[Comparison between automatic and manual analysis in the diagnosis of obstructive sleep apnea-hypopnea syndrome]

To compare automatic and manual analysis of neurological and respiratory variables obtained with the SomnoStar alpha 4100, a 16-channel polysomnographic system.

Twenty-eight patients suspected of obstructive sleep apnea-hypopnea syndrome were enrolled and given conventional polysomnographic tests. The order of automatic and manual reading of respiratory episodes, sleep stages, and arousals was randomized. We assessed agreement with the intraclass correlation coefficient and plotted standardized differences against standardized means, using the Bland-Altman method.

Poor agreement was observed between the 2 types of analysis of sleep stages, especially for REM and deep sleep stages. Agreement was good for apneic episodes among the respiratory variables; however, automatic analysis underestimated hypopneas. If manual analysis is considered the gold standard at the apnea-hypopnea index cut point greater than 10, automatic analysis obtained a sensitivity of 55%, a specificity and positive predictive value of 100%, a negative predictive value of 47%, and an overall diagnostic yield of 67.8%.

The automatic analysis of the SomnoStar 4100 system provides an unsatisfactory reading of sleep stages and respiratory episodes, especially hypopneas.