Utility of home oximetry as a screening test for patients with moderate tosevere symptoms of obstructive sleep apnea

Objective: To determine the value of home oximetry as a screening test in p atients with moderate to severe symptoms of obstructive sleep apnea (OSA). Design: Retrospective, observational study. Setting: The Sleep Unit of a tertiary referral, university hospital. Patients: 116 patients referred for evaluation of moderate to severe sympto ms of OSA in which both home oximetry and polysomnography (PSG) were perfor med. Interventions: NA Results: Three numerical oximetry indices were evaluated: average of desatu rations greater than or equal to 4% and average of resaturations greater th an or equal to 3% per hour of analysis time (DI4% and RI3%, respectively); and cumulative percentages of time spent at saturations below 90% (CT90%). A qualitative assessment was also performed. Oximetry indices were compared with apnea/hypopnea index (AHI) by simple linear regression and Bland-Altm an analyses. Optimal cut-off points, in terms of sensitivity and specificit y, for the oximetry indices were searched using ROC analysis, at an AHI thr eshold of greater than or equal to 10. The correlation between AHI and the desaturation indices was r=0.50 for CT90%, r=0.60 for DI4%, and r=0.58 for RI3%. No bias was found between PSG and oximetry indices in Bland-Altman pl ots. Neither the numerical indices nor the qualitative analysis achieved an adequate (> 0.8) area under the ROC curve. A CT90% <0.79 excluded OSA with 84% sensitivity. A DI4% greater than or equal to 31.4 or a RI3% greater th an or equal to 40.5 diagnosed OSA with 97% specificity. Using these values, 38% of the patients would have been correctly classified by oximetry alone , 10% would have been incorrectly classified, and 50% could not have been c lassified with certainty. Eleven (15%) OSA patients would have been missed by oximetry. Conclusions: Correlation between home oximetry and PSG was not high. Oximet ry was more useful to confirm than to exclude OSA in our study. Qualitative assessment was not better than numerical analysis. The greatest value of o ximetry in this setting seems to be as a tool to rapidly recognize and trea t more severe OSA patients in waiting list for PSG.