Improving the Efficiency and Ease of Healthcare Analysis Through Use of Data Visualization Dashboards

Abstract

The digitization of a patient's health record has profoundly impacted medicine and healthcare. The compilation and accessibility of medical history has provided clinicians an unprecedented, holistic account of a patient's conditions, procedures, medications, family history, and social situation. In addition to the bedside benefits, this level of information has opened the door for population-level monitoring and research, the results of which can be used to guide initiatives that are aimed at improving quality of care. Cerner Corporation partners with health systems to help guide population management and quality improvement projects. With such an enormous and diverse client base-varying in geography, size, organizational structure, and analytic needs-discerning meaning in the data and how they fit with that particular hospital's goals is a slow, difficult task that requires clinical, statistical, and technical literacy. This article describes the development of dashboards for efficient data visualization at the healthcare facility level. Focusing on two areas with broad clinical importance, sepsis patient outcomes and 30-day hospital readmissions, dashboards were developed with the goal of aggregating data and providing meaningful summary statistics, highlighting critical performance metrics, and providing easily digestible visuals that can be understood by a wide range of personnel with varying levels of skill and areas of expertise. These internal-use dashboards have allowed associates in multiple roles to perform a quick and thorough assessment on a hospital of interest by providing the data to answer necessary questions and to identify important trends or opportunities. This automation of a previously manual process has greatly increased efficiency, saving hours of work time per hospital analyzed. Additionally, the dashboards have standardized the analysis process, ensuring use of the same metrics and processes so that overall themes can be compared across hospitals and health systems.

Update on Data Reuse in Health Care

“…Increasingly, traditional sources of health-related data, such as claims databases, can be linked with clinical data from EHRs to better understand the safety of medications we prescribe to patients [8]. In fact, the wealth of information in all sources of data can be used to improve efficiency in clinical care [9] as well as compare the effectiveness of therapeutic options [10].…”

Modeling and simulating hospital operations in a 3D environment

“…They have been frequently developed for business managers to integrate and summarize data and key performance information in a visual display (Dowding et al 2015). Previous healthcare related dashboards have been developed to analyze bed occupancies (Daley et al 2013), readmission prevention (Stadler et al 2016) and performance management (Mesabbah and Arisha 2016). Other dashboards have been built as Discrete-Event Simulators (DES) to estimate capacity levels (Werker et al 2009), measure policy impact (Fletcher et al 2007), simulate patient scheduling (Lee et al 2013;Werker et al 2009;Quevedo and Chapilliquén 2015;Harper and Gamlin 2003) and aid strategic decision making (Leskovar et al 2011;Vanberkel and Blake 2007;Ballard and Kuhl 2006).…”

“…Those that give focus to patient scheduling tend to focus on outputs related to utilization and patient waiting (for example, Lee et al (2013); Al-Araidah et al (2012)), while those that focus on hospital management tend to measure bed occupancy and patient length of stay or readmissions (for example, Daley et al (2013);Stadler et al (2016)). …”

“…The ability to view the right level of detail for the right user is also vital to ensure the right information is dealt with by the right people (Pauwels et al 2009). Too much data can be overwhelming to users, with usability and utility of a tool being easily lost in noise from too much data (Stadler et al 2016). Similarly, the time taken to obtain value from data can be prohibitive if the tools do not present the data in an easy to understand format (Stadler et al 2016;Halford et al 2005).…”

“…Too much data can be overwhelming to users, with usability and utility of a tool being easily lost in noise from too much data (Stadler et al 2016). Similarly, the time taken to obtain value from data can be prohibitive if the tools do not present the data in an easy to understand format (Stadler et al 2016;Halford et al 2005). It has been argued by Halford et al (2005) that accuracy of interpretation decreases when the complexity of the tool increases.…”