Cognitive Impairment after Aneurysmal Subarachnoid Hemorrhage: Temporal Patterns and Electrophysiological Risk Stratification.

Abstract

Cognitive impairment (CI) after aneurysmal subarachnoid hemorrhage (aSAH) is common, but its temporal course and prediction using multimodal data remain unclear. This prospective cohort study assessed time to first-detected CI among survivors eligible for follow-up cognitive testing. CI was defined as an education-adjusted Montreal Cognitive Assessment score <26 at 7 to 24 months after admission. A sensitivity endpoint using MoCA <23 was also examined. First-detected CI-free survival was compared between patients with aSAH and those with unruptured intracranial aneurysms (UIAs). Within the aSAH cohort, six prediction schemes were tested, including core clinical variables alone and clinical variables plus baseline cognition, plasma biomarkers, event-related potentials (ERPs), MRI, or CSF markers. Patients with aSAH had lower 24-month first-detected CI-free survival than patients with UIA, 43.7% vs. 71.4%, with an 8.1-month loss in restricted mean survival time, 95% CI 6.2 to 9.6. They also had a higher hazard of first-detected CI, HR 7.21, 95% CI 3.59 to 14.49. Among the six prediction schemes, only the ERP CP2 pair model met all prespecified performance criteria. In temporal internal validation, this model achieved an AUROC of 0.928 and a Brier score of 0.103. Incremental value was greatest in patients aged ≥60 years and those with Fisher grade 3 to 4. aSAH was associated with earlier CI and substantial loss of CI-free time compared with UIA. An ERP-based model showed the most stable internally validated performance, especially in older patients and those with greater hemorrhagic burden. External multicenter validation is needed before clinical use.