Cross-cultural Adaptation Of Quality Of Life-alzheimer’s Disease And Alzheimer’s Disease Knowledge Scale Chinese Version

Objectives: To support the use of the Chinese version of Quality of Life-Alzheimer’s Disease(QOL-AD) and Alzheimer’s Disease Knowledge Scale (ADKS) as a generic instrument tomeasure QOL of AD and knowledge about AD in mainland China.Methods: Subjects of QOL-AD were patients with AD from three communities and three ThirdLevel1st Class hospitals,200and185cases surveyed for initial and final scale, respectively.Subjects of ADKS were elderly residents aged60years or older from three communities,206and173surveyed for initial and final scale, respectively. Item response theory was used to assesseach item of the two measures. Cross-cultural adaptations for the final scales were evaluated byreliability, validity, confirmatory factor analysis, item response theory, and the ceiling and flooreffects.Results:(1) Items analysis for initial scales: Discrimination and difficulty index of few items ofinitial QOL-AD were not within the range of the theoretical values. ICC showed ideally and theinformation curve covered a large area with high information. Discrimination and difficultyindex of21items of initial ADKS were within the range of the theoretical values, while for theothers either discrimination or difficulty index or both were beyond the range. Less than half ofitems had the ideal ICC and the standard error of the information curve was large, whichindicated less information.(2) Cross-cultural adaptation for final scales: QOL-AD: Analysisof dispersion tendency indicated the variances of all items were greater than one. Reliabilityevaluation: Cronbach’s α coefficients were0.95and0.91for patient report and caregiver report,respectively. The split half reliability coefficients were0.98and0.90for patient report andcaregiver report, respectively. Validity evaluation: Correlation coefficients between each itemand total score were0.65~0.87for patient report, while0.53~0.86for caregiver report.Correlation coefficients between each item and item13were0.49~0.69for patient report, while0.29~0.74for caregiver report. Item response theory: Discrimination index of the final versionwere1.45~3.56and the difficulty index of each item were-3~3with no converse threshold. ICCof each item was perfect and showed high information. ADKS: The replies of each item wereshown in accuracy rate which were almost50%(45.1%~57.8%). Reliability evaluation:Cronbach’s α coefficient and split-half reliability coefficient of the final version of ADKS was0.87and0.86, respectively. Validity evaluation: Overall discrimination validity was assessed by analysis of variance (F=66.16，P<0.001). Chi-square tests were used to evaluatediscrimination validity of each item (chi-square statistics all greater than3.58and P values allless than0.05). Confirmatory factor analysis: Standard factors loads were0.12~0.57. Modelfitting indexχ2/df=1.15, GFI=0.85, AGFI=0.82, NFI=0.82, IFI=1.01, CFI=1.00, RMR=0.13,RMSEA=0.029. Item response theory: Discrimination and difficulty index of the items wereall within the range of the theoretical values. ICC showed ideally with high information. Ceiling and floor effects: We did not found greater than15%of the respondents to achieve thelowest or highest possible score with the lowest and highest score4and29, respectively.Conclusion: The reliability, validity and the items of the Chinese version of the QOL-AD andADKS have good psychometric properties and can be used as a measure of generic QOL inpatients with AD and knowledge about AD in mainland China.