Perioperative respiratory adverse event risk assessment in children with upper respiratory tract infection: Validation of the COLDS COLDS score

摘要

Summary Background The decision to proceed with anesthesia and surgery has been controversial in pediatric patients with an upper respiratory tract infection. The COLDS score was proposed by Lee and August as a potential risk stratification scheme, but no validation has been performed on this scale. Aims The aim of this study was to evaluate the utility of the COLDS score in predicting perioperative respiratory adverse events and optimize its predictive ability. Methods COLDS scores, incidence of perioperative respiratory adverse events, surgical procedure type, and age were prospectively collected for 536 patients who met inclusion criteria. Area under the receiver operating characteristic curves was computed for total COLDS score and individual COLDS score categories. Multivariable regression was used create an optimized score. To quantify the decrease in risk associated with case cancelation due to illness, the other risk factors in COLDS were assessed separately from upper respiratory infection status and a risk model was created. Results The area under the receiver operating characteristic curve for the total COLDS score was 0.69, suggesting that the COLDS score has a moderate predictive ability for perioperative respiratory adverse events. When split into individual component scores, the area under the receiver operating characteristic curve ranged from 0.55 to 0.63. We also found that the area under the receiver operating characteristic curve for the scoring system was higher in younger children than for children aged 4‐6 (area under receiver operating characteristic curve of 0.70‐0.71 vs 0.66). The area under the receiver operating characteristic curve for the optimized scoring system was 0.71. Conclusion The COLDS score has the potential to be a valuable risk assessment tool for prediction of perioperative respiratory adverse events and appears to have a better predictive value in certain subpopulations.