Annotation of clinical datasets using openEHR Archetypes as a solution for data access issues faced in biomedical projects

Abstract

Clinical datasets are kept in diverse and disparate formats that are specific to systems these datasets are created in and might not be related to any known clinical data modelling standard. This makes their reuse or utilisation of clinical information in processes like biomedical research difficult. We conduct case studies of six biomedical research projects, assess data access-related difficulties encountered and reach conclusions on the overall impacts those difficulties had on the projects. We conclude that manual mapping of composite concepts found in clinical datasets has been reported as the most constraining issue and that the affected aspects of the projects were the length, cost and limited validity of project results. We then suggest that composite concepts are annotated using standards-based information models - openEHR Archetypes in particular. We then justify that approach and provide guidelines on the structure of the method that would facilitate its application.