JMIR Publications

ABSTRACT

Background:

Patient monitoring is central to the safety of state-of-the-art perioperative and intensive care medicine. While current state-of-the-art patient monitors display vital signs in the form of numbers and curve forms, Visual Patient technology creates an easy to interpret virtual patient avatar model, which, in a previous study, enabled anesthesia providers to perceive more vital sign information during short glances than conventional monitoring. In this study, we used eye-tracking technology to study the deeper mechanisms underlying information perception in both, conventional and avatar-based patient monitoring.

Objective:

In this study, we used eye-tracking technology to study the deeper mechanisms underlying information perception in both, conventional and avatar-based patient monitoring.

Methods:

In this prospective, multi-center study with a within subject design, we showed 32 anesthesia experts (physicians and nurse anesthetists) a total of four 3- and 10-second monitoring scenarios alternatingly as either routine conventional or avatar-based monitoring in random sequence. All participants observed the same scenarios with both monitoring technologies. After each scenario, we asked participants to report the status of the vital signs. Using an eye-tracker, we recorded the participants’ gaze paths as they observed the scenarios. From the eye-tracking recordings, we evaluated which vital signs the participants had visually fixated, how often and for how long during a scenario, and therefore, could have potentially read and perceived this vital sign. We compared the frequencies and durations with which the participants had visually fixated the vital signs between the two monitoring technologies.

Results:

Participants visually fixated more vital signs per scenario, median (IQR): 10 (9-11) vs. 6 (4-8), p<0.001 in avatar-based monitoring (median of differences: 3 vital signs (95% confidence interval [95%CI 3-4]). In a multivariable linear regression, monitoring technology (conventional vs. avatar-based monitoring, difference -3.3, p<0.001) was an independent predictor of the number of visually fixated vital signs. Only scenario duration affected the difference in vital sign fixations between technologies. The difference was more prominent in 3-second scenarios, difference -1.5, p=0.04. Study center, profession, gender, and scenario order did not influence the differences between conventional and avatar-based monitoring. In all four scenarios, the participants visually fixated nine of the 11 total vital signs shown statistically significantly longer using the avatar (all p < 0.001). Four critical vital signs, i.e., pulse rate, blood pressure, oxygen saturation, and respiratory rate were visible almost the entire time of a scenario with avatar-based monitoring, while with conventional monitoring, these were only visible for fractions of the observations. Visual fixation of a certain vital sign was associated with the correct perception of that certain vital sign in both technologies. Phi coefficient for avatar: 0.358, for conventional monitoring: 0.515, both p<0.001.

Conclusions:

This study uncovered, by use of eye-tracking, one of the mechanisms responsible for the improved information transfer in avatar-based monitoring. The design of the avatar technology, which presents the information about multiple vital signs integrated into forms and colors of the corresponding anatomical parts of a patient model enabled parallel perception of multiple vital signs, and thereby increased the number of visually fixated vital signs and the time available to fixate each vital sign. With this finding confirmed by eye-tracking, this study adds a new and higher level of empirical evidence as to why avatar-based monitoring improves the perception of vital sign information compared to conventional monitoring.