: Clinical trial protocols contain Schedule of Activities (SoA) that specify patient visit timing and procedures. Extracting SoAs from legacy protocols into interoperable standards such as Fast Healthcare Interoperability Resources (FHIR) and the Unified Study Definitions Model (USDM) is limited by widely varying table formats and complex trial activities. ProtocolMiner automatically extracts SoAs from PDF protocols, interprets their semantics, and outputs standards-compliant representations.
Materials and Methods: Semantic analysis was applied to determine visit timing, procedures, cycles, and exceptions in 29 diverse protocols. The methodology maps to FHIR PlanDefinition and ActivityDefinition resources, as well as USDM components including ScheduleTimeline, Encounter, and Activity classes.
Results: ProtocolMiner created accurate, detailed timelines for 22 of 29 SoAs, with minor errors in seven. FHIR successfully represented complex timing relationships including relative timing with windows, open-ended repetitions with stopping criteria, and conditional execution. USDM mapping enables integration with CDISC-compliant study design workflows.
Conclusion: Automated SoA extraction from legacy clinical trial protocols and transformation into FHIR and USDM is feasible with high fidelity. ProtocolMiner bridges the gap between legacy protocols and emerging digital standards, enabling computational protocol management and cross-system interoperability in clinical research.