Reply

To the Editor:

We thank Caudri and de Jongste for their interest in our article and thoughtful comments. Their first comment has been addressed at length in the discussion of our article and also in the meticulous debate presented by Bush and Eber.1 In brief, there is no question that asthma may be diagnosed without difficulty by the primary physician on the basis of typical history, response to therapy, and, when needed, additional tests. However, even though this statement holds for many or even most children in the community with clinical symptoms suggesting asthma, it may not be true for a relatively small percentage of children with less-specific complaints or who ignore mild to moderate symptoms or do not respond characteristically to treatment. Because of the high incidence of asthma, this group still includes a substantial number of children referred to special clinics. It is this minority, albeit a considerable load, that consumes more healthcare resources and can benefit the most from early diagnosis, with the emphasis on “early.” We agree that the diagnostic yield of asthma also will increase in this minority population by adding IgE levels and other tests, such as skin tests for allergy, adenosine, exercise, and methacholine challenge tests. However, when the patient arrives for the first visit to a special clinic after being referred by his or her physician, the doctor at the clinic has the choice to either ask for a battery of these tests and see the patient again for evaluation or, as our data suggest, perform a cheap, cost-effective FeNO test and get immediate results. This may improve early institution of treatment. Given the test's ease of performance compared with provocation tests and induced sputum, and its better specificity and sensitivity compared with spirometry, FeNO may offer the best single, cost-effective test for diagnosing asthma in children.

We agree with the comment regarding our exclusion criterion of children without asthma treated with steroids. In fact, all children receiving steroids (37 cases) were excluded from our analysis. We failed to mention that 4 of these 37 children were determined to not have asthma. We should have defined this group of 37 children treated as “asthma and non-asthma treated with ICS.” This oversight did not affect our results, however, because these 37 children were excluded regardless of their final diagnosis and did not contribute to the study. We thank Caudri and de Jongste for their observation.