OBJECTIVES : The Kansas City Cardiomyopathy Questionnaire (KCCQ) has been widely used to measure heart failure (HF) patients’ perception of their health status, symptoms, physical and social function and quality of life. However, as a disease specific measure, the KCCQ is not designed to estimate health-state utilities, precluding its use in cost-effectiveness research. This study used data from the COAPT trial to map KCCQ scores to the SF-6D utility index.

METHODS : Data were obtained from the COAPT trial--a prospective, randomized controlled trial of transcatheter mitral valve repair vs. standard care in 614 patients with HF and secondary mitral regurgitation. Patients completed the SF-36 and the KCCQ at baseline and at 1, 6, 12, and 24 months. We calculated SF-6D utility scores and KCCQ overall summary (KCCQ-OS) scores for both the full and short form (KCCQ-12) for each subject. Using mixed linear models with random effects for subjects, we regressed SF-6D utility scores on the KCCQ-OS and the KCCQ-12, testing for age and sex effects. Model fit was evaluated using conditional AIC, R² and correlation.

RESULTS : Both the KCCQ-OS and the KCCQ-12 score were closely correlated with SF-6D derived health utility (correlation of 0.76 and 0.77 for the KCCQ-OS and KCCQ-12). Model testing revealed a conditional R² of 0.71 for the model using the KCCQ-OS and 0.73 for the KCCQ-12. The addition of age and gender to the models had negligible impact on the results. Based on these data, the final mapping equations were 0.44 + 0.0035*(KCCQ-OS) and 0.45 + 0.0034*(KCCQ-12).

CONCLUSIONS : The KCCQ-OS and the KCCQ-12 are excellent predictors of health state utility in patients with HF, as compared with the SF-6D utility estimates. Further studies are needed to validate these predictive models in other heart failure populations.