Abstract Background Automated infrared pupillometry (AIP) provides a quantitative, non-invasive, and reproducible method for assessing the pupillary light reflex, which is mediated by brainstem pathways. Although AIP has been increasingly used in critical care settings, interpretation of pupillary parameters may be confounded by the administration of sedatives and analgesics. However, the influence of these medications on individual AIP parameters in postoperative intensive care unit patients remains insufficiently characterized. This proof-of-concept pilot study aimed to identify pupillometric parameters that remain relatively stable under routine sedative and analgesic exposure following elective cardiovascular surgery. Methods This prospective observational study included 10 patients admitted to the coronary care unit (CCU) after elective open-heart surgery. AIP parameters and bispectral index (BIS) values were measured postoperatively at predefined time points from CCU admission to extubation. Temporal changes in each parameter were analyzed descriptively using within-patient comparisons over the postoperative course. Results The Neurological Pupil Index (NPi) and latency (LAT) showed minimal temporal variation throughout the observation period. In contrast, BIS values increased progressively after surgery, accompanied by gradual increases in constriction percentage (CH), mean contraction velocity (CV), and maximum contraction velocity, whereas mean dilation velocity remained relatively unchanged over time. When comparing predefined postoperative time points immediately after CCU admission and after extubation, only CH showed an increase (median interquartile range: 0.45 0.38–0.52 vs. 0.68 0.60–0.73; p = 0.02). Conclusions In this small postoperative cohort, NPi and LAT were relatively more stable than other pupillometric parameters under routine sedative and analgesic exposure. However, given the limited sample size and the absence of neurological complications, these findings should be interpreted as descriptive and hypothesis-generating, and no conclusions can be drawn regarding clinical utility for neurological monitoring or prognostication. In contrast, CH and CV require cautious interpretation, as they may be influenced by pharmacological agents rather than by the underlying neurological status.
Okamoto et al. (Mon,) studied this question.