The proposed research plan seeks to understand the impact of sleep disruption in the Neurological Intensive Care Unit (ICU) on older patients with acute brain injury (ABI). In current practice, the neurocritical care community performs frequent serial neurological examinations ("neurochecks") in an effort to monitor patients for neurological deterioration following brain injury. Many neurocritical patients are older and/or cognitively fragile, and delirium is common. Although ICU delirium is multifaceted, frequent neurochecks may represent a modifiable risk factor if the investigators can better understand the risks and benefits of various neurocheck frequencies. This project will randomize patients with acute spontaneous intracerebral hemorrhage (ICH) to either hourly (Q1) or every-other-hour (Q2) neurochecks and evaluate the impact of neurocheck frequency on delirium. Second, longer-term cognitive outcomes will be investigated in patients with ICH randomized to Q1 versus Q2 neurochecks with the goal of identifying whether hourly neurochecks increase the risk for dementia.

Usual care: Patients with ICH are cared for in the NeuroICU by a specialized team including Board-certified neurointensivists, nurse practitioners, and neuro-trained bedside nurses (neuroRNs). These neuroRNs each undergo intensive specialized training on the neurological exam, diagnosis, and acute management of neuro-complications, and maintain their knowledge via regular audits and through annual continuing education. The CAM-ICU tool is the most well-known and robustly utilized tool to assess ICU delirium, and has been validated in the poststroke population. The investigators have tight adherence to detailed protocols for multidisciplinary management of potential confounders including coagulopathy, cerebral edema, blood pressure (goal 130-150mmHg systolic), nutrition, glucose control, and early mobility. Pain management utilizes Tylenol and short-acting opiates in non-intubated patients. In intubated patients, our goal RASS (Richmond Agitation Sedation Score) is 0 to -2 and the investigators prioritize analgosedation with fentanyl, propofol and/or dexmedetomidine.

This intervention: Patients will be randomized (random number generator) to Q1 or Q2 neurochecks, which will be performed by the bedside expert neuroRNs. Once randomized, the investigators will monitor patients during their ICU admission with data prospectively collected to include: patient and clinical demographic information (e.g., age, severity and location of ICH), pre-admission sleep quality (via Pittsburgh Sleep Quality Index), medical complications during ICU admission, incident delirium per ICU stay (as measured by CAM-ICU screening tool performed by the bedside nurses each shift; yes/no) as well as time elapsed in the hospital prior to developing delirium, frequency of neurochecks at time of incident delirium, ICU and hospital lengths of stay (LOS), consecutive and total duration of delirium (as measured by CAM-ICU), discharge destination (including death), techniques used to manage delirium (e.g., antipsychotic medication administration with total dosages), and pharmacotherapy for pain (e.g., opiate frequency and dosing). Safety data (i.e., mortality, upgrade in frequency of neurochecks, adverse events) will be collected throughout the study (see details in DSMP). Given a reported association between ApoE allele and early onset delirium and delirium duration, all enrolled patients will undergo saliva sampling (4 buccals swabs) for ApoE allele status.

Long-term outcomes:

Patients enrolled in will be followed longitudinally for longer-term neuropsychological assessment at 6-months post-discharge. Assessment will be performed using the comprehensive cognitive and emotional batteries through NIH Toolbox® along with supplemental cognitive measures (e.g., Auditory Verbal Learning, AVLT) and additional mood/emotional screening questionnaires (e.g., HAM-A, IES-R). The NIH Toolbox® is chosen because of its validation in neurological patients with stroke and relative resilience against learning effects. By giving a wide range of verbal and nonverbal tests and using typical methods of cognitive assessment (e.g., delayed AVLT), the investigators can address our outcomes independent of verbal dysfunction from dominant hemisphere lesions and over time as a function of domain-specific impairments. This aim will also capture retention statistics and participant impressions/willingness to participate after 6 months.

Inclusion Criteria (Part 1):

1. Adult patients (age >18 years) with spontaneous acute <45cc in volume with radiographic and clinical stability for ≥6 hours following admission to the ICU. These criteria are based on the literature and experience of the investigative team.
2. Additional intraventricular hemorrhage (with or without external ventricular drain) is allowable.
3. Only first admission to the NeuroICU during the hospitalization will be eligible.

Inclusion criteria (Part 2):

a. any patient included in part 1 alive at 6 months post-discharge

Exclusion Criteria:

1. Patients with unstable intracranial bleeding
2. Patients with known history of intracranial neurological injury
3. Pre-existing cognitive impairment (known or highly suspected based on family-provided history, Activities of Daily Living Questionnaire)
4. Pre-existing diagnosed sleep disorder
5. Comatose or heavily sedated
6. Death expected within 30 days or other terminal illness
7. ICH score >4 (equivalent to mortality risk >72%)
8. Pregnancy
9. Incarcerated
10. Non-English or non-Spanish speaking

Per NIH guidelines, will provide anonymized data to qualified investigator

No sooner than at the time of study closure

The full de-identified dataset will be made available for sharing utilizing a repository. All data sharing practices are scrutinized for HIPAA compliance and other best practices around data use agreements (DUAs). Access to databases and associated software tools generated under the project will be available for educational, research and non-profit purposes.