Visual and different automatic scoring profiles of respiratory variables in the diagnosis of sleep apnoea-hypopnoea syndrome

Abstract

The purpose of our study was to explore the diagnostic accuracy of different methods of scoring night time recording of respiratory variables (NTRRV) for the diagnosis of the sleep apnoea-hypopnoea syndrome (SAHS). Within a 2 week period, we performed a partially attended uight time recording of respiratory variables and a full polysomnography (PSG) for reference in patients with suspected SAHS. Night time recording of respiratory variables was carried out using equipment which records, and continuously displays on a monitor, oximetry, airflow, chest and abdominal motion and body position. Night time recording of respiratory variables was scored manually and automatically, according to different combinations of the parameters described previously. Full polysomnography was performed in the Sleep Laboratory following conventional standards. Thirty six patients were studied. Visual analysis and different automatic scoring profiles of night time recording of respiratory variables were compared to full polysomnography in terms of agreement, sensitivity and specificity. Visual scoring of night time recording of respiratory variables gave the finest agreement-sensitivity-specificity relationship. Automatic scoring of nighttime recording of respiratory variables showed a trend to underestimate the apnoea-hypopnoea index (AHI) with respect to full polysomnography due mainly to underrecognition of hypopnoeas. Agreement-sensitivity-specificity relationships of automatic night time recording of respiratory variables with respect to full polysomnography varied depending on the automatic profile used. Some had a good agreement and sensitivity whilst others had a good specificity. These findings show that visual scoring of night time recording of respiratory variables is the most accurate method of analysis when compared to full polysomnography. The usefulness of the automatic methods of scoring of respiratory variables depends on the end-point chosen and is not reliable enough to be used in all situations. Night time recording of respiratory variables represents a real complement to conventional full polysomnography in clinical practice.