Capnometry as a predictor of admission in bronchiolitis.

Objectives: Bronchiolitis is a dynamic condition, and predicting clinical deterioration can be difficult. The objective of this study was to determine whether capnometry readings among bronchiolitic children admitted to the hospital are significantly different from those discharged from the emergency department. Methods: We prospectively studied a convenience sample of children younger than 24 months with clinical bronchiolitis. A single end-tidal CO2 (ETCO2) reading was taken before treatment, and a clinical work of breathing score was assigned to each patient. Treating physicians and nurses were blinded to capnometry readings. The decision to admit was based on the judgment of the attending physician. Descriptive statistics and appropriate hypothesis testing were performed. A receiver operating characteristic curve was constructed for the association between admission and capnometry readings. The alpha was set at 0.05 for all comparisons. Results: One hundred five children with bronchiolitis were included for study. Capnometry readings for admitted (mean, 32.6 mm Hg; 95% confidence interval [CI], 30.3-34.9 mm Hg) and discharged (mean 31.4 mm Hg; 95% CI 29.8-33.0 mm Hg) bronchiolitic children were not significantly different. Capnometry readings for low (mean, 31.7 mm Hg; 95% CI, 29.5-33.8 mm Hg), intermediate (mean, 32.1 mm Hg; 95% CI, 30.1-34.1 mm Hg), and high (mean, 30.5 mm Hg; 95% CI, 19.3-41.7 mm Hg) work of breathing (score) ranges were not significantly different. Conclusions: Capnometry readings are not useful in predicting admission for children younger than 2 years with clinical bronchiolitis. There are no significant differences in capnometry readings among bronchiolitic children with low, medium, and high work of breathing scores.