A scoping review of real-world data sources for retrospective oncology analysis in Japan

Background Cancer remains the leading cause of mortality in Japan, creating significant demand for high-quality real-world evidence to support clinical research and decision making. This study aims to provide a comprehensive overview of oncology studies using real-world data (RWD) in Japan, assessing the suitability of these data sources for oncology research. Methods Searches were conducted in PubMed, Ichushi-Web, and Google Scholar for studies published from 1 January 2020 to 28 February 2025. Eligibility criteria focused on retrospective observational studies utilizing readily available RWD databases. The availability of information on patient demographics, clinical characteristics, genomic information, treatment patterns, and clinical outcomes, as well as data source limitations reported by authors, was extracted and analyzed. Results Of 484 studies screened, 121 met the eligibility criteria. The majority utilized administrative claims data (52%), followed by registries (44%) and electronic health record (EHR)-derived data (6%). Limitations varied by data source. Claims data lacked clinical detail, resulting in challenges in accurate study cohort identification and outcomes research; registries contained deep genomic information but had high clinical data missingness and restricted cohorts; EHR-derived data provided detailed clinical insights but were less established in scale. Conclusion This scoping review highlights the strengths and limitations of Japanese RWD sources, revealing the frequent use of claims data in oncology research despite their limited clinical granularity. Future research should increasingly leverage clinically rich and longitudinal data sources, particularly EHR-derived datasets, to enable more precise cohort identification and capture meaningful clinical outcomes, better supporting clinical and research needs in Japanese oncology.