Real-World Evidence BRIDGE: A Tool to Connect Protocol With Code Programming

Area of Science:

• Health Informatics
• Data Management
• Pharmacovigilance

Background:

• Documentation of programming decisions in Real World Evidence (RWE) studies is crucial for transparency and reproducibility.
• Existing methods for documenting RWE study designs can be fragmented and lack standardization.
• The Vaccine Monitoring Collaboration for Europe (VAC4EU) initiative highlighted the need for improved RWE study documentation.

Purpose of the Study:

• To enhance the documentation of programming decisions in Real World Evidence (RWE) studies.
• To develop a machine-readable metadata schema for RWE study designs.
• To facilitate the translation of study designs into analytical pipelines.

Main Methods:

• Analyzed statistical analysis plans (SAPs) from the VAC4EU initiative.
• Designed a machine-readable metadata schema (RWE-BRIDGE) with study sections, codelists, and time anchoring.
• Developed a guide and a Shiny app for populating and validating the metadata schema.

Main Results:

• Developed RWE-BRIDGE, a relational database schema with 12 tables across four study design sections.
• The schema includes definitions for study variables, cohorts, outcomes, and data retrieval.
• RWE-BRIDGE is available on GitHub and has been applied in several RWE studies within VAC4EU.

Conclusions:

• RWE-BRIDGE provides a systematic approach to detailing RWE study parameters, including variables, time anchoring, and algorithms.
• The metadata schema promotes FAIR principles, enhancing collaboration and transparency between researchers and programmers.
• RWE-BRIDGE is flexible, adaptable to various data models and programming languages, and supports specific SAP needs.