Informant questionnaire on cognitive decline in the elderly (IQCODE): um estudo de validação

• Autor(a) principal: Sanchez, Maria Angélica dos Santos
• Data de Publicação: 2013
• Tipo de documento: Tese
• Idioma: por
• Título da fonte: Biblioteca Digital de Teses e Dissertações da UERJ
• Texto Completo: http://www.bdtd.uerj.br/handle/1/8599
• topic: Aging; Dementia; Psychometrics; Caregivers; Envelhecimento; Demência; Psicometria; Cuidadores; Cognição na velhice; Distúrbios da memória em idosos; Demência; Envelhecimento; Psicometria; Cuidadores; Questionários; CNPQ::CIENCIAS DA SAUDE::SAUDE COLETIVA

The Informant questionnaire on cognitive decline in the elderly (IQCODE) is an instrument used to screen for cognitive decline in older people, based on the report obtained from a reliable informant. The present study aimed to validate a Portuguese version of the IQCODE to be used in Brazil. The questionnaire was applied in two samples: the first one the clinical sample was made up of patients treated in the ambulatory unit of the Geriatric Department of the Universidade do Estado do Rio de Janeiro; the second one the community sample was formed by community-dwelling individuals who participated in the research protocol of the Frailty in the Brazilian Elderly Study (Fibra-RJ). In the clinical sample the convergent validity and the IQCODE-BR accuracy were evaluated for demential syndrome detection. In order to establish a cutoff point, the area under the ROC curve, the sensitivity, specificity, and also the negative and positive predictive values were analyzed. The same analyses described above were accomplished in the community sample. In order to evaluate the influence of both elderly and informants sociodemographic characteristics, including the emotional condition of the caregivers on the scores of the tests, multivariate analyses employing a logistic regression model were also performed. The statistical program SPSS, version 19.0 was used for data storage and analysis. In order to do factorial scrutiny the one-dimensional structure originally proposed was reassessed. The confirmatory factor analysis was done using weighted least squares means variance adjusted (WLSMV), and further confirmatory factor analyses were done employing exploratory procedures (AFC/E). The adjustment model was evaluated by the indexes root mean square error of approximation (RMSEA), Comparative Fit Index (CFI) and Tucker-Lewis index (TLI). The data was analyzed by the software MPlus7. There were 169 elderly in the clinical sample. The sensitivity and specificity were 83.3% and 80.7%, respectively. The area under the Roc curve was 0.83% (95%IC: 0.766; 0.893) and the best cutoff point to identify cognitive decline was 3.52. As for the community sample, 417 elderly were evaluated and the best cutoff point was 3.26; the sensitivity and the specificity were 89% and 72%, respectively. The area under the ROC curve was 0.88 (95%IC: 0.837; 0.917). In the multivariate analysis, only the informants´ burden levels maintained significant association with the scores of IQCODE-BR. High correlation among factors was observed through the dimensional structure analysis and seven pairs of items were detected as redundant. In an attempt to improve the model, six of these pairs were removed and new analyses were performed. This being done, the correlation among factors disappeared, and the adjustment indexes RMSEA = 0.058 (90%IC: 0.053; 0.064), CFI = 0.983 and TLI = 0.981, were satisfactory. In conclusion, the results of the present study suggest that the IQCODE is an adequate screening tool and a useful aid for the diagnosis of cognitive decline in the elderly. The removal of redundant items seems to make it more refined without losing the capacity to detect the original construct, although these data must be confirmed with further studies.