Minimal Clinically Important Difference in Quality of Life Scores for Patients with Heart Failure and Reduced Ejection Fraction

AIMS: While the associations of health-related quality of life scores in heart failure (HF) e.g. the Kansas City Cardiomyopathy Questionnaire (KCCQ) with clinical outcomes are well established, their interpretation in the context of what magnitudes of change are clinically important to patients is less clear. The main objective of this study was to correlate the changes in the KCCQ and Patient Global Assessment (PGA) in patients with HF with reduced ejection fraction (HFrEF) to determine minimal clinically important difference (MCID). METHODS AND RESULTS: We analysed data from 459 patients of the FAIR-HF trial. Both KCCQ and PGA were assessed at 4 and 24 weeks after enrolment. An anchor-based approach was used to calculate MCID at week 4 and 24. PGA was chosen as the clinical anchor against which changes in the KCCQ scores were calibrated. For each category of change in PGA, the corresponding differences were calculated by the mean scores of various domains of KCCQ along with 95% confidence intervals (CIs). There was fair correlation between PGA and changes in overall summary scores (OSS) (r = 0.31; P < 0.001), clinical summary scores (CSS) (r = 0.36; P < 0.001) and physical limitation scores (PLS) (r = 0.31; P < 0.001) from baseline to week 4. KCCQ OSS, CSS and PLS MCID for 'little improvement' at week 4 were 3.6 (1.0-6.2), 4.5 (1.8-7.2) and 4.7 (1.4-8.0) points, respectively. OSS, CSS and PLS MCID for 'little improvement' at week 24 were 4.3 (0.2-8.4), 4.5 (0.5-8.5) and 4.0 (-0.9-9.0) points, respectively. CONCLUSION: The MCID threshold for KCCQ score was generally consistent and numerically lower than the threshold of 5-point change considered for clinical outcome prognosis and were stable between 4 and 24 weeks. This suggests that even changes smaller than the traditional 5-point improvements in KCCQ may be clinically meaningful. Also, these results can aid in the clinical interpretation of patient-reported outcomes, and better endpoint selection in future studies.