Development and Description of a Synthetic,High-Fidelity,Emergency Department Patient Dataset for the Evaluation of Healthcare IT Products

Abstract

Developing novel interfaces for high-risk situations, such as the Emergency Department, requires a sufficient quantity of detailed patient data to support usability and evaluation activities, yet patient privacy restrictions often prevent the use of actual patient data for these activities. We developed a synthetic dataset to provide a suitable alternative to the use of actual patient data that can be integrated into a variety of research activities. The Emergency Department Information Systems (EDIS) Dataset was developed through close collaboration of experts in Emergency Medicine, Human Factors, and Systems Engineering and provides an ecologically valid set of data for 54 patients, treated in an Emergency Department operating at steady-state, with realistic patient loads and flow. The dataset includes both static and dynamic data for each patient case over a 500-minute time period. A sample application of the dataset is provided to demonstrate how the dataset was used to support the design and evaluation of novel EDIS interface displays and its potential adaptation for future HIT research. This dataset provides a readily adaptable alternative to researchers in need of synthetic patient data to support HIT research and development activities. The EDIS dataset and supporting material are freely available through the University at Buffalo Institutional Repository and can be directly accessed with the URL: hdl.handle.net/10477/75188.