Carbon dioxide (CO₂) using capillary blood gas (CBG) analysis is commonly used children with acute bronchiolitis. Evidence to support its use is limited.

A retrospective observational study was conducted over two bronchiolitis seasons (2014 -2016) of infants admitted to a tertiary teaching hospital using patient electronic medical records. Using logistical regression models (STATA/IC 12.1) the association between CBG pCO₂ and markers of disease severity (length of stay (LOS) and high dependency admission (HDU)) was examined.

332 children were assessed with 526 CBG performed in 158 infants (mean age 0.31 years, 54% male, 27% premature, 77% RSV positive). The initial CBG pCO₂ was a mean 5.9kPa (SD1.1) and a maximum mean of 6.4kPa (SD1.5). Median LOS was 3 days (range 0–35). A CBG pCO₂ >7.0kPa during the admission (in 23% infants (36/158)) was significantly associated with younger age (OR 0.005 (95%CI 0.0007, 0.03); p<0.0001), the use of supplemental oxygen (OR 1.9 (95%CI 1.1, 3.3); p=0.033) (adjusted for age) and inspired fraction of oxygen (FiO₂) (slope coefficient 2.01 (95%CI 1.08, 2.94), p<0.0001) (adjusted for age). In 62% (98/158) a CBG was performed in ED and a pCO₂ >7kPa (N=26/98) in ED was significantly associated with LOS (IRR 1.4 (95%CI 1.1,1.8); p=0.008) and HDU admission (OR 3.5 (95%CI 1.7,7.8); p=0.001).

CBG pCO₂ >7 kPa identifies children in ED with more severe disease with longer length of stay and risk of admission to HDU. Our results suggest that CBG pCO₂ may be a possible marker of severity in future intervention trials for bronchiolitis.