To assess whether eye movements add prognostic value beyond malignant EEG features and characterize the relationship between eye movements, malignant EEG features, and discharge outcomes in cardiac arrest (CA) patients.

Neurological monitoring of CA patients is essential for evaluating recovery potential. Eye movement quantification via EEG may serve as a novel noninvasive biomarker indicating preservation of neural pathways for consciousness recovery. Characterizing eye movement relationships with other prognostic markers can clarify their contribution to comatose CA assessment.

We conducted a retrospective cohort study of 48 CA patients who underwent continuous EEG/EOG monitoring post-arrest at Boston Medical Center (2020-2023). Team members quantified eye movements (that were normalized per hour) and malignant EEG features, including absence of reactivity, burst suppression, periodic discharge, polyspike waves, or discontinuous patterns. Among discharged patients, outcomes included corneal reflex, spontaneous or stimulus-induced eye opening, and command following. Logistic regression models were adjusted for age and sex, with a likelihood ratio test assessing the impact of eye movements on model fit.

Among 48 patients (65% male, median age 60), 19% had malignant features, and 31% were deceased at discharge. Preliminary analyses reveal a statistically significant negative association between eye movements and death and statistically significant positive associations between eye movements and discharge outcomes. Including eye movements improved model fit for death (LR 11.44, p=0.022), pupillary reflex (LR 8.58, p=0.072), corneal reflex (LR 12.36, p=0.015), eye-opening to stimuli (LR 9.51, p=0.049), spontaneous eye opening (LR 12.22, p=0.016), and command following (LR 13.05, p=0.011).

Incorporating eye movements in predictive models may enhance assessment of neurological outcomes in CA patients, positioning eye movements as a potential non-invasive biomarker for neuroprognostication in the ICU.