Unreliability of automatic scoring of MESAM 4 in assessing patients with complicated obstructive sleep apnea syndrome

Background: Portable devices are used for unattended recording of patients with suspected obstructive sleep apnea syndrome (OSAS). The MESAM 4 (MAP; Martinsried, Germany) is a computerized ambulatory polysomnographic system that records four parameters: breathing noise, heart rate, arterial oxygen saturation (Sao₂), and body position. Design and method: We evaluated the reliability of the oxygen desaturation index (ODI) automatically calculated by the MESAM 4 device in evaluating patients with "complicated" OSAS. These patients present Sao₂ drops due to apneas associated with a fall in baseline Sao₂ during sleep, as occurs in the "overlap syndrome." Ten patients with complicated OSAS underwent nocturnal MESAM 4 recordings, and we compared the visual and automatic scorings of the ODI. Results: The ODI obtained with visual scoring was significantly higher than ODI automatically calculated by the MESAM 4 in all patients. In some patients, this difference was so significant that it could bias clinical judgment of OSAS severity. We demonstrated that the system did not identify those desaturation events that were superimposed on a fall in baseline Sao₂. The error depends on the algorithm by which the device recognizes the desaturation events and calculates the baseline Sao₂. Conclusion: Automatic analysis of MESAM 4 recordings may be misleading in evaluating OSAS patients who have a fall in baseline Sao₂ during sleep. In this case, visual scoring performed by a trained polysomnographer is recommended.