CPAP indication based on clinical data and oximetry for patients with suspicion of obstructive sleep apnea: A multicenter trial

BACKGROUND AND OBJECTIVE: The usefulness of pulse oximetry for the management of obstructive sleep apnea is controversial. The aim of this study was to assess the accuracy for indication of Continuous Positive Airway Pressure (CPAP) treatment in patients with suspected obstructive sleep apnea (OSA) based on clinical and oximetry data as compared to polysomnography (PSG). METHODS: This multicenter observational study involved seven sleep laboratories. Patients with suspicion of OSA who completed a standardized sleep questionnaire and a diagnostic PSG were enrolled. Eight observers logged on to a website independently and blindly. Seven observers only accessed the clinical data, curve and pulse oximetry results (Os-SO2-test method), while the eighth observer had full access to all indicators of PSG (O-PSG-reference method). Once observers assessed the information available on the website, they had to choose between three CPAP treatment options (yes/no/do not know) based on their knowledge and criteria. RESULTS: 411 subjects (228 men), median age 54 years, were available for evaluation. Os-SO2 had lower sensitivity (S), greater specificity (Sp) and positive likelihood ratio (PLR) to prescribe CPAP in patients more symptomatic (Epworth Sleepiness Scale-ESS > 10 or comorbidities) than those with fewer symptoms (ESS < 11 without comorbidities) (S 45-75% versus 45-91%, p 0.028); Sp 93.8-100% versus 68.5-96.6%, p 0.004; PLR > 10 versus 2.9-17, p<0.01). CONCLUSIONS: Due to its low false positive rate, a strategy based on pulse oximetry and clinical data was a consistent tool to indicate CPAP treatment in most symptomatic patients with a suspicion of OSA.