A Systematic Review of Knowledge Visualization Approaches Using Big Data Methodology for Clinical Decision Support

Abstract

This chapter reports on results from a systematic review of peer-reviewed studies related to big data knowledge visualization for clinical decision support (CDS). The aims were to identify and synthesize sources of big data in knowledge visualization, identify visualization interactivity approaches for CDS, and summarize outcomes. Searches were conducted via PubMed, Embase, Ebscohost, CINAHL, Medline, Web of Science, and IEEE Xplore in April 2019, using search terms representing concepts of: big data, knowledge visualization, and clinical decision support. A Google Scholar gray literature search was also conducted. All references were screened for eligibility. Our review returned 3252 references, with 17 studies remaining after screening. Data were extracted and coded from these studies and analyzed using a PICOS framework. The most common audience intended for the studies was healthcare providers (n = 16); the most common source of big data was electronic health records (EHRs) (n = 12), followed by microbiology/pathology laboratory data (n = 8). The most common intervention type was some form of analysis platform/tool (n = 7). We identified and classified studies by visualization type, user intent, big data platforms and tools used, big data analytics methods, and outcomes from big data knowledge visualization of CDS applications.