Can Visualizations Complement Quantitative Process Analysis Measures? A Case Study of Nurses Identifying Patients Before Administering Medications. (June 2013)

• Record Type: Journal Article
• Authors: Marquard, Jenna L.; Jo, Junghee; Henneman, Philip L.; Fisher, Donald L.; Henneman, Elizabeth A.
• Is Part Of: Journal of cognitive engineering and decision making. Volume 7:Number 2(2013)
• Page Start: 198
• Page End: 210
• Publication Date: 2013-06
• Subjects: medication administration -- medication error -- visualization -- Markov chain -- sparkline; Human-computer interaction -- Periodicals; User-centered system design -- Periodicals; 004.019
• DOI: 10.1177/1555343412457683
• Languages: English
• ISSNs: 1555-3434

Abstract

The objective of this study is to demonstrate the effectiveness of visualizations for exploring one error-prone health care process: nurses verifying patients' identities during the medication administration process. We employed three types of process visualizations (i.e., Markov chains, sparklines, and timeline belt visualizations) to explore process execution data from an experiment wherein nurse participants ( N = 20) administered medications to three patients in a simulated clinical setting. One patient had an embedded error, with the medication being incorrect for the patient. The visualizations allowed us to view aggregate and individual-level process execution data, providing insights into the orders in which participants completed process steps. Although we used eye tracking videos, the system developed in this study can automatically generate visualizations using large process execution data sets produced from an array of sources, including observations, sensors, and health IT audit trails. In this article, we demonstrate that the visualizations provide insights complementary to quantitative measures regarding what process steps participants likely used to identify errors, with the visualizations requiring less work to produce. Therefore, the visualizations may be an effective means for efficiently comparing sets of process execution data (e.g., different individuals completing a process, pre– and post–technology implementation, pre– and post–quality improvement intervention).