Validity of a portable cardio-respiratory system to collect data in the home environment in patients with obstructive sleep apnea

As part of the development of a portable cardio-respiratory system for detection of sleep-disordered breathing in the home, data gathered from the LifeShirt® (LS) system were compared to that from traditional polysomnography (PSG) in a laboratory environment and then again in the home to ascertain (1) the degree of concordance between the two and (2) to verify whether or not reliable data could be collected outside the laboratory. Ten subjects were recorded for one night each during three different conditions, PSG in the lab (PSG), LS and PSG in the lab (LS-L), and LS at home (LS-H). Total sleep time (p=0.097), time in stage 1+2 sleep (p=0.245), time in stage 3+4 sleep (p=0.633), REM sleep time (p=0.157), and total awakenings (p=0.364) were not different between PSG and LS-L. No significant difference between the apnea/hypopnea index (AHI) as determined by PSG (28.2+21.1/hr) LS-L (30.8+16.3/hr) and the LS-H (27.3+21.4/hr.) was observed. Oximetry data were not significantly different between the two devices (mean: PSG, 93±5%; LS-L, 92±3%; and LS-H, 92±3%; p=0.41). There was a strong linear relationship between PSG and LS-H and PSG and LS-L for AHI (r=0.96; p<0.0001 and r=0.82; p=0.004, respectively). Agreement for AHI was determined using the method of Bland-Altman (bias=0.848, SD=6.41), as well as the concordance correlation coefficient (ρ_c=0.96). Sensitivity and specificity for detection of OSA were high, but varied slightly with the threshold definition used. For an AHI of >5/hr, sensitivity and specificity were both 100%; for an AHI>15/hr, they were 87.5% and 100%, respectively. Conclusions: In summary, a high degree of concordance between LS and traditional PSG was observed, suggesting that LS may be a viable option for the physicians to consider for home detection of OSA.