Introduction: Determining whether using inhaled corticosteroid (ICS) is effective in gaining control in children with severe asthma is difficult. Often multiple influencing factors change in a temporally complex manner. Thus, the success of a particular medication is difficult to disentangle, and the clinical diagnosis of ‘steroid resistance’ is often based on non-quantified clinical criteria. Aim:In a well controlled setting, we determined whether analysing twice daily lung functions (PEF) in comparison to drug changes could add information on individual drug response. Methods: We retrospectively investigated lagged cross correlations between the PEF series over 6 weeks and changes of medication doses in 90 patients. Based on their statistical significance, the patient was classified as responsive to the medication changes (ICS, LABA). This was compared to the overall clinical response based on GINA level of control during the observation period. Results: In 42 patients we observed changes in ICS. Only in 20(48%) significant ICS-responses were seen. In comparison to the overall clinical treatment response, 11 patients were true positives, 9 true negatives, 12 false negatives, and 10 false positives. Conclusion: The concordance between our method and clinical response was poor. However, this method might help phenotyping patients with severe asthma, support the effectiveness of ICS treatment (true pos, true neg), and identify disconcordant subgroups (false neg, false pos), where other treatments are more important for therapy success. E.g. 10(24%), who were clinically identified as steroid resistant, showed one or more statistically significant responses to ICS.
Sprang et al. (Sun,) studied this question.