Peri-operative adverse respiratory events in children

Three quarters of all critical incidents and a third of all peri-operative cardiac arrests in paediatric anaesthesia are caused by adverse respiratory events. We screened for risk factors from children's and their families' histories, and assessed the usefulness of common markers of allergic sensitisation of the airway as surrogates for airway inflammation and increased risk for adverse respiratory events. One hundred children aged up to 16 years with two or more risk factors undergoing elective surgery were included in the study. Eosinophil counts, IgE level, specific IgE for D. pteronyssinus, cat epithelia and Gx2 (grass pollen) were measured for each child and adverse respiratory events (bronchospasm, laryngospasm, oxygen desaturation < 95%, severe persistent coughing, airway obstruction and postoperative stridor) were recorded. Twenty-one patients had an adverse respiratory event but allergic markers were poor predictors. Binary logistic regression showed a lack of predictive value of the eosinophil range and adverse respiratory events (p = 0.249). Receiver operating characteristic (ROC) curves for the presence of adverse respiratory events vs level of specific IgE antibody (to Gx2 (AUC 0.614), cat epithelia (0.564) and D. pteronyssinus (0.520)) demonstrated poor predictive values. However, the presence of risk factors was strongly associated with adverse respiratory events (p < 0.001) and a ROC-curve analysis indicated a fair capacity to predict adverse respiratory events (AUC 0.788). There was a significant difference (p = 0.001) between the presence of adverse respiratory events in patients with more than four (p = 0.006), compared with less than four (p = 0.001), risk factors. We conclude that while risk factors taken from the child's (or family) history proved good predictors of adverse respiratory events, immunological markers of allergic sensitisation demonstrated low predictive values. Pre-operative identification of children at high risk for an adverse respiratory event should rely on clinical, rather than immunological, assessment