Improving Providers’ Decision-Making and Reducing Information Overload Using Information Visualization in Electronic Health Records

Project Summary/Abstract
Importance: Poor design of electronic health records (EHR) can cause providers to overlook crucial data,
consequently impeding decision-making and adversely affecting patient safety and quality of care. Typically
spending 3.3 hours per day using the EHR, ICU providers monitor 2.5 million data points per month and
respond to 187 alerts per patient per day Visualization is the science of transforming textual data into a visual
display. Dashboards are novel information visualization methods to organize data and hold the potential to
reduce information and cognitive overload, thereby improving providers' decision-making through increased
efficiency and reliability.
Objective: Our goal is to investigate the efficacy of novel information visualization methods to improve ICU
providers' decision-making, efficiency, and satisfaction.
Specific aims: (1) Characterize the effect of EHR information overload on the quality of care by measuring
prescribing providers' performance, information seeking load, information processing load using real-time
patient data in institutional Epic® and Cerner® EHR in a laboratory setting.
(2) Evaluate the effect of information visualization on providers' decision making; performance; efficiency;
fatigue; workload; and satisfaction through real patient data in a randomized controlled trial.
(3) Determine the acceptability, ease of use, and barriers to implementing information visualization dashboards
into practice in an ICU setting, using guided interviews with ICU providers.
Study Design: We will recruit ICU providers from four academic medical centers: two Epic® sites (UNC, Mayo
Clinic) and two Cerner® sites (MedStar, University of Pittsburgh). The first study will aim to characterize the
effect of current EHR information overload on providers' information processing abilities. Second, we will test
the efficacy using an information visualization dashboard to improve the decision-making, efficiency,
performance of providers compared to their institutional HER through a randomized trial. Third, we will conduct
guided interviews to understand provider's perceptions around the implementation and use of visualization
dashboard in real ICU settings.
Discussion: We anticipate that completion of these aims will yield the following outcomes: 1) to add new
knowledge around information overload-related patient safety risks in current EHRs; 2) to determine the
efficacy of a visualization tool compared to prominent EHR interfaces Epic® and Cerner®; and 3) to create
new knowledge around visualization implementation guidelines and usability best-practices. Because the
visualization tool may improve the decision-making process for providers, we expect its use will increase
patient safety and decrease the number of medical errors.