Validation of Sleep Healthcare System Clinical Trial
Official title:
Validation of Sleep Healthcare System
Sleep quality affect working and learning performance; poor quality of sleep is one of the
common problems of modern people. Traditionally, polysomnography is a recognized standard for
sleep quality assessment. Subjects are put adhesive electrodes, chest and abdomen band,
oximetery, and oronasal cannula and stay in certified sleep laboratory for monitoring. These
sensors setup are cumbersome and be likely to induce discomfort. An alternative to assess the
quality of sleep is actigraphy, which allows users to wear for more than two weeks. In recent
years, many of the smart watches, which often measure wrist photoplethysmography (PPG) signal
and body movement, are prevailing to make long-term sleep monitoring feasible, but its
accuracy and effectiveness still need to be verified.
Obstructive sleep apnea (OSA) is a common disorder characterized by intermittent hypoxia and
sleep fragmentation. OSA is associated with cardiovascular morbidity and mortality, metabolic
dysregulation, and neurocognitive dysfunction, which results in the negative impact on
prognosis. PSG is the gold standard for OSA diagnosis which is expensive and less accessible.
Therefore, modality other than PSG is necessary to speed up diagnosis and treatment. Center
of Sleep Disorder in National Taiwan University Hospital has been operated since June 2006.
Up to Dec.2015, totally 8,819 patients have been referred for sleep studies (NTUH cohort)
where 1,435 patients are under long-term CPAP and 396 patients are under MAD. Using data from
4,618 patients in NTUH cohort, we have already established an OSA prediction mode
(apnea-hypopnea index, AHI≥5/hr) with accuracy 82.37% (sensitivity 87.03%, positive
predictive value 91%). Regarding the molecular mechanism, our previous study showed that by
plasma metabolomics profiling, we could identify candidate metabolites associated with OSA
severity. The 11 candidate metabolites were identified by comparing profiling in 100 patients
with AHI <15/hr and with AHI>=15/hr, respectively. Six identified metabolites were selected
to establish an AHI prediction model which gave sensitivity 66%, specificity 72%, and AUROC
0.736. Furthermore, 15 plasma metabolites associated with excessive daytime sleepiness (EDS)
or polysomnographic parameters were identified. Among those metabolites, L-Kynurenine and
g-Glutamylleucine were metabolites associated with EDS which generated the AUROC to EDS
prediction as 63% in study group and 76.7% in validation group. The online system (Good
Sleep) for diagnosis of sleep disorder has been set up under the collaboration between, NTU,
NTUH, and MediaTek. It aims on population with low probability of sleep disorder which
compliments the NTUH cohort, high probability of sleep disorder. The online system provides
the diagnosis and solution of sleep disorder, sleep tracking, and education via both website
and App. The system is almost set and needs the input from general population to validate the
accuracy.
The sleep healthcare system, which includes questionnaires, smart watches,24-hr BP and
＂LARGAN＂ECG Holter for long-term home sleep monitoring, is proposed to allow users to detect
potential subjects who have sleep disorders by filling out the questionnaire. The aims of the
present project include: (1) All 300 voluntary.Stage1, Recruit 140 voluntary participants
from MediaTek to validate agreement of sleep efficiency via online system, actigraph devices,
smart watches and daily blood pressure for one week.Stage2, Recruit 160 voluntary
participants from patients with moderate-severe OSA (AHI≥15/hr) to validate agreement of
sleep efficiency via online system, actigraph devices, smart watches, ECG Holter and 24 hour
blood pressure for one day. (2) All participants will take an overnight PSG test, blood
sampling, basal metabolism measurement, ECG Holter, body composition and E-Prime at the sleep
center to validate the performance of online system on diagnosis of OSA in low risk
population. (3) Analyze the of PSG parameters in both low and high risk population (to build
up the out of center devices for OSA home testing). (4) Integrate the clinical parameters and
plasma metabolic profile, before and after treatment, to identify factors associated with OSA
related sequels and long-term prognosis.
n/a