Leaf

UWM ITS Clinical Data Explorer

As a result of technology advancements in clinical data warehouse performance and storage capacity, the amount and the complexity of clinical data around clinical staffs and clinical researchers are constantly increasing. However, due to cognitive limitations of human beings, information processing by human mind is very slow and limited. Especially in design of informatics tools in clinical environment, enabling various members of the clinical staff (physicians, technicians, nurses, students, managers) to take advantage of effectively presenting and interacting with data should be prioritized.

Information visualization can assist users to reveal deeper details of data by exploiting their visual recognition abilities1. Information Visualization is well-studied in medicine2, public healthcare3, electronic medical data4, and medical imaging5, but is relatively new to clinical research informatics. Current clinical informatics research includes plentiful examples in visualizing problem- and domain-specific clinical data, such as distributed time-oriented clinical records and their analysis6, but our goal is to help researchers extract retrospective data obtained through query tools and delve into it by data visualization for a more general purpose.

Leaf is a next-generation self-service clinical data explorer sponsored by Institute of Translational Health Science, University of Washington. It supplies aggregate counts, information tables, basic visualizations and exporting functions of patient populations from Caradigm clinical data repository, which is effective at estimating patient cohort sizes and exporting cohort information for purpose of quality improvement and research. Its exporting destination REDCap is a mature web application for building and managing online surveys and databases and has an extendable architecture where plugins with additional features can be developed. Currently REDCap can export patient data extracted from Leaf to data analysis tools such as R, SPSS, SAS, Stata, however, the distance between cohort identification and cohort information visualization is far more than intuitive and time-consuming for Leaf users.

Visualization systems such as HARVEST7 focus primarily on visualizing individual patients’ longitudinal medical history other than an entire cohort. SMART apps build a plugin inside i2b2, providing an EMR-like view and a natural-feeling medical review process for each patient8. Gnaeus9 is an example of a cohort visualization tool, but it does not assist in finding the cohort. Important research has also been undertaken on visualization of patient histories, such as the LifeLine and the KNAVE projects7. However, none of them match our attempts to bridge the gap between real-time clinical data from the whole patient population in University of Washington Medical Center and its affiliated medical institutes and clinical knowledge discovery. Meanwhile, a user-friendly application in clinical settings will be more welcomed and intuitive for potential users. Therefore, Tableau as a successful visualization software focused on business intelligence is a handy choice for us to fill in the gap.

References:

1. Harris DR, Henderson DW, Sciences T. i2b2t2 : Unlocking Visualization for Clinical Research. :98-104.
2. Chittaro L. Information visualization and its application to medicine. Artificial intelligence in medicine.2001;22(2):81-88.
3. Shneiderman B, Plaisant C, Hesse BW. Improving healthcare with interactive visualization. Computer.2013;46(5):58-66.103
4. West VL, Borland D, Hammond WE. Innovative information visualization of electronic health record data: a systematic review. Journal of the American Medical Informatics Association. 2015;22(2):330-339.
5. Bui AA, Hsu W. Medical Data Visualization: Toward Integrated Clinical Workstations. In: Medical Imaging Informatics. Springer; 2010. p. 139-193.
6. Shahar Y, Goren-Bar D, Boaz D, Tahan G. Distributed, intelligent, interactive visualization and exploration of time-oriented clinical data and their abstractions. Artificial intelligence in medicine. 2006;38(2):115-135.
7. Hirsch JS, Tanenbaum JS, Gorman SL, Liu C, Schmitz E, Hashorva D, et al. HARVEST, a longitudinal patient record summarizer. Journal of the American Medical Informatics Association. 2015;22(2):263-274.
8. Wattanasin N, Porter A, Ubaha S, Mendis M, Phillips L, Mandel J, et al. Apps to display patient data, making SMART available in the i2b2 platform. In: AMIA Annual Symposium Proceedings. vol. 2012. American Medical Informatics Association; 2012. p. 960.
9. Federico P, Unger J, Amor-Amoros A, Sacchi L, Klimov D, Miksch S. Gnaeus: utilizing clinical guidelines for knowledge-assisted visualization of EHR cohorts. In: EuroVis Workshop on Visual Analytics (EuroVA) vol. 2015. The Eurographics Association; 2015.