To the Editor:

Dr. Hansen's argument (in his letter that appears in this issue [1]) that a response to a medical intervention in a patient should be considered as positive only if the functional parameter exceeds intra-individual variability is interesting and needs serious consideration. His statement is based on studies documenting more discriminative responses after acute inhalation of a β₂-agonist (2) or a rehabilitation program (3), when individual rather than group variability was considered. Our comments are intended to bring further contribution to the topic.

Most of the problems in biology arise not only from how the data are analyzed but also from the choice of the parameters that best identify the outcome. As an example, when using volitional tests such as the FEV₁ or 6-minute-walk test, the pros of the statistical approach might well be overweighed by the cons of functional parameters whose variability is largely determined by the patient's cooperation and, specifically for FEV₁, by the variable effect of the inspiratory maneuver preceding forced expiration on bronchial tone (4) and thoracic gas compression (5). Based on an average measurement variability within 8% and 150 ml, the European Respiratory Society/American Thoracic Society guidelines suggested that changes in the FEV₁ larger than 12% and 200 ml are required to define bronchoreversibility (6). Although we do agree with Dr. Hansen that smaller changes may be significant in subjects able to give highly repeatable measurements, there are some requirements for short-term natural variability of a given test to be used as a diagnostic tool or an outcome in clinical practice. These include the knowledge of short-term variability in healthy subjects, long-term variability in patients (7), and minimum clinically important difference. Therefore, threshold values need to be set even if a statistical approach is used, as suggested by Dr. Hansen.

Besides the above considerations, we believe that the most promising application of short-term intra-individual variability may be in the prediction of clinically relevant events, such as asthma attacks or deterioration of lung function (8). To this aim we report that the choice of lung function tests requiring minimal or no patient cooperation, such as the forced oscillation technique, has been revealing a very promising and encouraging tool for assessment of home variability of airway caliber (9).