CARD1: Human Factors Evaluation of HeartMate 3 LVAD Peripherals: Insights gained from an Eye Tracking based Simulation Study

Abstract

Background: Although left ventricular assist devices (LVADs) have demonstrated excellent clinical outcomes and design improvements over the past decade, human factors issues including limited usability and unintuitive handling continue to challenge this therapy. The aim of this study was to evaluate user experience of former LVAD patients and laypersons (LP) with the HeartMate 3 (HM3) LVAD peripherals in simulated everyday and emergency scenarios. Methods: This cross sectional, single center cohort study included former non-HM3 LVAD patients post heart transplantation (HTX) and LP without any previous LVAD peripherals handling experience. Seven predefined everyday and emergency scenarios, including battery exchanges (without alarm, in response to alarm, in dim light, within carry bag), power supply change to AC power, driveline dis- and reconnection and controller exchange were simulated. Subjects’ gaze behavior was recorded using eye tracking (ET) technology (Tobii Pro Glasses 2, Tobii Pro Lab), whereas areas of interest (AOIs) were defined for LVAD components used. Completion rate, duration to success (DTS), pump-off time, and percental fixation duration per AOI were defined as outcome measures to identify differences between initially successful and unsuccessful subjects and to correlate subject performances with demographics. Results: Thirty gender and age matched, untrained subjects (15 HTX vs. 15 LP, median age: 64 (IQR=11) vs. 63 (10) yrs., p=0.84), 20 vs. 20% female, p=1.00) completed 210 scenarios, whereas 82.9 vs. 81.9% (p=1.00) could be solved at first attempt. Changing power supply from batteries to AC power revealed highest complexity for both groups (DTS=251±93s, p=0.76): only 26.7% of subjects succeeded at first attempt (33.3% vs. 20.0%, p=0.68) and 56.7% at second attempt (66.7% vs. 46.7%, p=0.04) with significantly more LP failing. The first attempt resulted in 10 unnecessary hazards from disconnecting the driveline (5 HTX vs. 5 LP, pump-off time 2-118 s, p=0.25) and significant differences in percental fixation durations for five AOIs in participants with vs. without initial success (p<0.029, see Figure). Significantly decreased DTS during battery exchange (no alarm 32.5 (18.8)s vs. alarm 26.5 (20.8)s and dim light 21.0 (19.0)s, p<0.001, without group differences, p>0.34) indicate a high level of learnability and routine establishment. Exchanging batteries in the bag was more time consuming (75.0 (45.0) s, p=0.09), especially in elderly patients (r=0.61, p<0.001). Conclusion: This ET based human factors study provides insights into user experiences of untrained former non-HM3 LVAD patients and LP in handling HM3 peripherals in simulated everyday and emergency scenarios. It highlights unintuitive and potentially hazardous handling characteristics and provides guidance for future user centered design of LVAD wearables.