C. Butler and L.G. Heaney raise several important points regarding our recent report 1, which identified risk factors of frequent exacerbations in difficult-to-treat asthma. First, they questioned why we excluded patients on oral corticosteroids from the analysis. In our clinic, and in most other pulmonary clinics, it is common practice to taper oral corticosteroids to the lowest possible dose whenever possible, and to increase the dose no more than strictly necessary in case of the worsening of asthma symptoms. This is a continuing process, mostly initiated by the patients themselves, without interference of a pulmonologist. We chose not to include patients on oral corticosteroids, because we felt that initial exacerbations could not be distinguished clearly from temporary deteriorations in symptoms.
C. Butler and L.G. Heaney also question the appropriateness of our definition of difficult-to-treat asthma. We adapted the European Respiratory Society Task Force definition of difficult asthma, i.e. “failure to achieve asthma control when maximally recommended doses of inhaled therapy are prescribed for at least 6–12 months” 2. Our patients were symptomatic despite the regular use of high doses of inhaled corticosteroids combined with long-acting bronchodilators. They were nonsmokers, and had a limited smoking history of <10 pack-yrs. They were only included in our study if they had been previously assessed and treated by a respiratory specialist, and closely supervised by the same specialist for ≥1 yr. We assumed that this was a long enough period to exclude unidentified or alternative diagnoses. Our patients have now been followed for another 5 yrs, and there was only one patient in whom the diagnosis of asthma was ultimately rejected; this patient suffered from chronic embolic syndrome presenting as recurrent severe wheezy attacks. More aggressive treatment of comorbid factors, such as chronic rhinosinusitis or gastro-oesophageal reflux, resulted in a better asthma outcome in ∼20% of the patients.
With regards to poor adherence, we agree that this can be an important exacerbating factor in patients with asthma. However, poor compliance with treatment is notoriously difficult to estimate 3. In contrast to the UK, individual prescription records were not readily available in the Netherlands at the time of the study. Therefore, unfortunately, the (indirect) method to assess compliance as referred to by C. Butler and L.G. Heaney could not be applied.
C. Butler and L.G. Heaney are also concerned about our diagnosis of gastro-oesophageal reflux. In our study, the diagnosis of reflux was based on 24-h pH measurement (n = 39) and/or on the basis of a trial with proton-pump inhibitors (n = 66). The latter diagnostic test was performed in those who were too dyspnoeic to undergo a 24-h pH measurement. All other patients who had not undergone one of these diagnostic tests were considered as having no reflux, which was most likely an underestimation of the real prevalence of this comorbid factor. In our opinion, this emphasises the importance of this exacerbating factor in patients with difficult-to-treat asthma even further.
Finally, C. Butler and L.G. Heaney expressed concern about our definition of respiratory infections. We based the diagnosis of respiratory infection on symptoms (episodes of increased dyspnoea, accompanied with increased production of purulent sputum) and the prescription of a course of antibiotics by the respiratory specialist. Since healthcare providers in the Netherlands are particularly conservative when it comes to the prescription of antibiotics for respiratory diseases 4, we assumed that there was a serious suspicion of upper or lower respiratory infection. We agree that this is not a rock-hard definition, but, in our opinion, workable for the purpose of this study.
- © ERS Journals Ltd
