Abstract Body: Introduction: The advent of wireless monitoring technologies to detect stroke onset during sleep has the potential to accelerate reperfusion treatment in the 20-25% of patients with wake-up stroke. Monitoring technologies will hasten stroke detection in patients who sleep through stroke onset but not in patients who are awakened by stroke onset. The relative frequencies patients who stay asleep after onset compared with patients who wake up with onset has not been well
delineated.
Methods: We performed a systematic search and meta-analysis of studies comparing diffusion-weighted imaging–fluid-attenuated inversion recovery (DWI-FLAIR) mismatch frequencies in wake-up stroke (WUS) compared with clear onset time stroke (COS). Differences in WUS vs COS decay curves in frequency of FLAIR-DWI mismatch with longer time from symptom detection to imaging were used to estimate the time of stroke onset during sleep.
Results: Two studies of WUS and COS patients met selection criteria, collectively enrolling 443 patients (201 WUS, 242 COS). The frequencies of FLAIR-DWI mismatch at different intervals from symptom-detection-to-imaging are shown in Figure 1. For COS patients, frequency of mismatch presence declined gradually: <2h 60%; 2-3h 53%; 3-4h 41%; 4-5h 37%; 5-6h 11%. In contrast, for WUS patients, <2h patients had similar mismatch rates to <2h COS patients but
mismatch frequency then declined precipitously in all later time windows: <2h 50%; 2-3h13%; 3-4h 13%; 4-5h 6%; 5-6h 12%. Among the WUS patients: 16.9% had detection-to-imaging time in the <2h time period in which mismatch was nearly equally present as in COS patients; 70.1% had detection-to-imaging times with mismatch rates whose COS equivalents indicated stroke onset 1-8h prior to awakening; and 13.0% had detection-to-imaging times of indeterminate
stroke onset time.
Conclusions: About 5 of every 6 patients with stroke first detected on awakening had remained asleep for 1-8 hours after stroke onset, while 1 of every 6 awoke at the time of, or soon after, onset. These findings indicate that wireless monitoring technologies to detect stroke start during sleep would accelerate stroke detection in many WUS patients.