Detection of cognitive impairment using self-rated AD8 and informant-reported AD8.

<bold>Background/purpose: </bold>Screening of dementia can help to initiate proper management of the disorder. The use of the Ascertain Dementia 8-item Questionnaire (AD8) in screening has been promoted in Taiwan recently. The purpose of this study was to compare the psychometric properties and appropriateness of informant-reported and self-rated AD8 in cognitive impairment screening in Taiwan.<bold>Methods: </bold>The AD8 were administered to 153 participants and their informants recruited from two neurology out-patient clinics. The discriminative abilities for early cognitive impairment [Clinical Dementia Rating scale (CDR) 0.5 and 1] of informant-based and self-rating AD8 were determined and compared with their areas under the receiver operating curve. κ coefficients representing the agreement between self-rated and informant-reported AD8 scores were also calculated.<bold>Results: </bold>Participants and their informants were aged 76.9 years and 56.0 years on average, respectively. Only informant-reported AD8 was significantly associated with CDR level (Spearman ρ=0.469, p<0.001) and Cognitive Abilities Screening Instrument score (Spearman ρ=-0.458, p<0.001). The item-by-item agreements between self-rated and informant-reported AD8 were poor (κ coefficients: -0.030 to 0.206). The area under the receiver-operator characteristic curve was 0.59 for self-rated AD8 scores, and 0.77 for informant-reported AD8 scores, indicating that the discriminating ability of AD8 scores between CDR 0 and CDR 0.5 or greater is better when reported by informant than when rated by self.<bold>Conclusion: </bold>Informant-rated AD8 gave more accurate screening results than self-reported AD8 in an out-patient clinic setting. [ABSTRACT FROM AUTHOR]