Oximeter performance: the influence of acquisition parameters.

STUDY OBJECTIVE This study was designed to determine whether different desaturation indexes (DIs) would be obtained in patients with sleep-disordered breathing by systematically altering two acquisition parameters: the recording setting and the display mode. DESIGN Prospective clinical study. SETTING Community sleep-disorders center. PATIENTS The study included 75 patients who were suspected of having sleep-disordered breathing. INTERVENTIONS Each patient had simultaneous pulse oxyhemoglobin saturation (SpO2) traces at three recording settings (3 s, 6 s, and 12 s) during the diagnostic phase of split-night polysomnography. On-line and memory displays of those data at each recording setting were obtained. DIs for > or = 3% desaturation events per hour were calculated for each of the six traces. RESULTS The mean on-line DIs significantly differed from each other, with slower (longer) recording settings resulting in lower values than faster (shorter) settings. The memory DIs all significantly underestimated the on-line DIs. Pearson correlations ranged from 0.82 to 0.90 between the on-line/memory DI pairs, but Bland-Altman analysis detected disagreement at higher levels of disordered breathing. CONCLUSIONS These findings confirm that significantly different SpO2 data are obtained at various acquisition options. The recording setting and display mode parameters should be disclosed in all reports employing oximetry with the fastest recording setting and on-line display mode preferable for case finding of sleep-disordered breathing.