Overview
The proposed study is a randomized controlled trial among an anicipated 180 participants with acute ischemic stroke and mod/severe OSA diagnosed by ambulatory polysomnography (PSG) comparing PAP treatment with usual care concerning the primary outcome of functional recovery.
Description
This study has 4 aims: 3 main aims and an additional exploratory aim.
Specific aim 1: Test whether PAP treatment for OSA in ischemic stroke improves stroke severity, post-stroke symptoms, post-stroke functional recovery (primary outcome), and quality of life.
Specific aim 2: Determine whether acute vs. subacute initiation of PAP in ischemic stroke results in greater improvement in post-stroke outcomes (stroke severity, symptoms, stroke recovery, quality of life).
Specific Aim 3: Conduct in-depth qualitative interviews with 25-30 stroke patients, family members, and acute care providers to elicit their experience with PAP use, including factors that influence adherence.
Exploratory aim: Evaluate whether polysomnographic measures have prognostic utility in understanding the variability in stroke recovery.
Eligibility
Inclusion Criteria:
- Having an acute ischemic stroke with brain imaging within 48 hours of symptoms onset;
- Being within 5 days of neurologic symptoms onset;
- Moderate (15 =< overall AHI < 30) /severe (overall AHI >= 30) OSA, thus with an obstructive apnea-hypopnea index AHI >= 15.
Exclusion Criteria:
- Past use of prescribed PAP for OSA;
- Suspected sleep disorder other than OSA (e.g., narcolepsy) (because such patients should be referred for a formal PSG in a sleep laboratory);
- Life expectancy is less than 6 months (e.g., hospice patients);
- Patients who require mechanical ventilation (because such patients could not participate in the intervention protocol);
- Non-English language patients (because the intervention strategy involves forming a relationship between the patient and research staff);
- Central sleep apnea with > 50% of respiratory events classified as central apnea;
- Resting oxygen saturation < 90%.
- Inability to provide their own informed consent. To enhance the generalizability of our study, all stroke severity will be included. However, we will exclude patients who cannot provide their own consent. This is because patients will need to participate actively in the protocol with a behavioral intervention. An assessment of the patient's competence to provide consent will be made based on published recommendations.