Using Eye-Tracking Technology to Capture the Visual Attention of Nurses During Interpretation of Patient Monitoring Scenarios from a Computer Simulated Bedside Monitor

Introduction: This study analysed the utility of eye tracking technology for gaining insight into the decision making processes of nurses during their interpretation of patient scenarios and vital signs. Methods: Five patient monitoring scenarios (vignette, vital signs [ECG, BP etc.] and scoring cri...

Full description

Saved in:
Bibliographic Details
Published in:Journal of electrocardiology 2016-11, Vol.49 (6), p.927-927
Main Authors: Currie, Jonathan, Bond, Raymond R, McCullagh, Paul, Black, Pauline, Finlay, Dewar D, Peace, Aaron
Format: Article
Language:English
Subjects:
Cardiovascular
Computer simulation
Decision making
Monitoring systems
Nurses
Patients
Technology
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Introduction: This study analysed the utility of eye tracking technology for gaining insight into the decision making processes of nurses during their interpretation of patient scenarios and vital signs. Methods: Five patient monitoring scenarios (vignette, vital signs [ECG, BP etc.] and scoring criteria) were designed and validated by critical care experts. Participants were asked to interpret these scenarios whilst ‘thinking aloud’. Visual attention was measured using infrared light-based eye-tracking technology. Each interpretation was scored out of 10. Subjects comprised of students (n = 36) and qualified nurses (n = 11). Scores and self-rated confidence (where 1 = low, 10 = high) are presented using mean ± SD. Significance testing was performed using a t-test and ANOVA where appropriate (α = 0.05). Multivariate regression was performed to determine if a machine could use eye gaze features to accurately predict competency (dependent variable = score). Independent eye gaze only variables were used in the regression models if they statistically significantly (p b 0.05) correlated with the score. Results: Scores across all scenarios were calculated (students=4.58±1.13 vs. qualified=6.85±0.82) with statistical significance between groups (p = b0.01). Mean self-rated confidence was also calculated (students = 5.79 ± 1.05 vs. qualified = 7.49 ± 1.00, p=b0.01). There was a weak positive correlation between confidence and score amongst students (r = 0.323, p = 0.06), although no meaningful correlation with qualified nurses (r = −0.099, p = 0.77). However, for all participants there was a moderate correlation between confidence and score (r = 0.592, p = b0.01). The fitness of the regression models for predicting competency based on eye gaze features only is as follows: • Scenario 1: R2 = 0.407, Std error = 1.243 (p = 0.09) • Scenario 2: R2 = 0.746, Std error = 1.439 (p = 0.01) • Scenario 3: R2 = 0.385, Std error = 1.564 (p = 0.03) • Scenario 4: R2 =0.687, Std error = 1.340 (p = 0.44) • Scenario 5: R2 = 0.766, Std error = 0.960 (p = 0.02) The following table also shows where subjects fixated the most and least on the different vital signs on the bedside monitor (note the lower fixation duration on the ECG by students in comparison to qualified nurses). Conclusion: The study has shown that eye-tracking measurements can provide insight into the decision-making of nurses and can be used to predict competency.
ISSN:0022-0736
1532-8430
DOI:10.1016/j.jelectrocard.2016.09.017