Abstract
The use of repeated measurements to assess asthma has been little investigated, despite its relevance regarding temporal phenotypes. Using repeated answers to a single question asked 7 times over 15 yrs, 3 asthma patterns were a priori defined. The aim was to compare these patterns to prescription data to evaluate their validity.
Asthma answers (have you ever had asthma attacks?) were collected from 7 questionnaires in the French E3N study (70,428 women, 55-82 yrs). The first pattern “never asthma” included women who never answered yes (n=64,061, 91%). The second pattern “consistent asthma” included women who answered yes at least once, never followed by a no in subsequent questionnaire (n=2,853, 4%). The third pattern “inconsistent asthma” included women who answered yes followed by at least one no (n=3,514, 5%). Using prescription database, we identified inhaled corticosteroids (ICS) and bronchodilators (IBD) dispensed (2004-2009)
“Consistent asthma” claimed back more anti-asthma drugs than “never asthma”; “inconsistent asthma” had intermediate means of reimbursements (yearly means ICS/IBD: 2.24/1.17, 0.10/0.04, 1.23/0.73 resp.). Both “consistent” and “inconsistent” patterns had increasing reimbursements of ICS and IBD according to the number of positive answers to the question through the 7 questionnaires.
Our results showed a good reliability of our a priori asthma patterns, underlying the importance of temporal variations in asthma.
- © 2012 ERS
