Background 

Though oximetry is useful in screening for Obstructive Sleep Apnea (OSA), the optimal cut-off values of oximetry indices for differentiating Primary Snoring (PS) from OSA remain unclear.

Objective

The aim was to determine the utility of overnight oximetry indices in differentiating PS from OSA in children with suspected OSA.

Methodology 

In this retrospective study, we collected clinical, polysomnography (PSG) and oximetry data of children aged 1-18 years with suspected OSA. Patients with craniofacial anomalies, syndromes, neuromuscular conditions and central apnea index ?5 were excluded. Oximetry indices analysed were oxygen desaturation index (ODI₃, defined as the number of ?3% desaturation episodes per hour of sleep), SpO2 nadir and the sleep duration with SpO2 <94%. 

Results 

Of 1203 children, 1081 (89.9%) had OSA. ROC analysis showed ODI₃ to have the largest area under curve of 0.869. ODI₃ cut-off of 2.40 achieved 79.7% sensitivity and 78.7% specificity, in differentiating PS from OSA. The optimal cut-off values of ODI₃ for predicting OSA severity were 2.00 (71.5% sensitivity, 70.5% specificity) for mild OSA, 3.70 (82.8% sensitivity, 93.4% specificity) for moderate OSA and 4.30 (99.3% sensitivity, 96.7% specificity) for severe OSA. There was excellent correlation between ODI₃ and both Obstructive Apnea Hypopnea Index (r=0.963; p<0.05) and Respiratory Arousal Index (r=0.808, p<0.05).

Conclusions 

This study provides optimal cut-off values for ODI₃ in differentiating PS from OSA. As overnight oximetry is cheap and widely available unlike PSG, it has the potential to be incorporated into cost-effective clinical decision making algorithms.