Authors: Kang M et al.
Journal: Anesthesiology, December 4, 2025. DOI: 10.1097/ALN.0000000000005875
Summary
This prospective observational study evaluated whether preoperative saccadic eye movement tasks can serve as a noninvasive and accurate predictor of postoperative delirium (POD) in elderly patients undergoing arthroplasty. Given the limitations of traditional cognitive screening tools and invasive biomarker testing, the authors sought to determine whether oculomotor function—reflecting frontoparietal and subcortical network integrity—could improve preoperative risk stratification.
The study enrolled 316 patients aged 65 years or older undergoing elective arthroplasty. Preoperative assessments included standard cognitive tests (MMSE and MoCA), serum and cerebrospinal fluid biomarkers (including neurofilament light chain), and detailed saccadic task measurements. POD was diagnosed postoperatively using the Confusion Assessment Method by trained evaluators.
Postoperative delirium occurred in 8.2% of patients. Several saccadic parameters differed significantly between patients who developed POD and those who did not, including reaction time, saccadic accuracy, gain abnormalities, and peak velocities across multiple saccade paradigms. While cognitive scales and serum biomarkers showed statistically significant group differences, their predictive performance was modest.
In contrast, saccadic task–based models demonstrated substantially superior predictive accuracy. Logistic regression models using saccadic parameters outperformed MMSE, MoCA, and neurofilament light chain levels. Machine learning models further improved performance, achieving high discrimination for POD prediction. These findings suggest that subtle deficits in executive and attentional networks detectable through eye movement analysis precede clinically overt delirium.
Overall, the study positions saccadic eye movement testing as a practical, noninvasive behavioral biomarker that may enhance POD risk prediction when used alongside existing assessments.
Key Points
-
Postoperative delirium occurred in 8.2% of elderly arthroplasty patients.
-
Multiple saccadic eye movement parameters differed between POD and non-POD patients.
-
Saccadic testing predicted POD more accurately than MMSE, MoCA, or neurofilament light chain levels.
-
Machine learning models using saccadic parameters achieved high predictive accuracy.
-
Saccadic tasks provide a fast, noninvasive assessment of executive and attentional network integrity.
-
Saccadic eye movement testing may serve as a useful adjunct for preoperative delirium risk stratification.
Thank you for allowing us to use this article from Anesthesiology.