ABSTRACT
The 12-item version of the Kansas City Cardiomyopathy Questionnaire (KCCQ-12) was originally developed for patients with heart failure but has been used and tested among patients with severe aortic stenosis (AS) who underwent transcatheter aortic valve implantation. Whether the instrument is suitable for patients with AS who underwent surgical aortic valve replacement (SAVR) is currently unknown. Thus, we aimed to investigate the psychometric properties of the KCCQ-12 before and after SAVR among patients with severe AS. We conducted a prospective cohort of 184 patients with AS who completed the KCCQ-12 and the EuroQol 5 Dimension 5 Levels before and 4 weeks after surgery. Construct validity was investigated with hypothesis testing and an analysis of Spearman's correlation between the two instruments. Structural validity was investigated with explorative and confirmatory factor analyses and reliability with Cronbach's α. All analyses were conducted on data from the two time points (preoperatively and four weeks after surgery). The hypothesis testing revealed how the New York Heart Association class was significantly correlated with the preoperative KCCQ-12 total score (higher New York Heart Association class, worse score). A longer length of hospital stay and living alone were significantly associated with poorer postoperative KCCQ-12 total score. KCCQ-12 and EuroQol 5 Dimension 5 Levels were moderately correlated in most domains/the total score/Visual Analogue Scale score. Principal component analyses revealed two 3-factor structures. The confirmatory factor analyses did not support the original model at any time point. Cronbach's α ranged from 0.22 to 0.84 in three preoperative factors and from 0.39 to 0.76 in the postoperative factors. The total Cronbach's α was 0.83 for the suggested preoperative 3-factor model and 0.83 for the postoperative model. In conclusion, the Danish version of the KCCQ-12 tested in a population of patients with AS who underwent SAVR appears to have acceptable construct validity, whereas structural validity cannot be confirmed for the original four-factor model. Overall reliability is good.
