Design and evaluation of a safety-centered user interface for radiation therapy.

PURPOSE: As radiation therapy treatment grows more complex over time, treatment delivery has become more susceptible to adverse events and patient safety risks from use error. The radiation therapy monitoring and treatment delivery user interface explored in this study was redesigned using ecological interface design, a human factors engineering method, and evaluated to improve treatment safety. METHODS AND MATERIALS: An initial design concept was created based on previously completed analysis and informally evaluated in focus groups with radiation therapists. Sixteen newly graduated radiation therapists used both the redesigned and current system in a usability test to determine if the redesigned system better supported detection of errors. RESULTS: The redesigned system successfully improved the error detection rate of 2 errors: wrong treatment volume and wrong treatment site (P < .03 and P < .01, respectively). It also improved level 2 and level 3 situation awareness (ie, comprehension of the meaning of the information and the projection of the behavior of the technology: P < .01 and P < .01, respectively) and achieved a higher user satisfaction. CONCLUSIONS: The ecological interface design approach was found to be effective in redesigning a radiation therapy treatment delivery interface. Radiation therapists were able to deliver simulated radiation therapy with a higher rate of error detection and improved higher-level situation awareness, and participants preferred the redesigned interface to the current interface. Overall, the redesigned interface improved the radiation therapists' system understanding and ability to detect errors that affect patient safety.

The usability of ventilators: a comparative evaluation of use safety and user experience.

BACKGROUND: The design complexity of critical care ventilators (CCVs) can lead to use errors and patient harm. In this study, we present the results of a comparison of four CCVs from market leaders, using a rigorous methodology for the evaluation of use safety and user experience of medical devices. METHODS: We carried out a comparative usability study of four CCVs: Hamilton G5, Puritan Bennett 980, Maquet SERVO-U, and Dräger Evita V500. Forty-eight critical care respiratory therapists participated in this fully counterbalanced, repeated measures study. Participants completed seven clinical scenarios composed of 16 tasks on each ventilator. Use safety was measured by percentage of tasks with use errors or close calls (UE/CCs). User experience was measured by system usability and workload metrics, using the Post-Study System Usability Questionnaire (PSSUQ) and the National Aeronautics and Space Administration Task Load Index (NASA-TLX). RESULTS: Nine of 18 post hoc contrasts between pairs of ventilators were significant after Bonferroni correction, with effect sizes between 0.4 and 1.09 (Cohen's d). There were significantly fewer UE/CCs with SERVO-U when compared to G5 (p = 0.044) and V500 (p = 0.020). Participants reported higher system usability for G5 when compared to PB980 (p = 0.035) and higher system usability for SERVO-U when compared to G5 (p < 0.001), PB980 (p < 0.001), and V500 (p < 0.001). Participants reported lower workload for G5 when compared to PB980 (p < 0.001) and lower workload for SERVO-U when compared to PB980 (p < 0.001) and V500 (p < 0.001). G5 scored better on two of nine possible comparisons; SERVO-U scored better on seven of nine possible comparisons. Aspects influencing participants' performance and perception include the low sensitivity of G5's touchscreen and the positive effect from the quality of SERVO-U's user interface design. CONCLUSIONS: This study provides empirical evidence of how four ventilators from market leaders compare and highlights the importance of medical technology design. Within the boundaries of this study, we can infer that SERVO-U demonstrated the highest levels of use safety and user experience, followed by G5. Based on qualitative data, differences in outcomes could be explained by interaction design, quality of hardware components used in manufacturing, and influence of consumer product technology on users' expectations.