Presentation (CTI)

OBJECTIVES: Claims databases are widely used for real-world evidence generation in Japan but often provide limited depth of clinical information, including laboratory results, biomarkers, and disease severity indicators. We aimed to describe DATuM IDEA, a Japanese real-world database integrating electronic health record (EHR)-derived clinical data with claims data under the Next-Generation Medical Infrastructure Act.
METHODS: We conducted a descriptive data-resource profile using the April 2026 data cutoff. We summarized governance, anonymization and opt-out procedures, patient-level EHR-claims linkage, data domains, coding and standardization processes, institutional coverage, patient demographics, and selected clinical content.
RESULTS: Data have been collected since 2019. As of the April 2026 data cutoff, DATuM IDEA contained 1,929,950 unique patients from 63 institutions across 33 prefectures. The database is updated quarterly with an approximately 6-month data lag. Patient-level linkage is enabled by an anonymized identifier assigned under the certified anonymization framework, allowing deterministic linkage between EHR-derived clinical information and claims data within each institution. The database includes demographics, diagnoses, medications, procedures, surgeries, healthcare costs, visit records, and results from more than 3,000 types of laboratory tests, including biomarker tests. The patient population was broadly distributed by age and sex, with a mean age of 53.7 years and 48.3% male. Neoplasms were recorded in 557,531 patients, including 534,857 with solid-site neoplasms and 42,307 with hematological neoplasms. The predominance of acute-care and cancer care hospitals may be advantageous for studies of specialized care, inpatient treatment, procedures, oncology, and healthcare resource use, while limiting capture of primary-care-centered pathways and care outside participating institutions.
CONCLUSIONS: DATuM IDEA provides a legally governed model for privacy-preserving integration of EHR and claims data in Japan. By linking EHR-derived clinical information with claims-based healthcare utilization and cost data, it may support clinically granular RWE and health economic analyses beyond EHR- or claims-only data sources.