Child behavior checklist (CBCL) como instrumento de rastreio para transtorno do espectro autista: estudo psicométrico
Carregando...
Data
2024-03-15
Autores
Kitanishi, Norton Yoshiaki [UNIFESP]
Orientadores
Caetano, Sheila Cavalcante [UNIFESP]
Tipo
Dissertação de mestrado
Título da Revista
ISSN da Revista
Título de Volume
Resumo
Objetivo: A identificação precoce do Transtorno do Espectro Autista (TEA) por meio do rastreamento é crucial para o diagnóstico e intervenção precoces, mas isso pode ser desafiador em países de baixa e média renda (PBMR) devido ao alto custo das ferramentas de diagnóstico e triagem. O Child Behavior Checklist (CBCL) 1.5-5 tem sido estudado como um instrumento de rastreio de nível 1 para TEA, utilizando as subescalas problemas do espectro do autismo e síndrome retraimento, mas há uma falta de validade de construto, especialmente em PBMR. Este estudo tem como objetivo fornecer validade de construto para o CBCL 1.5-5 como instrumento de triagem para TEA em uma amostra representativa de base populacional de 1.292 crianças brasileiras de 3 a 5 anos e uma amostra de 70 TEA de 1 a 5 anos. Métodos: A análise fatorial confirmatória foi utilizada para avaliar a qualidade do ajuste; e a força da correlação entre os itens e o TEA. Resultados: Os achados indicaram um bom ajuste ao modelo e alta confiabilidade na amostra total para o modelo de problemas do espectro do autismo (SB-X2(54) = 2089.148, p =1.00, CFI = 0,96, TLI = 0,95, RMSEA = 0,037, SRMR = 0,041, ω = 0,869), exceto para um item com baixa carga fatorial. O modelo de síndrome retraimento também mostrou bom ajuste e alta confiabilidade (SB-X²(20) = 1508.647, p = 1.00; CFI= 0,974, TLI= 0,963, RMSEA= 0,034, RMRR= 0,033, ω= 0,776), com um item com baixa carga fatorial. Conclusão: as escalas Problemas do Espectro do Autismo e Retraimento da CBCL 1.5-5 podem ser rastreadoras válidas do nível 1 do TEA na população brasileira, pois apresentaram boa consistência interna e validade de construto.
Purpose: Early identification of Autism Spectrum Disorder (ASD) through screening is crucial for early diagnosis and intervention, but this can be challenging in low and middle-income countries (LMIC) due to the high cost of diagnostic and screening tools. The Child Behavior Checklist (CBCL) 1.5-5 has been studied as a level 1 screening tool for ASD, using the autism spectrum problems and withdrawn syndrome subscales, but there is a lack of construct validity, especially in LMIC. This study aims to provide construct validity for the CBCL 1.5-5 as a screening tool for ASD in a representative population-based sample of 1,292 Brazilian children aged 3 to 5 years and a sample of 70 ASD aged 1 to 5 years. Methods: Confirmatory factor analysis was used to evaluate the goodness of fit; and the strength of the correlation between the items and ASD. Results: Findings indicated a good model fit and high reliability in the total sample for autism spectrum problems model (SB-X2(54) = 2089.148, p =1.00; CFI= 0.96, TLI= 0.95, RMSEA= 0.037, SRMR= 0.041, ω= 0.869), except for one item with a low factor loading. The withdrawn syndrome model also showed a good fit and high reliability (SB-X²(20) = 1508.647, p = 1.00; CFI= 0.974, TLI= 0.963, RMSEA= 0.034, SRMR= 0.033, ω= 0.776), with one item with a low factor loading. Conclusion: the Autism Spectrum problems and Withdrawn CBCL 1.5-5 scales can be valid ASD level 1 screeners in the Brazilian population, as they showed good internal consistency and construct validity.
Purpose: Early identification of Autism Spectrum Disorder (ASD) through screening is crucial for early diagnosis and intervention, but this can be challenging in low and middle-income countries (LMIC) due to the high cost of diagnostic and screening tools. The Child Behavior Checklist (CBCL) 1.5-5 has been studied as a level 1 screening tool for ASD, using the autism spectrum problems and withdrawn syndrome subscales, but there is a lack of construct validity, especially in LMIC. This study aims to provide construct validity for the CBCL 1.5-5 as a screening tool for ASD in a representative population-based sample of 1,292 Brazilian children aged 3 to 5 years and a sample of 70 ASD aged 1 to 5 years. Methods: Confirmatory factor analysis was used to evaluate the goodness of fit; and the strength of the correlation between the items and ASD. Results: Findings indicated a good model fit and high reliability in the total sample for autism spectrum problems model (SB-X2(54) = 2089.148, p =1.00; CFI= 0.96, TLI= 0.95, RMSEA= 0.037, SRMR= 0.041, ω= 0.869), except for one item with a low factor loading. The withdrawn syndrome model also showed a good fit and high reliability (SB-X²(20) = 1508.647, p = 1.00; CFI= 0.974, TLI= 0.963, RMSEA= 0.034, SRMR= 0.033, ω= 0.776), with one item with a low factor loading. Conclusion: the Autism Spectrum problems and Withdrawn CBCL 1.5-5 scales can be valid ASD level 1 screeners in the Brazilian population, as they showed good internal consistency and construct validity.
Descrição
Citação
KITANISHI, Norton Yoshiaki. Child behavior checklist (CBCL) como instrumento de rastreio para transtorno do espectro autista: estudo psicométrico. 2024. 81 f. Dissertação (Mestrado Psiquiatria e Psicologia Médica)- Escola Paulista de Medicina, Universidade Federal de São Paulo, São Paulo, 2024.