Identificação da demência na atenção primária – adaptação e validação de um rastreador (CSID breve) e de um programa para treinamento de Agentes Comunitários De Saúde
Data
2021-04-19
Autores
Machado, Danusa de Almeida [UNIFESP]
Orientadores
Ferri, Cleusa Pinheiro [UNIFESP]
Tipo
Tese de doutorado
Título da Revista
ISSN da Revista
Título de Volume
Resumo
Introdução: O envelhecimento populacional e aumento da expectativa de vida são fenômenos globais e estão associados com o aumento da prevalência e incidência de doenças como as demências. As demências são um importante problema de saúde pública, e são ainda subdiagnosticadas no mundo inteiro. O não reconhecimento dessa condição pode ser um fardo psicossocial e financeiro para famílias e serviços de saúde. Iniciativas que ofereçam respostas a essa questão são urgentes, especialmente no contexto da Atenção Primária. Objetivos: Este trabalho teve por objetivos testar estratégias de identificação da demência na atenção primária, por meio da adaptação e validação de um rastreador (Community Screening Instrument for Dementia – CSID brief) e de adaptação e validação de um Programa para Treinamento de Agentes Comunitários de Saúde (ACS). Métodos: O instrumento para rastreio de demência na comunidade (CSID breve) foi traduzido e validado para a população brasileira, seguindo as etapas: tradução inicial, síntese das versões traduzidas, retrotradução, síntese das traduções reversas em revisão por comitê de especialistas, e reavaliação da versão final pelo comitê revisor. O um Protocolo de Treinamento dos ACS na identificação da demência na comunidade teve duração de oito horas e foi conduzido por uma equipe multidisciplinar, que valorizou a experiência e o conhecimento dos ACS. Foram adotados recursos audiovisuais, relatos de casos e vinhetas clínicas para realização de discussão em grupo sobre o envelhecimento normal e o envelhecimento não normal. Especificamente sobre as demências, foram abordados os temas: definição, evolução, sintomas comportamentais e psicológicos, causas, diagnóstico diferencial, opções de tratamento. Participaram 14 ACS que atuam na Unidade Básica de Saúde (UBS) em São José dos Campos, SP. Após o treinamento, os ACS indicaram 60 pessoas com 65 anos ou mais, que foram por eles consideradas como possíveis casos de demência por um período de seis meses. Dessas, 45 aceitaram participar do estudo. Posteriormente, dos 481 idosos cadastrados na UBS que não foram indicados, foram sorteados 60 nomes e desses, foram entrevistados 40. Foram convidados também os informantes dos idosos participantes, e participaram 75. Resultados: Do total de 85 participantes, 22 (25,9%) foram diagnosticados com demência por um especialista que adotou os critérios diagnósticos do DSM-IV. Desses 22, 20 estavam entre os idosos indicados pelos ACS e foram, portanto, indicações corretas. A taxa de diagnóstico de Demência de acordo com o padrão ouro (diagnóstico clínico/DSM-IV) foi 9,4%. Os parâmetros psicométricos do treinamento foram: Sensibilidade 90,9%, Especificidade 60,3%, VPP 44,4% e VPN 95,0%. Observou-se os seguintes parâmetros do CSID breve na subescala Participante com o ponto de corte ≥5: Sensibilidade 76,5%, Especificidade 88,4%, VPP 72,2%, VPN 90,5%, área da Curva ROC 0,92 (p=0,03; IC 0,86-0,99) e α de Cronbach 0,76. Na subescala do Informante, a pontuação ≥5 apresentou uma Sensibilidade de 77,3%; Especificidade de 96,0%; VPP 89,4%, VPN 90,5%, RV+19,3, área sobre a curva de 0,89, com p=0,05 (IC 0,80-0,98) e α de Cronbach 0,81. O cálculo das propriedades diagnósticas do CSID breve combinado mostrou que o escore ≥7 possui uma Sensibilidade de 94,1% e Especificidade de 80,0%. Os valores Preditivos Positivo e Negativo foram, respectivamente, 66,7% e 97,0%. A razão de verossimilhança positiva foi de 4,7, área da Curva ROC 0,93 (p=0,03; IC 0,87-1,00) e α de Cronbach 0,86. Observou-se que o instrumento rastreador na versão combinada apresenta excelentes propriedades psicométricas, e que o Treinamento dos ACS contribuiu para a identificação de casos novos de demência na comunidade (17,8% do total). Conclusões: O CSID breve e o Treinamento demonstraram ser estratégias válidas, reprodutíveis e sensíveis para uso na identificação da demência na comunidade, atendendo a uma importante questão de saúde pública no contexto da Atenção Primária.
Introduction: Population aging and increased life expectancy are global phenomena and are associated with an increase in prevalence and incidence of diseases such as dementia. Dementias are a major public health problem, and are still underdiagnosed worldwide. Failure to recognize this condition can be a psychosocial and financial burden for families and health services. Initiatives that offer answers to this question are urgent, especially in the context of Primary Care. Objectives: This study aimed to test strategies for identifying dementia in primary care, through the adaptation and validation of a screening Instrument (Community Screening Instrument for Dementia - CSID brief) and the adaptation and validation of a Program for Training Community Health Workers (CHW). Methods: Community Screening Instrument for Dementia (brief CSID) was translated and validated for the Brazilian population, following the steps: initial translation, synthesis of translated versions, back-translation, synthesis of reverse translations under review by an expert committee, and reassessment of the final version by the review committee. The CHW Training Protocol for identifying dementia in the community lasted eight hours and was conducted by a multidisciplinary team that valued the experience and knowledge of the CHW. Audiovisual resources, case reports and clinical vignettes were adopted for group discussion on normal and non-normal aging. Specifically on dementias, the following topics were addressed: definition, evolution, behavioral and psychological symptoms, causes, differential diagnosis, treatment options. 14 CHW who work in a Basic Health Unit (BHU) in São José dos Campos, SP participated. After training, the CHW indicated 60 people aged 65 or over, who they considered as possible cases of dementia for a period of six months. Of these, 45 agreed to participate in the study. Subsequently, of the 481 elderly people registered in that UBS who were not appointed, 60 names were selected and 40 of them were interviewed. Informants of the elderly participants were also invited, and 75 participated. Results: Of the total of 85 participants, 22 (25.9%) were diagnosed with dementia by a specialist who adopted the DSM-IV diagnostic criteria. Of these 22, 20 were among the elderly who were indicated by the CHW and therefore, were correct indications. The rate of Dementia diagnosis according to the gold standard (clinical diagnosis/DSM-IV) was 9.4%. The psychometric parameters of the training were: Sensitivity 90.9%, Specificity 60.3%, PPV 44.4% and NPV 95.0%. The following parameters of the brief CSID were observed in the Participant subscale with the cutoff ≥5: Sensitivity 76.5%, Specificity 88.4%, PPV 72.2%, NPV 90.5%, area under the ROC Curve of 0, 92 (p = 0.03; CI 0.86-0.99) and Cronbach's α of 0.76. In the Informant's subscale, the score ≥5 presented a Sensitivity of 77.3%; 96.0% Specificity; PPV 89.4%, NPV 90.5%; RV + 19.3, area on the curve of 0.88, with p = 0.05 (CI 0.80-0.98) and Cronbach's α of 0.81. The calculation of the diagnostic properties of the combined brief CSID showed that the score ≥7 has a sensitivity of 94.1% and specificity of 80.0%. The positive and negative predictive values were 66.7% and 97.0%, respectively. The positive likelihood ratio was 4.7, area under the ROC Curve 0.93 (p = 0.03; CI 0.87-1.00) and Cronbach's α 0.86. It was observed that the screening instrument in the combined version has excellent psychometric properties, and that the CHW Training contributed to the identification of new cases of dementia in the community (17.8% of the total). Conclusions: The brief CSID and the Training proved to be valid, reproducible and sensitive strategies for use in the identification of dementia in the community, addressing an important public health issue in the context of Primary Care.
Introduction: Population aging and increased life expectancy are global phenomena and are associated with an increase in prevalence and incidence of diseases such as dementia. Dementias are a major public health problem, and are still underdiagnosed worldwide. Failure to recognize this condition can be a psychosocial and financial burden for families and health services. Initiatives that offer answers to this question are urgent, especially in the context of Primary Care. Objectives: This study aimed to test strategies for identifying dementia in primary care, through the adaptation and validation of a screening Instrument (Community Screening Instrument for Dementia - CSID brief) and the adaptation and validation of a Program for Training Community Health Workers (CHW). Methods: Community Screening Instrument for Dementia (brief CSID) was translated and validated for the Brazilian population, following the steps: initial translation, synthesis of translated versions, back-translation, synthesis of reverse translations under review by an expert committee, and reassessment of the final version by the review committee. The CHW Training Protocol for identifying dementia in the community lasted eight hours and was conducted by a multidisciplinary team that valued the experience and knowledge of the CHW. Audiovisual resources, case reports and clinical vignettes were adopted for group discussion on normal and non-normal aging. Specifically on dementias, the following topics were addressed: definition, evolution, behavioral and psychological symptoms, causes, differential diagnosis, treatment options. 14 CHW who work in a Basic Health Unit (BHU) in São José dos Campos, SP participated. After training, the CHW indicated 60 people aged 65 or over, who they considered as possible cases of dementia for a period of six months. Of these, 45 agreed to participate in the study. Subsequently, of the 481 elderly people registered in that UBS who were not appointed, 60 names were selected and 40 of them were interviewed. Informants of the elderly participants were also invited, and 75 participated. Results: Of the total of 85 participants, 22 (25.9%) were diagnosed with dementia by a specialist who adopted the DSM-IV diagnostic criteria. Of these 22, 20 were among the elderly who were indicated by the CHW and therefore, were correct indications. The rate of Dementia diagnosis according to the gold standard (clinical diagnosis/DSM-IV) was 9.4%. The psychometric parameters of the training were: Sensitivity 90.9%, Specificity 60.3%, PPV 44.4% and NPV 95.0%. The following parameters of the brief CSID were observed in the Participant subscale with the cutoff ≥5: Sensitivity 76.5%, Specificity 88.4%, PPV 72.2%, NPV 90.5%, area under the ROC Curve of 0, 92 (p = 0.03; CI 0.86-0.99) and Cronbach's α of 0.76. In the Informant's subscale, the score ≥5 presented a Sensitivity of 77.3%; 96.0% Specificity; PPV 89.4%, NPV 90.5%; RV + 19.3, area on the curve of 0.88, with p = 0.05 (CI 0.80-0.98) and Cronbach's α of 0.81. The calculation of the diagnostic properties of the combined brief CSID showed that the score ≥7 has a sensitivity of 94.1% and specificity of 80.0%. The positive and negative predictive values were 66.7% and 97.0%, respectively. The positive likelihood ratio was 4.7, area under the ROC Curve 0.93 (p = 0.03; CI 0.87-1.00) and Cronbach's α 0.86. It was observed that the screening instrument in the combined version has excellent psychometric properties, and that the CHW Training contributed to the identification of new cases of dementia in the community (17.8% of the total). Conclusions: The brief CSID and the Training proved to be valid, reproducible and sensitive strategies for use in the identification of dementia in the community, addressing an important public health issue in the context of Primary Care.
Descrição
Citação
MACHADO, Danusa de Almeida. Identificação da demência na atenção primária – adaptação e validação de um rastreador (CSID breve) e de um programa para treinamento de Agentes Comunitários de Saúde. 2021. 172 p. Tese (Doutorado em Psicobiologia) – Escola Paulista de Medicina, Universidade Federal de São Paulo, São Paulo, 2021.