“Any fool can make a rule and any fool will mind it”

In principle, accurate guideline recommendations should lead to optimal management based on a secure diagnosis. However, current IPF diagnostic guidelines do not meet the needs of a major sub-group (possibly the majority) of patients with idiopathic pulmonary fibrosis (IPF). A great many IPF patients have HRCT appearances of “possible UIP”. A surgical biopsy is very often impracticable due to age, disease severity, co-morbidities or patient refusal. A guideline-based diagnosis cannot be made in these patients, although the diagnosis is often obvious. Inflexible diagnostic criteria, although essential for treatment trials, must necessarily be structured around an inflexible diagnostic algorithm. With this approach, non-standardised information (i.e. not available in all patients) must be omitted, including observed disease behaviour prior to and on treatment, findings on bronchoalveolar lavage, likelihoods in relation to age and a wealth of ancillary clinical information. However, when a diagnosis cannot be made using guideline criteria, a probable or highly probable “working diagnosis” of IPF can and should be made in most IPF patients by means of clinical reasoning, integrating all available non-standardised information.