Continuous oximetry recordings on the first post-operative night after pediatric adenotonsillectomy-a case-control study

Background: Children with obstructive sleep apnea (OSA) with recurrent dips in oxygen saturation (SpO₂) during sleep are known to be at increased risk of post-operative airway compromise after adenotonsillectomy (AT). We aimed to determine the extent of desaturation on the first post-operative night in children known to have recurrent desaturation pre-operatively and to compare the extent of desaturation in that group with results in children known to have normal oximetry recordings pre-operatively. Methods: Prospective sequential recruitment of 57 children who had overnight oximetry performed on the first night after adenotonsillectomy was undertaken, including 28 with a McGill Oximetry Score (MOS) of 2–4 pre-operatively (high risk group) and 29 with a normal/inconclusive pre-operative MOS (low risk group). Oximetry parameters (mean SpO₂, SpO₂ nadir, and rates of SpO₂ dips below 90% and dips of ≥4%) were compared to the pre-operative oximetry result. Demographic and clinical factors, and the occurrence of post-operative complications, were derived from the medical record. Results: In the high risk group, the MOS improved in 23/28 children, but remained abnormal in 82%. Conversely, in the low risk group 26/29 (90%) had a normal post-operative oximetry. The remaining 3, all of whom had severe OSA on pre-operative polysomnography, had a lowered baseline SpO₂ post-operatively. Mean SpO₂ was slightly lower post-operatively in both groups. In the high risk group, all other SpO₂ measures improved post-operatively. Respiratory adverse events were more common in the high risk group as expected (39% compared to 3% in the low risk group, p = 0.001). An adverse event requiring clinical intervention was significantly more likely if the post-operative oximetry was abnormal (result unknown to the treating team), occurring in 73% of children with an abnormal compared with 32% of children with a normal post-operative oximetry (p = 0.002). Conclusion: Most children with an abnormal oximetry pre-operatively continued to have an abnormal oximetry on the first night after AT, albeit somewhat improved. While adverse events were more frequent in children with an abnormal post-operative oximetry, half (54%) did not suffer a clinical respiratory adverse event despite having repetitive desaturations on downloadable oximetry. These findings support close clinical observation of children at high risk of complications post-operatively, especially those with abnormal oximetry pre-operatively, rather than focusing on recurrent dips in SpO₂ on post-operative oximetry downloads in the absence of clinically evident complications.