| Abstrakt: |
Background: The Cognitive Domains and Functional Assessment Questionnaire (CDFAQ) assesses cognitive domains and functional decline based on the DSM‐5 criteria for Neurocognitive Disorders. It has two versions, one for the patient (CDFAQ‐PV) and another for the informant (CDFAQ‐IV) assessing: Complex Attention (CA), Executive Functions (EF), Learning and Memory (LM), Language (L), Perceptual‐Motor Domain (PM) and Social Cognition. Recently, the CDFAQ was translated and adapted to English. Method: To describe the performance of the CDFAQ‐PV and CDFAQ‐IV assessing Brazilian older adults compared with the Informant Questionnaire on Cognitive Decline in the Elderly Long Version – IQCODE‐LV. This study was approved by the ethical committee of the UFMG and the CDFAQ was applied to 132 subjects in the University Hospital. The CDFAQ‐PV was applied to 66 older adults and the CDFAQ‐IV, as well as the IQCODE‐LV, to their 66 respective informants. Non‐parametric correlation with IQCODE‐LV and accuracy assessment by Receiver Operating Characteristic (ROC) with cutoff for dementia were performed. Result: The median age of the patients was 70.5 years (IQR = 13.5) and 5 years of schooling (IQR = 7), while the informants had a median age of 55 years (IQR = 24.5) and 12 years of schooling (IQR = 8). The CDFAQ‐IV showed a strong positive correlation (Rho:0.708, p<0.001) to IQCODE‐LV. Interestingly, the LM Domain had an even higher positive correlation (Rho:0.758, p<0.001), while the CDFAQ‐PV had moderate (Rho:0.405, p = 0.001) and its LM Domain had a weak (Rho:0.348, p = 0.004) positive correlations. The CDFAQ‐IV had a good area under the curve (0.871) as well as its LM (0.861), L (0.832), EF (0.854) Domains, while SC, PM were considered only fair. The cutoff values were assessed by Youden's J statistics: CDFAQ‐IV Total score ≥66.5 (SE:88.9%; SP:89.6%) and the LM Domain score ≥11,5 (SE:88.9; SP:77.1%) had the best performances. Conclusion: The strong correlation of the CDFAQ‐IV and its LM Domain to the IQCODE‐LV, as well as the good area under the curve, points a good accuracy for assessing neurocognitive disorders when applied to their informants. Nevertheless, this is still preliminary data, and more assessments in different scenarios, as well as, comparing with objective cognitive assessment is still needed. [ABSTRACT FROM AUTHOR] |
