To the Editors:
I would like to comment on the editorial by Johnson and Raghu 1, which was published in the November issue of the European Respiratory Journal. In a very clear way, the authors overviewed the use of outcome measures in clinical trials of idiopathic pulmonary fibrosis (IPF).
I would like to propose some concerns, which could affect the external validity of randomised controlled trials in IPF 2, 3. The first concern is about the diagnostic accuracy of the disease. IPF is a rare disease and no single accurate test for the diagnosis of IPF exists. Studies of the accuracy of diagnosing IPF are performed mostly in tertiary referral centres, and, even in these studies, an important interobserver variability exists. In most of these studies, prior knowledge of the presence of a form of interstitial lung disease existed, which may evoke an observer bias and, therefore, influence the results on the diagnostic accuracy 3.
The incidence of IPF in a general pulmonary practice is low. Diagnostic accuracy (i.e. sensitivity and specificity) also depends on the prevalence of the disease. A lower prevalence of the disease results in a higher number of false-positive and false-negative diagnoses 3. In recent published trials, participating centres were selected from tertiary care centres or secondary care centres with particular interests in the management of IPF.
If a treatment for idiopathic pulmonary fibrosis is found, the chance that this treatment will be given to patients without idiopathic pulmonary fibrosis is high in a pulmonary care practice with a very low incidence of idiopathic pulmonary fibrosis. Or to summarise these concerns using a question: will it be possible to generalise these trials in a general pulmonary care practice 2?
- © ERS Journals Ltd