Background To determine which diary card variables are the most predictive for administration of additional courses of corticosteroids using the TRUST (The Regular Use of Salbutamol Trial) data set.
Methods Logistic regression models were used to identify the extent to which a change in diary card variable affected the odds ratio (OR) for administering a course of oral or increased inhaled corticosteroids. The complete TRUST diary card data were used with over 200 000 days of diary card observations from 983 mild to moderate asthmatic subjects.
Results An increase in daytime symptoms of 1–5 U over baseline was associated with an increase in the OR for starting all types of corticosteroids from two- to 60-fold.
Conclusions These results indicate that an increase in daytime symptoms of two or more over baseline strongly predicts the administration of additional corticosteroids. The results have significant implications for both clinical practice and design of clinical trials in asthma.