The National Institutes of Health ’ s Biomedical

Research data warehouse Ontology a b s t r a c t The US National Institutes of Health (NIH) has developed the Biomedical Translational Research Information System (BTRIS) to support researchers' access to translational and clinical data. BTRIS includes a data repository, a set of programs for loading data from NIH electronic health records and research data management systems, an ontology for coding the disparate data with a single terminology, and a set of user interface tools that provide access to identified data from individual research studies and data across all studies from which individually identifiable data have been removed. This paper reports on unique design elements of the system, progress to date and user experience after five years of development and operation. Institutions that engage in both patient care and clinical research are increasingly turning to the development of clinical data repositories (or ''warehouses'') as resources for gathering data collected during routine patient care with the intent of re-using them to answer secondary questions [1]. The National Institutes of Health (NIH) is no exception. The Clinical Center, a 240-bed hospital on the NIH's Bethesda, Maryland campus is devoted exclusively to research. Since 1976, the Clinical Center has maintained two contiguous electronic health record systems (EHRs). The data from those systems, together with data from research systems at many of the NIH's 26 other institutes and centers, comprise an important collection of longitudinal information about common and rare conditions , diagnosed and treated in conventional and unconventional ways. In order to maximize the value of these data, the NIH launched the development of the Biomedical Translational Research Information System (BTRIS; pronounced BEE-triss), charged with providing a single, central resource that could support the data requirements of primary researchers, as well as those who seek to re-use old data to answer new questions related to research or hospital operations. Some aspects of BTRIS have been previously described briefly, including its general design [2], and a query tool for retrieval of data with personal identifiers removed [3]. The purpose of this paper is to provide further details about BTRIS's database design, terminology management, and query tools, and to provide a comparison with the architectures of other major current clinical data repository efforts. From 1976 to 2004, the Clinical Center's electronic health record was a version of the Medical Information Management System from Technicon Data Systems (Tarrytown, NY). Locally referred to as MIS …