RELIABILITY AND VALIDITY OF THE SPANISH VERSION OF THEGMS–AGECAT PACKAGE FOR THE ASSESSMENT OF DEMENTIA AND COGNITIVE DISTURBANCES

This report confirms the reliability and validity of the Geriatric Mental State–Automated Geriatric Examination for Computer Assisted Taxonomy (GMS–AGECAT) package in a Spanish elderly population. No changes in the original English version are considered necessary. A representative sample (N=1080) of the elderly (65+ years) community in Zaragoza, Spain, stratified by age and sex was assessed. A two-phase screening was designed: phase I (lay interviewers): Spanish versions of GMS–AGECAT and Mini-Mental (MMSE); phase II (psychiatrists,N=324): the same instruments and History and Aetiology Schedule (HAS). Diagnosis: DSM-III-R criteria. Stringent test–retest reliability coefficients were calculated by comparing lay interviewers’ (phase I) vs psychiatrists’ (phase II) ratings. DSM-III-R psychiatric diagnoses in phase II were the gold standards for the validity study. Test–retest reliability coefficients of the cognitive sections of GMS were: case/no case distinction, mean kappa=0.71; dementia/no dementia distinction, mean kappa=0.80. GMS validity coefficients: detection of cases, sensitivity=98.4%; specificity=76.5%. Test–retest reliability coefficients of AGECAT: case/no case distinction, mean kappa=0.59; organic syndrome, dementia, mean kappa=0.68. Validity coefficients of AGECAT (psychiatrists’ outputs): case/no case distinction, sensitivity=90.6%; specificity=89.3%; dementia/no dementia distinction, sensitivity=86.4%; specificity=94.3%.