Blood gas tensions in adult asthma: a systematic review and meta-regression analysis

The last half-century has seen substantial changes in asthma treatment and care. We investigated whether arterial blood gas parameters in acute and non-acute asthma have changed historically.

We performed a systematic search of the literature for studies reporting P(aO2) , P(aCO2) and forced expiratory volume in 1 s, percentage of predicted (FEV1%). For each of the blood gas parameters, meta-regression analyses examined its association with four background variables: the publication year, mean FEV1%, mean age and female fraction in the respective studies.

After screening, we included 43 articles comprising 61 datasets published between 1967 and 2013. In studies of habitual-state asthma, mean P(aO2) was positively associated with the publication year (p = 0.001) and negatively with mean age (p < 0.01). Mean P(aCO2) showed a positive association with publication year (p = 0.001) and a negative association with female fraction (p < 0.05). In acute asthma studies, blood gas levels were unassociated with publication year and mean age, mean P(aO2) was positively associated with FEV1% (p < 0.05) whereas mean P(aCO2) showed a negative association with FEV1% (p < 0.05) for studies with mean FEV1% <40. In neither acute nor habitual-state studies was mean arterial pH associated with any of the predictor variables.

In studies of habitual-state asthma, mean reported P(aO2) and P(aCO2) levels were found to have increased since 1967. In acute asthma studies, mean P(aO2) and P(aCO2) were associated with mean FEV1% but not with either publication year or patient age.