How reliable is the manual correction of the autoscoring of a level IV sleep study (ApneaLink™) by an observer without experience in polysomnography?

Autor: Nigro, Carlos, Malnis, Silvana, Dibur, Eduardo, Rhodius, Edgardo
Zdroj: Sleep & Breathing; Jun2012, Vol. 16 Issue 2, p275-279, 5p
Abstrakt: Objective: This study aims to compare the manual correction of the automatic analysis of ApneaLink™ between a skilled observer in the interpretation of sleep studies and a subject trained only in the scoring of ApneaLink™ device. Methods: Ninety-six subjects performed the ApneaLink™ and polysomnography (PSG) simultaneously in the sleep laboratory. Two blind observers, who were independent from the results of the PSG, performed first the automatic scoring and then the hand correction from the ApneaLink™ device. The scorers of ApneaLink™ represented two physicians with different levels of training (scorer A: 20 years of experience in reading polysomnography plus 3 years of experience in the interpretation of ApneaLink™, scorer B: 1 year of experience in the analysis of ApneaLink™). The interobserver agreement was assessed with the intraclass correlation coefficient (ICC) and kappa statistics. The diagnostic accuracy of the manual analysis ApneaLink™ device was evaluated by the area under the receiver operator curve (AUC-ROC). Results: Ninety patients were included (69 men; mean age, 49.6; median RDI, 13.9; median BMI, 29.3 Kg/m). The ICC between the manual apnea/hypopnea index from ApneaLink™ and the respiratory disturbance index of the PSG for each observer was similar (scorer A, 0.902; CI 95% 0.80-0.95; vs. scorer B, 0.904; CI 95% 0.86-0.94; p = 0.9). The agreement between the observers on the presence or absence of obstructive sleep apnea syndrome (OSAS) was very good (kappa, 0.83; CI 95% 0.69-0.98). The AUC-ROC was similar between the observers (scorer A, 0.88; CI 95% 0.78-0.98; scorer B, 0.83; CI 95% 0.71-0.95; p = 0.5). Conclusions: The non-expert observer showed a very good agreement with the expert observer on the results of the manual correction of the ApneaLink™ autoscoring. Both observers had similar diagnostic accuracy to identify subjects with OSAS when compared with PSG. [ABSTRACT FROM AUTHOR]
Databáze: Complementary Index
Externí odkaz: