We used data from the prospective multicenter IPF-PRO Registry to evaluate relationships between lung function measures and PROs in a contemporary cohort of patients with IPF. We found that there was only a weak relationship between decline in lung function and deterioration in HRQL assessed using PROs. No threshold of lung function decline performed well at classifying patients into those with versus without a concurrent deterioration in HRQL.

Weak correlations between lung function measures and PROs, and even more so between changes in lung function measures and changes in PROs, have also been observed in studies assessing the psychometric properties of PROs in patients with IPF [15, 16]. Consistent with this, clinical trials of therapies for IPF have demonstrated a significant reduction in the rate of FVC decline, but no significant difference in changes in PROs [17, 18]. These data suggest that physiologic measures and PROs assess different aspects of the impact of IPF. The impact of a given severity of disease on a patient’s HRQL may be influenced by a patient’s personal perspectives and lifestyle: a recent study in 50 patients with IPF found poor agreement between physicians’ ratings of patients’ clinical status, which were closely related to DLco % predicted, and the patients’ ratings, which were likely based on a more complex assessment of their overall well-being [19]. Further, as baseline lung function may affect the ability of a patient to perceive a decline in FVC, the same absolute change in FVC may have a different impact on symptoms or functional capacity in patients with different severities of disease. These observations support the evaluation of both changes in lung function and changes in HRQL in clinical practice and clinical trials, as recommended in a recent symposium on meaningful endpoints for use in clinical trials in IPF [20].

Understanding the minimal important change to the patient (MICP) of clinical measures, i.e., the smallest difference perceived as important by patients, would assist clinicians in the interpretation of changes. However, a prerequisite to performing such an analysis would be at least a moderate correlation between changes in the anchor and changes in the indicator of interest [21, 22]. In our study, the rather weak correlations between changes in lung function measures and changes in PROs meant that anchors based on these PROs would be inadequate for determining MICPs of lung function measures. It is possible that PROs developed in patients with ILDs, such as the Living with Pulmonary Fibrosis questionnaire [23, 24], may be better correlated with lung function, but this requires further investigation.

In conclusion, these analyses of data from the IPF-PRO Registry suggest that in patients with IPF, there is only a weak relationship between changes in lung function and changes in PROs. Our findings underscore the importance of including PROs in clinical trials, given that lung function changes may not correspond to patients’ experience of their disease.