Abstract
This study evaluated the accuracy of a portable recording device (MESAM IV) in identifying obstructive sleep apnoea (OSA). The MESAM IV system measures arterial oxygen saturation (Sa,O2), heart rate, snoring sounds and body position, and allows both automatic and manual scoring of the recordings. Nocturnal polysomnography and MESAM IV recordings were performed simultaneously in 150 patients with suspected OSA, and were analysed blindly by a different observer. Patients with an apnoea-hypopnoea index (AHI) greater than or equal to 10 were diagnosed as having OSA. In the evaluation of the discriminatory ability of MESAM IV scores, the cut-off point was set to minimize first the exclusion of truly diseased patients (i.e. false-negative interpretations), and then confirmation of nondiseased subjects (i.e. false-positives). When used as an exclusion test, the portable device reached a sensitivity of 0.98 and a specificity of 0.78; as a confirmation test, these values were 0.69 and 0.97, respectively. These results were achieved with manual scoring, which was superior to automatic scoring. Manual scoring was also better than automatic scoring when OSA was defined according to other threshold values (> or = 5, 15 and 20) for the AHI. The combination of MESAM IV manual scores could reduce the need for diagnostic polysomnography in three quarters of the patients clinically suspected of having obstructive sleep apnoea, substantially reducing costs associated with diagnostic procedures.
