To explore the predictive ability of capillary blood gas (CBG) pCO₂ for respiratory decompensation in infants aged ≤6 months with bronchiolitis admitted from the emergency department; to determine whether end-tidal CO₂ (etCO₂) capnography can serve as a less invasive substitute for CBG pCO₂.

This was a prospective cohort study of previously healthy infants aged ≤6 months admitted for bronchiolitis from the emergency department (ED). Initial CBG pCO₂ and etCO₂ capnography were obtained in the ED prior to inpatient admission. Simple logistic regression modeling was used to examine the associations of CBG pCO₂ and etCO₂ capnography with respiratory decompensation. Pearson's correlation measured the relationship between CBG pCO₂ and etCO₂ capnography.

Of 134 patients, 61 had respiratory decompensation. There was a significant association between CBG pCO₂ and respiratory decompensation (OR = 1.07, p = 0.003), even after outlying values were excluded (OR = 1.06, p = 0.005). End tidal CO₂ capnography was not significantly associated with decompensation (OR = 1.02, p = 0.17), even after outlying values were excluded (OR = 1.02, p = 0.24). There was a moderate correlation between etCO₂ capnography and CBG pCO₂ (r = 0.39, p < 0.001).

In infants with bronchiolitis, CBG pCO₂ provides an objective measure for predicting respiratory decompensation, and a single etCO₂ measurement should not replace its use.