The aim of the current study was to test the properties of a scale especial
ly developed for use in epidemiological surveys in low sociocultural popula
tions, and to determine whether this scale can be used in a clinical popula
tion. Material and Methods: Two hundred fifty subjects took part; 150 were
controls and 100 had dementia of various types according to DSM-IV and NINC
DS-ADRDA criteria. Results: Cronbach's alpha was equal to .78. In the group
of people under 75 years old, sensitivity did not exceed 90%. Specificity
was over 90% at the level 4/5. The respective scores for the group of peopl
e over 74 were 6/7 and 3/4. Discussion: The Epidemiological Dementia Index
seems to be less powerful when used in a clinical population. This provides
further evidence that it is not suitable to use the same instruments in bo
th epidemiological studies and clinical practice.