Seonhun Lee, Brenda A. Nyarko, Gina L. Georgadarellis, Frank C. Sup, Karen K. Giuliano
Intravenous (IV) smart pumps are widely used medical devices in acute care, yet persistent usability challenges continue to contribute to medication administration errors. Although these devices share standardized functional requirements, their interface designs vary considerably, influencing cognitive workload, error likelihood, and interaction efficiency. This study compared perceived cognitive workload and error patterns while programming four IV smart pumps to examine how interface design influences performance during a standardized IV fluid administration task. A within‑subjects, repeated‑measures laboratory study was conducted with 31 nursing students. Participants programmed four IV smart pumps (Becton Dickinson [BD] Alaris, Baxter Sigma Spectrum, Intensive Care Unit [ICU] Medical Plum 360, and Ivenix Infusion System) while wearing eye‑tracking glasses. Programming errors were recorded, and perceived cognitive workload was assessed using the National Aeronautics and Space Administration -Task Load Index (NASA-TLX). Repeated‑measures ANOVA evaluated differences across devices. Pump type significantly affected overall workload (p < .001). The Ivenix infusion system produced the lowest perceived cognitive workload, shortest programming times, and no observed programming errors. In contrast, the other pumps were associated with higher perceived cognitive workload, more corrected and uncorrected errors, and greater variability in task duration. Error patterns suggested mismatches between interface structure and user expectations, particularly during channel selection and drug‑library navigation. Although IV smart pumps perform the same clinical functions, they vary substantially in usability, programming accuracy and perceived cognitive workload. These findings underscore the importance of designing devices to be aligned with their intended use, environments in which they will be used, and clinical user needs.