Ming Liu

Background/Objectives: Carotid artery stenosis is a well-established risk factor for vascular cognitive impairment. Although carotid endarterectomy (CEA) reduces the risk of ischemic stroke, its effect on cognitive function remains uncertain. Most previous studies have relied on mean changes in cognitive scores and have not defined postoperative cognitive decline using patient-level thresholds for the Montreal Cognitive Assessment (MoCA). Moreover, few studies have integrated perioperative variables across baseline, intraoperative, and postoperative domains to examine heterogeneity in cognitive outcomes. We therefore defined postoperative cognitive decline using threshold-based MoCA changes and prospectively evaluated perioperative factors associated with distinct cognitive outcomes after CEA. Methods: We prospectively enrolled older patients undergoing elective CEA under general anesthesia. Of 183 patients screened, 162 were ultimately enrolled and analyzed. All patients were managed according to standardized anesthetic and surgical protocols. Preoperative clinical, laboratory, and imaging variables, intraoperative hemodynamic parameters, and postoperative clinical indicators were collected. The primary outcome was cognitive decline at discharge, defined as a decrease of ≥2 points from baseline in the MoCA score. The secondary outcome was cognitive decline at 30 days, defined as a ≥3-point decrease in the telephone MoCA (t-MoCA) score relative to the corresponding baseline domains. Results: Among the 162 enrolled patients, preoperative triglyceride level, preoperative cerebral hypoperfusion volume, surgical laterality, intraoperative mean arterial pressure (MAP) decline, and visual analogue scale (VAS) on postoperative day 1 were independently associated with early postoperative cognitive decline. A prediction model incorporating these five factors achieved an area under the curve (AUC) of 0.752 (95% CI: 0.670–0.835) for cognitive decline before discharge. For 30-day cognitive decline, the model demonstrated an AUC of 0.717 (95% CI: 0.606–0.828). All patients with complete 30-day follow-up were stratified into four cognitive trajectory subtypes, and their longitudinal cognitive changes were visualized using a Sankey diagram. Conclusions: A perioperative model incorporating metabolic, cerebral perfusion, surgical, hemodynamic, and pain-related variables may help identify patients at increased risk of cognitive decline after CEA. The trajectory analysis further reveals heterogeneous postoperative cognitive patterns, including transient, delayed, and persistent decline. These findings support individualized perioperative risk assessment and structured cognitive follow-up. However, no internal or external validation was performed, and further testing is required in larger multicenter cohorts.

Ming Liu, Denggui Wang, Zhirong Yan, and Min Zhou. Korean J Pain -0001;0:. https://doi.org/10.3344/kjp.24386