Objective: To evaluate the psychometric properties of the Minnesota Living with Heart Failure Questionnaire (MLHF-Q) in patients with atrial fibrillation.
Design: A prospective study of the patients who underwent DC electrical cardioversion.
Setting: Clinics of cardiology and thoracic surgery of the University Hospital in Groningen, the Netherlands.
Main outcome measures: The disease-specific MLHF-Q and generic measures of quality of life were administered. The sensitivity to change over time was tested with effect sizes (ES). Internal consistency of MLHF-Q scales was estimated with Cronbach's alpha. To evaluate the construct validity multitrait-multimethod analysis was applied. The 'known group validity' was evaluated by the comparison of mean scores and effect sizes between two groups of the New York Heart Association (NYHA) classification (NYHA I versus II-III). Stability of MLHF-Q scales was estimated in a subgroup of patients who remained stable. Perfect congruence analysis and factor analysis were applied to confirm the a priori determined structure.
Results: Cronbach's alpha was > or = 0.80 of the MLHF-Q scales. Perfect congruence analysis (PCA) showed that the results resemble quite well the a priori assumed factor structure. Multitrait-multimethod analysis showed convergent validity coefficients ranging from 0.59 to 0.73 (physical impairment dimension) and 0.39 to 0.69 (emotional dimension). The magnitude of change can be interpreted as medium (ES = 0.50). The results of a 'test-retest' analysis in a stable group can be valued as satisfactory for the MLHF-Q scales (Pearson's r > 0.60). The physical dimension and the overall score of the MLHF-Q discriminated significantly between the NYHA I and II-III groups (p < 0.001) with large effect sizes (ES > 1.0).