Zhongguo quanke yixue (Feb 2022)
Reliability and Validity of the Chinese Version of the Partners in Health Scale in Patients with Chronic Heart Failure
Abstract
BackgroundChronic heart failure (CHF) is a common cardiovascular disease. Improving the self-management ability of CHF patients will contribute to quality of life improvement and reduction of rehospitalization and mortality rates. The Partners in Health (PIH) Scale is a measure designed by Flinders University, Australia, to assess the generic knowledge, attitudes, behaviors, and impacts of self-management in chronic disease patients, and is mainly used to assess the implementation effect of self-management projects in chronic disease patients.ObjectiveTo translate the PIH Scale into Chinese, then test the reliability and validity of the Chinese version in CHF patients, providing CHF patients with a tool for precisely assessing their self-management abilities.MethodsThe PIH was translated into Chinese with the guidance of the Brislin's translation model, then was revised according to the results of the review of a panel of experts, and a pre-test, and then the Chinese version of PIH (C-PIH) was developed. The demographic questionnaire, C-PIH, and Minnesota Living with Heart Failure Questionnaire (MLHFQ) were used in two surveys (one was conducted between April and June 2010, and another between April and June 2011) with 410 CHF patients selected from two grade A tertiary hospitals in Beijing using convenience sampling. Measurement of ceiling and floor effects, and item-total correlation were used for item analysis. Expert evaluation was used to evaluate the content validity analysis. Spearman's rank correlation coefficient was used to measure the criterion-related validity. KMO test, Bartlett's test of sphericity, exploratory factor analysis and confirmatory factor analysis were used for construct validity analysis. Monofactor analysis was used for validity analysis of known-groups. Reliability analysis was estimated by using the Cronbach's α.ResultsItem analysis indicated that only item 3 (level of adhering to medication) of the C-PIH showed ceiling effect. Item-total correlation coefficients of the scale ranged from 0.424 to 0.761 (P<0.001) . The scale-level content validity index of the scale was 0.966. Item-level content validity indices ranged from 0.800 to 1.000. C-PIH was positively correlated with MLHFQ in terms of total score (rs=0.200, P<0.05) . The KMO value was 0.872 and Bartlett's test of sphericity was χ2=1 139.142 (P<0.001) , indicating that the sample size was appropriate for factor analysis. By exploratory factor analysis, 3 factors with an eigenvalue greater than 1.000 were extracted, including knowledge (7 items) , coping (3 items) and adherence (2 items) , explaining 66.514% of the total variance. The loadings of items on each factor ranged from 0.571 to 0.869. The original model fit indices did not reach the critical value. After adding the suggested covariance correlation between errors-in-variables e1 and e2, e6 and e7, the fitting indices of the modified model were acceptable (χ2/df=2.393, RMSEA=0.0851, CFI=0.968, NFI=0.953, NNFI=0.963, GFI=0.905, AGFI=0.854, RFI=0.932, IFI=0.966) . Known-groups analysis demonstrated that the C-PIH total score varied significantly by level of education, economic income, NYHA class, and treatment (inpatient or outpatient) in CHF patients (P<0.001) . Good internal consistency was indicated with a scale Cronbach's α of 0.890, and three factors' (knowledge, coping and adherence) Cronbach's α of 0.894, 0.807, and 0.511.ConclusionThe C-PIH exhibited good reliability and validity, which may be used as a general self-management assessment tool in patients with CHF.
Keywords
