ABSTRACT
OBJECTIVES/HYPOTHESIS: To identify clinical risk factors for oxygen desaturation in the first 24 hours post-tonsillectomy, thus permitting the identification of those patients who warrant inpatient monitoring. STUDY DESIGN: A retrospective analysis of 4092 consecutive patients undergoing tonsillectomy over a two-year period. METHODS: Detailed clinical data were recorded for all patients who desaturated in the postoperative period (n = 294) and randomly selected controls (n = 368). Univariate and multivariate analysis was performed in order to identify independent risk factors for desaturation. RESULTS: There were 294/4092 patients (7.2%) who experienced desaturations (defined as sustained saturations <90%) in the first 24 hours postoperatively (mean nadir, 78.7%). Multivariate analysis identified seven independent clinical risk factors for desaturation in the initial 24 hours post-tonsillectomy: trisomy 21, weight, coexistent cardiac disease, a coexistent syndromic diagnosis, a clinical diagnosis of obstructive sleep apnea (OSA), a coexistent neurologic diagnosis, and a prior diagnosis of pulmonary disease. A policy that admits all patients exhibiting any one of these risk factors except OSA would have identified 92% of the patients who subsequently desaturated. However, such a policy would also have required admission of 60% of the patients in our control group. CONCLUSIONS: These findings are generally consistent with the Clinical Practice Guidelines recently published by the American Academy of Otolaryngology. In a tertiary care center, it may not be possible to identify an algorithm that admits all children at risk of desaturation while permitting the discharge of a high percentage of patients. LEVEL OF EVIDENCE: 3b.