FLOWER: Following Longitudinal Outcomes With Epidemiology for Rare Diseases

This study does not require data entry by treating site staff or physicians. Centralized data structuring is completed by xCures study staff. Data elements are aggregated, normalized, and coded to OMOP-based ontologies (SNOMED, LOINC, ICD-10, CTCAE, RxNorm, and MedDRA) in one process, permitting standardization of verbatim terms from medical records. The data collection platform supports 21 CFR Part 11-compliant data annotation with formal QC/QA process, medical review, and source data verification.

Beyond EMR data, raw DICOM images (MRI, CT files) can be collected from all sites of care and anonymized for integration with the clinical data. Molecular profiling and somatic or germline genomics results, and biochemical lab data, when available, are collected from commercial and academic sources and centralized. Additionally, patient- and caregiver-reported outcome surveys (PROs) can be collected to supplement information not found in clinical records.

Together, these clinical, imaging, biomarker, and assessment data will provide a comprehensive and longitudinal documentation of rare diseases in near real-time in a single observational basket study.

Traditional rare disease research registries rely on patients reporting many aspects of their condition via surveys or rely on key opinion leaders at specific institutions managing a team to enroll patients and annotate necessary data. These put unnecessary burdens on patients and strain limited research resources at medical centers. Gathering the necessary data and in sufficient quantities is often a limitation to successfully defining the natural history of a rare disease.