Premium
Reliability and validity of revised Turkish version of Mini Mental State Examination (rMMSE‐T) in community‐dwelling educated and uneducated elderly
Author(s) -
Keskinoglu Pembe,
Ucku Reyhan,
Yener Görsev,
Yaka Erdem,
Kurt Pınar,
Tunca Zeliha
Publication year - 2009
Publication title -
international journal of geriatric psychiatry
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.28
H-Index - 129
eISSN - 1099-1166
pISSN - 0885-6230
DOI - 10.1002/gps.2252
Subject(s) - cronbach's alpha , turkish , receiver operating characteristic , kappa , psychology , gerontology , test (biology) , inter rater reliability , medicine , demography , psychometrics , clinical psychology , mathematics , developmental psychology , rating scale , paleontology , linguistics , philosophy , geometry , sociology , biology
Objective To evaluate the reliability and validity of the revised Turkish version of Mini Mental State Examination (rMMSE‐T) in educated and uneducated community‐dwelling elderly, to re‐organize the present Turkish version of MMSE and to determine cut‐off point of the revised test. Methods This cross‐sectional and analytical study involved totally 490 elderly subjects selected by cluster sampling method. Receiver operating characteristic (ROC) analysis, kappa analysis and Cronbach's α coefficients were used for statistical analysis. Results Areas under ROC curve in educated and uneducated elderly were found as 0.953 and 0.907. Cut‐off point of 22/23 of rMMSE‐T in educated elderly had the highest sensitivity (90.9), specificity (97.0) and positive likelihood ratio (30.3), whereas cut‐off point of 18/19 of the test in uneducated elderly had the highest sensitivity (82.7), specificity (92.3) and positive likelihood ratio (10.7). The Cronbach's α values of the rMMSE‐T for educated and uneducated elderly were higher than 0.7 (sign of good internal consistency of the test). A significant correlations between intrarater and interrater test–retest in educated elderly subjects were observed (0.966 ( p = 0.000); 0.855 ( p = 0.000), respectively), and also in uneducated elderly (0.988 ( p = 0.000); 0.934 ( p = 0.000), respectively). Kappa value of the test in educated and uneducated elderly showed a perfect agreement interraters (1.000) and a substantial agreement in intraraters (1.000, 0.784; 0.826, 0.656, respectively). Conclusion rMMSE‐T had a high reliability and validity. It will be more appropriate to use the revised test and the new cut‐off point for the diagnosis and screening of dementia among community‐dwelling Turkish elderly population. Copyright © 2009 John Wiley & Sons, Ltd.