A timeline-based framework for aggregating and summarizing electronic health records
Filip Dabek, Elizabeth Yakes Jimenez, Jesus J. Caban · 2017
Electronic Health Records (EHRs) contain a significant amount of longitudinal information about a patient including pre-existing conditions, earlier diagnosis, previous treatments, active medications, base-line measurements for different clinical results, and much more. Unfortunately, data integration within an EHR and across different EHRs continue to be a limiting factor that threatens patient safety and the efficiency of healthcare providers. The disparate nature of the clinical data even within a single EHR often results in clinicians having to access and review a number of reports, modules, and tabs to access different data elements and clinical results. Due to the fragmented nature of EHR interfaces and the number of interactions that are needed to access clinical data, clinicians often spend a considerable part of their time going through the EHR of a patient in order to get a comprehensive overview and to be able to provide quality care. Data visualization and the integration of analytic models within graphical interfaces present a unique opportunity to effectively combine multiple clinical data sources and reduce the cognitive burden that disparate reports often have for end-users. With the ability of visualization techniques to summarize different data elements, we present a timeline-based framework to effectively aggregate and summarize the disparate clinical data of a patient enclosed within an EHR. The interface combines a set of visualization techniques with machine learning summarization approaches to optimize the process of understanding a patient's history through views that allow for easily skimming and jumping through time, filters for limiting the amount of information shown, and a hierarchy of summaries that provide an interface to view and compare different time frames.