The Reliability And Validity Of Rowland Universal Dementia Scale In Chinese And Its Comparison With MMSE

Research Purpose:1.To explore the reliability and validity of the Chinese version of the Rowland Universal Dementia Scale(RUDAS)in mild and moderate dementia population in China,and assess its credibility in dementia screening.So as to provide a new screening tool for dementia screening in China.2.We compared the sensitivity,specificity and diagnostic value of RUDAS and MMSE for dementia screening,so as to provide a theoretical foundation for medical staff to choose an accurate and appropriate screening tool.Research Methods:1.Reliability and validity of Chinese version of Rowland Universal Dementia ScaleTotally 358 dementia patients were selected from 2 nursing homes and 3 community health service centers in Huzhou ranged from December in 2018 to June in 2019 by convenient sampling method.The general information questionnaire,MMSE and Chinese version of RUDAS were used to investigate the patients and evaluate the reliability and validity.We used SPSS21.0 and AMOS21.0 to analyze the data.Item analysis was used for item differentiation test,structural validity(including exploratory factor analysis and confirmatory factor analysis)were used for validity test.Cronbach's coefficient?and retest reliability were used for reliability test.2.Comparative of Chinese version of RUDAS and MMSE test performanceWe conducted a convenient sampling to select patients with normal cognitive function and mild or moderate dementia in a psychiatric hospital,three nursing homes and a community health service center in Huzhou ranged from June to December in 2019.A geriatric psychiatry director and two neurologists constituted a panel of experts.Dementia and normal cognitive function were diagnosed by a team of experts.The survey tools included:general information questionnaire,Chinese version of RUDAS and MMSE.SPSS21.0 and MedCalc?software were used for data analysis.And the multivariate linear regression analysis was used for analyzing the effects of demographic variables on RUDAS and MMSE scores.The results of diagnosis were taken as the gold standard,and the scores of RUDAS and MMSE were used as the independent variables to draw the ROC curve.We also compared the difference of the AUC between them.The sensitivity and specificity between the two scales were compared by McNemar test.Research Results:1.The results of item analysis showed that the differences of all items were statistically significant(P<0.01),and the t values of all items were>3.00,indicating that each item had a good discrimination.The Pearson correlation coefficients between the scores of each dimension and the total score of the Chinese version of RUDAS scale was 0.372~0.622,and the correlation coefficients of every dimensions were greater than 0.3,indicating that the content of each item in Chinese version of RUDAS scale could better reflect the measurement topic.Exploratory factor analysis extracted 6 factors and factor loads of each item ranged from 0.594 to 0.960,which contributed to 62.584%variance totally.Six factors were named as visual-spatial orientation,imitation and judgment,visual structure,delayed recall,verbal communication skills 1,verbal communication skills 2.Compared with the original scale,the dimension of verbal communication reduced dimension.In the confirmatory factor analysis,?~2/df=1.031,GFI=0.882,AGFI=0.853,CFI=0.996,TLI=0.996,RMSEA=0.014,all the indicators were in adaption standard range.Cronbach's?coefficient was 0.793,and Cronbach's?coefficients for six dimensions ranged from 0.503 to 0.930.The test-retest reliability for the Chinese version of RUDAS was 0.91.2.The general demographic data showed that there were no significant differences in gender,occupation,marital status,history of dementia between the normal group and the dementia group(P>0.05),and there were significant differences in age,education level and residence between the normal group and the dementia group(P<0.05).The ROC curve analysis results showed that the AUC value of RUDAS was 0.894,the optimal cut-off value was 21,the sensitivity was 84.49%,and the specificity was 84.45%.The ROC curve analysis results showed that the AUC value of MMSE was 0.862,the optimal cut-off value was 23,the sensitivity was 77.54%,and the specificity was 85.45%.We compared the ROC curve of Chinese version of RUDAS and MMSE and found that the difference of AUC value between them two was not statistically significant (P>0.05).The sensitivity of Chinese version of RUDAS scale was significantly higher than that of MMSE(P<0.05),and the specificity of Chinese version of RUDAS scale and MMSE was not significantly different(P>0.05).The results of multiple regression analysis showed that education level affected scores of RUDAS,and its overall contribution rates to RUDAS was 18%(F=16.453,P=0.000).Education level and age affected the score of MMSE,with an overall contribution rates of 22.1%to MMSE(F=20.937,P=0.000).Research Conclusion:1.There are stable reliability and validity of Chinese version of RUDAS in dementia screening,which can provide guidance and reference for further diagnostic research and application practice of this scale.2.The comparative study shows that the diagnostic value of Chinese version of RUDAS is equivalent to MMSE,the sensitivity of Chinese version of RUDAS is better than MMSE,and the bias of demographic variables on Chinese version of RUDAS is less.Therefore,the Chinese version of RUDAS is recommended for screening of dementia patients in China.