Preoperative Predictors of Severe Respiratory Events After Tonsillectomy: Consideration for Pediatric Intensive Care Admission.

Abstract

Few data are available to guide postadenotonsillectomy (AT) pediatric intensive care (PICU) admission. The aim of this study of children with a preoperative polysomnogram (PSG) was to assess whether preoperative information may predict severe respiratory events (SRE) after AT. Retrospective cohort study. Single tertiary center. Children aged 6 months to 17 years who underwent AT with preoperative polysomnography (2012-2018) were identified by billing codes. Data were extracted from medical records. SRE were defined as any 1 or more of desaturations <80% requiring intervention; newly initiated positive airway pressure; postoperative intubation; pneumonia/pneumonitis; respiratory code, cardiac arrest, or death. We hypothesized that SRE would be associated with age <24 months, major medical comorbidity, obesity (>95th percentile), apnea-hypopnea index (AHI) ≥ 30, and O₂ nadir <70% on PSG. Analysis was performed with multivariable logistic regression. Of 1774 subjects, 28 (1.7%) experienced SRE. Compared to those without, children with SRE were on average younger (3 vs 5 years, p < .01) with a greater probability of medical comorbidities (59% vs 18%, p < .001). After adjustment for sex, black race, obesity, and age <24 months, children with major medical comorbidity were more likely than other children to have SRE (odds ratio [OR]: 14.2; 95% confidence interval [CI]: [5.7, 35.2]), as were children with AHI ≥ 30 (OR: 7.7 [3.0, 19.9]), or O₂ nadir <70% (OR 6.1 [2.1, 17.9]). Age, obesity, sex, and black race did not independently predict SRE. PICU admission may be most prudent for children with complex medical co-morbidities, high AHI (>30), and/or low O₂ nadir (<70%).