The contribution of bias must be considered when interpreting the results of this study. Two potential sources of bias exist. These are as a result of the study design and in the calculation of the predicted normal values. The trial design has used the baseline FEV1 for many of the calculations described in this article. The baseline FEV1 was used as one of the entry criteria for the study and thereby may have given an erroneously high result as investigators and patients attempted to obtain the best possible value for FEV1 at the baseline visit. Also, the design allowed for concomitant treatment to be administered throughout the study; indeed, by the end of the 3 years of study, 45% of the placebo group had received inhaled oral or systemic corticosteroids. This could minimize any possible treatment differences, and the effect would increase as the study progressed. The second source of bias is in the calculation of predicted normal values, in which prediction equations used are > 30 years old and likely provide predicted normal values that are too low. Therefore, interpreting the absolute levels within the groups is likely not precise; however, the random allocation should allow accurate estimates of treatment differences between groups.
The bronchodilator response documented at the baseline visit was 9.9% improvement in FEV1, and the magnitude of the benefit is, in part, due to the baseline measurement bias described above. The subsequent measurements made at years 1, 2, and 3 show a consistent difference between the budes-onide and placebo treatment groups, with the placebo group having approximately 1% greater bron-chodilator response at each visit. This finding is in agreement with that in another study.
In summary, this study has demonstrated that early intervention with inhaled budesonide within the first 2 years of asthma diagnosis in patients with persistent asthma improves both prebronchodilator and postbronchodilator FEV1. In addition, postbronchodilator FEV1 values declined in both the budes-onide and placebo treatment groups. This suggests that there may be inhaled steroid-sensitive and inhaled steroid-insensitive components to airway structural changes associated with asthma, which cause the decline in percentage of predicted FEV1, or alternatively the complete inhibition of the decline in FEV1 requires a higher dose of inhaled corticosteroids than was used in this study.