Optimization Strategies to Reduce Alarm Fatigue in Patient Monitors

Abstract

In order to ameliorate alarm fatigue, three optimization strategies are proposed to reduce false alarms and repetitive non-actionable true alarms. The four-lead arrhythmia analysis, multi-parameter fusion and intelligent threshold reminder are adopted and evaluated in multi-center clinic study. The four-lead arrhythmia analysis algorithm includes lead optimization, beat matching, detection and classification combinations. The multi-parameter fusion algorithm aggregates the information obtained from ECG, SpO2 and IBP wave signals. The intelligent threshold reminder can help medical staff to adjust and recover alarm limits appropriately. The results show that more than 50% of false alarms can be reduced by the four-lead analysis and the multi-parameter fusion analysis. In some specific occasions, the intelligent threshold reminder can reduce repetitive non-actionable true alarms significantly. And no further false-negative events are generated after using the strategies in our experiments. We demonstrate that increasing the dimensionality of parametric analysis and controlling the alarm limits is beneficial for reducing alarm fatigue in intensive care units.