Pulse oximetric thresholds for tonsillectomy and adenotomy in children: significance of 1–2% decline in oxyhemoglobin saturation

Abstract

We aimed to establish optimal overnight pulse oximetric thresholds for determining the indication of tonsillectomy and adenotomy (TA) in children by revising the definition of 'desaturation'. One hundred and thirty four children scheduled for TA (TA group, 5.3 +/- 1.4 years old) and 112 otherwise healthy children scheduled for elective minor surgery (control group, 5.4 +/- 1.5 years old) were enrolled into this prospective study. Data were recorded and stored every 10 s using Nellcor N-395. Desaturation/resaturation events were defined as x% change (x = 1-4) of SpO2 (oxyhemoglobin saturation by pulse oximetry) in 10 s. The desaturation/resaturation indices were calculated as events per hour of total sleeping time. For each index, a wide range of temporary thresholds was set. The optimal thresholds for TA were the ones that maximized the weighted average for sensitivity, specificity (based on whether the index improved or not after TA), and the percentage of the control children whose indices were below the threshold. For all the indices, the optimal thresholds that fulfilled the above condition were determined. Compared with the x = 3-4% results, the application of x = 1-2% approximately doubled the TA patients whose preoperative 'positive' indices improved after TA, with the weighted averages of 84.3-92.3% as described above. By defining desaturation/resaturation as a 1-2% change in SpO2 from the preceding value, children with adenotonsillar hypertrophy whose pulse oximetric indices are expected to improve after TA can be detected by pulse oximetry with relatively high sensitivity and specificity.