378

Abstract

Introduction: Delirium in medical and surgical ICU patients is a predictor of worse clinical outcomes, but there are few data concerning neurological ICU patients. Hypothesis: We tested the hypothesis that delirium is associated with length of stay, functional outcomes and domain-specific quality of life in patients with acute intraparenchymal hemorrhage (IPH) and subarachnoid hemorrhage (SAH). Methods: This is a prospective, observational cohort study. Delirium was routinely assessed twice daily using the Confusion Assessment Method for the ICU (CAM-ICU) by trained staff. Delirium was defined as a change from the “new baseline” of neurologic status at hospital admission as previously published. Functional outcomes were recorded with modified Rankin Scale (mRS, scored from 0, no symptoms to 6, dead), and quality of life (QOL) outcomes were assessed with NIH Patient Reported Outcomes Measurement Information System (PROMIS) and Neuro-QOL at 28 days, 3 months and one year. Results: We enrolled a total 143 patients, 84 with IPH and 59 with SAH. Sixteen patients were not assessable due to persistent coma. Of 127 remaining, 41 (32%) had delirium. ICU and hospital length of stay, when considered as time-dependent covariates, were associated with delirium (P<0.001) after correction for Glasgow Coma Scale (GCS), any benzodiazepine use, and admitting diagnosis. In patients with SAH, levetiracetam was associated with delirium, although seizures were rare. Delirium was associated with worse mRS at 28 days (4 [3-5] vs. 3 [1-5], P=0.008) and 3 months (3 [2-5.75] vs. 2 [0-4], P=0.004). In mixed models, delirium was associated with worse QOL scores in the domains of applied cognition – executive function and general concerns after adjustment for age, admission diagnosis, point of follow-up and GCS. Conclusions: Delirium was common after SAH and IPH and predictive of subsequent worse functional and QOL outcomes (especially applied cognition) although benzodiazepine use and infection were uncommon. Further study should determine if the underlying risk factors for delirium are modifiable, and this may lead to better outcomes.