Pulmonology · Medical Literature
Validation and minimum important difference of the chronic respiratory disease questionnaire in patients with interstitial lung disease.
Abstract
The Chronic Respiratory Disease Questionnaire (CRQ) was developed to measure health-related quality of life (HRQoL) in patients with chronic obstructive pulmonary disease (COPD), with a minimum clinically important difference (MCID) of 0.5 points per item. It has been used in clinical trials of patients with interstitial lung disease (ILD), however its validity and MCID in this population is unknown. Is the CRQ a valid measure of HRQoL in people with ILD and what is the MCID in this population? Data from a multicentre randomised controlled trial of pulmonary rehabilitation (PR) in 142 patients with ILD were analysed. Concurrent validity with the St George's Respiratory Questionnaire for IPF (SGRQ-I), known groups validity between patients of varying TLCO and GAP Index, and responsiveness to change after PR were assessed. MCID was calculated using anchor- and distribution-based methods. Analyses were repeated for the idiopathic pulmonary fibrosis (IPF) subgroup. There was a strong association between CRQ and SGRQ-I scores (r = -0.786, p < 0.05). CRQ scores differed significantly between patients in the highest and lowest TLCO quartile (effect sizes -0.49 to -0.64) and between GAP Index groups (p < 0.05), except for the emotional function domain. The CRQ was responsive to PR (effect size -0.53). The MCID ranged from 0.45 to 0.65 points per item. Results were similar in the IPF subgroup. The CRQ is a valid and responsive measure of HRQoL in patients with ILD, including those with IPF, with an MCID of around 0.5 points per item. These findings support the use of the CRQ in ILD.
Clinical relevance
Is the Chronic Respiratory Disease Questionnaire (CRQ) a valid and responsive measure of health-related quality of life (HRQoL) in patients with interstitial lung disease (ILD), and what is its minimum clinically important difference (MCID) in this population? __SNAPSHOT__ {"studyDesign":"Secondary analysis of data from a multicentre randomised controlled trial (pulmonary rehabilitation vs. usual care); psychometric validation study","setting":"Three tertiary hospitals in Australia","sampleSize":"n=142 total ILD patients (intervention: 74, usual care: 68); IPF subgroup n=61","population":"Adults with a confirmed diagnosis of ILD (including IPF n=61, connective tissue disease-associated ILD n=23, dust-related ILD n=22, and other ILD n=36) who were clinically stable at enrolment","intervention":"8 weeks of supervised exercise training (pulmonary rehabilitation); CRQ and SGRQ-I assessed at baseline, 9 weeks, and 6 months","comparator":"Usual care","primaryOutcome":"Psychometric validation of the CRQ: concurrent validity (vs. SGRQ-I), known-groups validity (by TLCO quartile and GAP Index), responsiveness to pulmonary rehabilitation, and MCID estimation using anchor- and distribution-based methods","followUp":"9 weeks (primary) and 6 months","fundingSource":"Not reported"}
Key findings
Concurrent validity: The CRQ total score demonstrated a strong negative correlation with the SGRQ-I total score (r = −0.786, p < 0.05), consistent with the instruments measuring similar constructs in opposite scoring directions. Associations between individual CRQ and SGRQ-I domain pairs ranged from fair to moderate. Known-groups validity: CRQ scores differed significantly between patients in the highest and lowest TLCO quartiles (effect sizes −0.49 to −0.64) and between GAP Index severity groups (p < 0.05), with the exception of the emotional function domain, which did not discriminate between GAP Index groups. Responsiveness: The CRQ was responsive to change following pulmonary rehabilitation with an effect size of −0.53. MCID: Using distribution-based methods, the MCID was 0.48 points per item by effect size and 0.45 points per item by standardized response mean. Domain-specific MCIDs derived from these methods ranged from 0.54 to 0.65 points per item. The overall MCID range across anchor- and distribution-based approaches was 0.45 to 0.65 points per item, converging on approximately 0.5 points per item. Findings were similar in the IPF-only subgroup. NNT/NNH: not applicable to this psychometric validation study. No adverse events were reported as part of this instrument validation analysis.
Practice implications
This study provides the first prospective, multicentre evidence that the CRQ is a valid and responsive patient-reported outcome measure (PROM) for HRQoL assessment in ILD, including the IPF subgroup. The instrument demonstrated acceptable concurrent validity against the disease-specific SGRQ-I, differentiated between patients of varying physiological disease severity (by TLCO and GAP Index), and was sensitive to change following pulmonary rehabilitation. Importantly, the empirically derived MCID of approximately 0.5 points per item aligns with the previously established MCID in COPD, suggesting that the threshold already familiar to clinicians and trialists can be applied in ILD without modification. However, several important caveats apply: the emotional function domain did not demonstrate known-groups validity by GAP Index, suggesting it may be less sensitive to disease severity in ILD; the study population was limited to three Australian tertiary centres and was predominantly drawn from an exercise training trial, which may not represent the broader ILD population seen in general respiratory practice or in trials of antifibrotic therapies. The ILD cohort was heterogeneous (IPF, CTD-ILD, dust-related, and other diagnoses), and while results were reproduced in the IPF subgroup, statistical power for subgroup analyses across other ILD subtypes was limited. Generalizability to non-ambulatory, severely impaired, or non-English-speaking patients is uncertain. Only the abstract was available for this review, limiting assessment of full methodological detail including anchor selection, ROC curve performance, and handling of missing data. __LIMITATIONS__ Secondary analysis of an RCT not originally designed for psychometric instrument validation; sample size (n=142) may be insufficient for some psychometric methods, particularly subgroup analyses.,Single-country, tertiary-centre sample from three Australian hospitals limits generalizability to community-based or international ILD populations.,Heterogeneous ILD population: while IPF subgroup analyses were performed, power for other ILD diagnostic subgroups (CTD-ILD, dust-related) was not reported.,The emotional function domain did not demonstrate known-groups validity by GAP Index, representing a domain-specific limitation of the instrument in ILD.,Responsiveness was assessed only in patients who completed the exercise training intervention (n=49 completers of 74 randomised); attrition and per-protocol analysis may introduce selection bias.,Anchor-based MCID methods rely on the quality and responsiveness of the chosen anchor (Global Rating of Change and 6MWD); anchor performance characteristics (e.g., ROC AUC) were not detailed in the available abstract.,Funding source not reported; potential conflicts of interest cannot be assessed.,Follow-up of 6 months limits conclusions about longer-term construct stability or responsiveness to disease progression.,Only an abstract was available for this review; full methodological details, including covariate adjustment and imputation methods, could not be evaluated. __BOTTOM_LINE_CATEGORY__ potentially_practice_informing __BOTTOM_LINE__ This study establishes for the first time that the CRQ is a valid and responsive HRQoL instrument in ILD and IPF, with an MCID of approximately 0.5 points per item—identical to its established COPD threshold. Clinicians and clinical trialists conducting ILD pulmonary rehabilitation research may use the CRQ with confidence using the existing 0.5-point-per-item benchmark. Caution is warranted when interpreting the emotional function domain, which did not discriminate between GAP Index severity groups, and independent replication in larger, more diverse ILD cohorts is desirable before broad cross-disease application.