Auto report of a complete PSG, will be credible your result?

Introduction:In the last years have seen the development of systems to record polysomnography(PSG)with the ability to self-diagnose.The application of the international rules for the manual scoring is lengthy, expensive and are subject to inter-and intra-operator variability.However, it is the knowledge of sleep labs professionals these systems have many flaws and that the result has little credibility. Aim:Compare manual and automatic analysis of neurological and respiratory variables obtained through the program Alice5,Respironics Philips. Methods:The study included 34individuals with suspected sleep-disordered breathing and performed a conventional PSG in our sleep laboratory. Each record was subjected to two analyzes:1-Automatic analysis of Alice5 (PSG);2-manual PSG Scoring according to the criteria of AASM. Statistical analysis was performed using SPSS. Results: It was found that significant differences between the two scorings. With exception of REM sleep all mean sleep stages values and mean the arousal index were statistically significant.It appears that the alorythm of Alice has an excellent sensitivity to recognize REM sleep, however overestimates the fast rhythms leading to an unlikely increase in N1 sleep. Conclusions/Discussion:The” Alice 5” program showed serious different results in the staging of sleep and respiratory events. Taking into account the experience of the scorer it seems more likely that the automatic scoring algorithm is not yet sufficient to reach results of the human eye. A therapeutic decision based on the automatic computed results appears questionable.