Presentation (CTI)

OBJECTIVES: To assess whether specialty physician-led chart abstraction (PLCA) produces real-world data (RWD) meeting FDA “fit-for-use” standards for regulatory-grade evidence, with a focus on data completeness, reliability, and contextual validity to support health economics and outcomes research (HEOR) in oncology.
METHODS: A systematic synthesis was conducted across retrospective, non-interventional PLCA studies in oncology/hematology, leveraging an internal archive of studies published from March 2020 to June 2025. Eligible studies featured U.S.-based PLCA from practice and hospital electronic health records (EHRs) as well as primary source lab and imaging findings into standardized electronic case report forms, rigorous quality control, and adherence to PRISMA 2020 and RECORD/RECORD-PE guidelines. The primary outcome was pooled missingness rates for key HEOR-relevant variable categories (demographics, treatment patterns, clinical characteristics, outcomes, safety, utilization, and unique insights).
RESULTS: Four studies (follicular lymphoma, HER2+ breast cancer, high-risk myelofibrosis, hepatocellular carcinoma) were included, with PCLA by board-certified oncologists/hematologists at an array of geographically diverse practice settings (i.e., both community and academic, dispersed across the U.S.). PLCA achieved high completeness for core HEOR variables: demographics, treatment patterns, adverse events, and clinical characteristics (aggregate means per category > 98% complete). Outcomes showed greater variability, reflecting documentation constraints inherent to RWD (> 90%). PLCA enabled high capture of “why-level” insights (e.g., rationale for treatment, eligibility for subsequent lines) rarely available in claims/automated EHR extractions (100%). Missingness was transparently reported, consistent with FDA guidance.
CONCLUSIONS: PLCA delivers enhanced data completeness, contextual validity, and contemporaneity relative to conventional RWD sources, supporting FDA and HEOR expectations for reliability and relevance in regulatory-grade evidence. By integrating dispersed documentation and expert judgment, PLCA supports fit-for-use datasets for complex endpoints and enables context-rich analyses for economic modeling, comparative effectiveness, and outcomes research. Future work should benchmark PLCA against automated pipelines and formalize quality control taxonomies to further advance HEOR practice.