MenuBack to Poster-Presentations-Details
W-31: Retrospective Chart Review Studies: Opportunities and Challenges to Post-Market Evidence Generation
Krista Angela Payne
Vice President and Senior Research Scientist; RWE
To describe the challenges of implementing chart review studies due to data access, quality and timeliness of data availability, as well as proposed solutions for the generation of post-market evidence of treatment effectiveness, patterns of drug utilization, and clinical and safety outcomes.
Qualitative evaluation of 3 recent chart review studies in Europe and North America. Studies were assessed based on research questions/real-world data (RWD) requirements, types of practices providing the data, data collection methods and duration.
The studies were undertaken in gastroenterology, oncology and pulmonology practices. Number of patients and sites ranged from 220-340 patients and 8-27 sites; type of practices varied from small private practices to large university/academic institutions. Research questions focused on treatment patterns, clinical outcomes and resource utilization. Data were abstracted by local site staff and entered into an electronic data capture system. Data collection duration ranged from 8-12 months. Identified data access challenges included: lack of site interest to participate in chart review studies, lack of resources to complete the data entry, data abstraction duration longer than planned, and patient consent requirement in certain European countries leading to smaller overall enrolled population. Proposed recommendations to overcome these challenges include: improving site interest through better communication of the research benefits and permitting authorship where merited, identifying alternative resourcing models for sites interested in participating but who lack resources, and determining whether direct-to-EMR data extracts of fully anonymized data can be undertaken (including leveraging automated analytics tools).
While retrospective chart review studies can provide an abundance of data to answer diverse research questions to generate RWD, better site engagement, use of site EMRs and automatic analytic tools may allow more rapid and robust RWD generation. Exploring alternative approaches to retrospective data access from study sites and automated real-time output is warranted to generate RWD more efficiently.