Automatic scoring had unacceptable results. PSG scoring used arousals. Data loss under 10%. Evidence grade IV since blinding of scoring not reported.
Scoring same for apneas and hypopneas for PSG and PM without arousals. No prevalence data for AHI > 5. Oximetry sampling rates not given. Time In bed used for PM RDI was 35% longer than total sleep time used for AHI.
AHI of 10 gives best pair of NPV and PPV. Patients with heart failure. Arousals Included In AHI for PSG. Oximeter sampling rate of five seconds on PM and PSG.
Scoring same for apneas and hypopneas for PSG and PM without arousals. Oximeter sampling rate not disclosed. AHI of five gave best PPV (89.1%) due to high prevalence although high LR was minimally Increased at 1.63.
Note-. Data obtained with portable monitor simultaneous with polysomnography. Apnea/hypopnea Index per hour of sleep with polysomnography. Respiratory disturbance Index, apnea/hypopnea Index per hour of recording unless otherwise Indicated for portable monitor. "Cannot be calculated due to division by 0.
Abbreviations: AHI, apnea/hypopnea Index; corr, corresponding sensitivity or specificity when best sensitivity and specificity at different RDI or AHI thresholds (when this occurs, there are a number of nondiagnostic tests); est, estimate; Evld, evidence grade; H, high; LR, likelihood ratio; L, low; NPV, negative predictive value = true negatives/true negatives plus false negatives (%); Non (%), percent of nondiagnostic tests; PM, portable monitor; PSG, attended polysomnography; Prev (%), prevalence In percent; PPV, positive predictive value = true positives/true positives plus false positives (%); red, reduction; RDI, respiratory disturbance Index; Sens, sensitivity (%); Spec, specificity (%); sat, arterial oxygen desatu ration.
Was this article helpful?