Related Experiment Video
Updated: Jul 13, 2026
Simultaneous Eye Tracking and Single-Neuron Recordings in Human Epilepsy Patients
Published on: June 17, 2019
Do Alarm Cues on ECG Monitors Enhance ICU Nurses' Detection of Abnormal Data? Evidence From an Eye-Tracking
Shumei Zhang1, Huiling Hu2, Hui Ge2
1Intensive Care Unit, Cancer Hospital Chinese Academy of Medical Science, Beijing, PR China.
Background:
Most modern electrocardiographic (ECG) monitors include visual alarm cues to flag abnormalities. Whether these cues consistently improve nurses' recognition of abnormal data remains uncertain.
Aim:
The primary aim is to examine the effects of alarm cues on ICU nurses' behavioural performance when identifying abnormal physiological parameters while using bedside vital-sign monitors. The secondary aim was to investigate how alarm cues influence nurses' visual attention patterns.
Study Design:
A within-subjects experiment with ICU nurses using a simulated monitor under two conditions (with vs. without alarm cues). Behavioural and eye-tracking metrics were recorded across heart rate, blood pressure, oxygen saturation, respiration and ECG.
Results:
Thirty-four ICU nurses between October 2024 and January 2025 completed the experiment. Alarm cues statistically significantly shortened reaction time (449.86 ± 134.67 s vs. 408.57 ± 140.41 s; p = 0.025). Changes in accuracy were selective: identification of abnormal ECG statistically significantly improved (median: 93% vs. 86%, p = 0.012), whereas accuracy for abnormal respiration statistically significantly decreased (mean: 78% vs. 88%, p < 0.001). No statistically significant differences were observed for heart rate, blood pressure or oxygen saturation. Eye-tracking results showed that attention was drawn to cued regions, with statistically significantly longer fixation durations (ECG: 44.42 ± 27.51 s vs. 33.98 ± 25.27 s, p = 0.004; blood pressure: median 27.66 vs. 20.41 s, p = 0.025) and statistically significantly fewer visits to uncued areas (105.94 ± 34.87 vs. 89.82 ± 35.51, p = 0.003), indicating a more focused but less comprehensive search.
Conclusion:
Visual alarm cues on ICU monitoring interfaces can effectively enhance attentional focus and accelerate decision-making but do not uniformly improve recognition accuracy across physiological parameters. These results highlight the need for optimising alarm interface design to balance efficiency and situational awareness.
Relevance To Clinical Practice:
Alarm systems should be parameter-specific and adaptive, adjusting cue salience to data context to minimise attentional tunnelling. Co-locating cues with the corresponding numeric/waveform fields and using graded, multi-layered indicators (rather than a single global alert) can support rapid yet comprehensive scanning. Brief scan-path prompts or training that requires verification of non-cued parameters may mitigate missed abnormalities. These strategies can help nurses sustain fast responses without sacrificing holistic monitoring.
