Abstract
Abstract
Introduction
The WatchPAT®️ is an innovative Home Sleep Apnea Device (HSAT) that utilizes the peripheral arterial signal (PAT®️) for OSA diagnosis. We examined the diagnostic validity of the WatchPAT and assessed the correlation between its sleep indices and those measured by PSG in a cognitive normal predominantly older Black/African-American sample.
Methods
Preliminary data analysis on a limited sample of 26 participants without a prior diagnosis of OSA who underwent HSAT and 2-nights of nocturnal polysomnography (NPSG). An apnea-hypopnea index (AHI) ≥ 15 events/h characterized moderate to severe OSA. Pearson correlation statistics (Fisher's z Transformation) for PAT/NPSG indices including respiratory disturbance index (RDI), oxygen desaturation index (ODI), mean oxygen saturation (Spo2), REM Latency (min), awake period during sleep (WASO), and sleep onset latency (SOL) were determined.
Results
Of the 26 participants, 17 (65.4%) were Black/African-American, 9 (34.6%) were non-Hispanic White, and 19 (73.1% [13/19 (68.4%) Black/African-American]) were female. Mean (SD) age, BMI and education was 66.3 (4.5) vs. 69.9 (3.7) years, 29.6 (6.4) vs. 26.6 (5.5) kg/m2 and 15.6 (3.1) vs. 17.6 (1.3) years for Black/African-American vs. non-Hispanic White, respectively. 35.3% vs. 33.3% and 17.7% vs. 22.2% of participants met criteria for moderate to severe OSA based on HSAT and NPSG, for Black/African-American vs. non-Hispanic White, respectively. The HSAT had a sensitivity and negative predictive value of 100% for both races, specificity of 78.6% vs. 85.7%, and positive predictive value (PPV) of 50% vs. 66.7%, for Black/African-American vs. non-Hispanic White, respectively. Analyses stratified by sex suggested that the WatchPAT had better diagnostic validity in Black/African-American women than men, with specificity of 83.3% vs. 75.0%. Among Blacks/African-Americans, the correlation for PAT/NPSG AHI, ODI and RDI were modest ranging from r = 0.65 to 0.70 P = .004, and mild for Spo2 and Nadir oxygen desaturation (r = 0.53 [95%CI, 0.04-0.79]; P = .03 for both). WASO, REM Latency and SOL showed no significant correlations.
Conclusion
Our preliminary data show HSAT having lower specificity, PPV, insights on sleep architecture for OSA diagnosis, and respiratory indices’ correlations with those from PSG, in Blacks/African-Americans compared to general population samples. The measure performed better among women.
Support (If Any)
AASM 231-BS-20, AARGD-21-8488397, NIH/NIA/NHLBI (K23AG068534, L30-AG064670, CIRAD P30AG059303 Pilot, NYU ADRC P30AG066512 Developmental Grant, R25HL105444 SRG, R01AG12101, R01AG022374, R01AG13616, RF1AG057570, R01HL118624, R01AG056031)