Construction and validation of the neonatal nutritional risk screening tool / Construção e validação da ferramenta de avaliação de risco nutricional neonatal

ABSTRACT Objective: To develop and validate both the content and reliability of the Neonatal Nutritional Risk Screening Tool (FARNNeo). Methods: Methodological study, convergent care. The instrument was built prior to the literature review and was analyzed by eight judges, during three cycles of the Delphi technique. The judges assessed their relevance and clarity with responses on the Likert scale with three levels, in addition to suggestions. The validation of the instrument was calculated using the agreement rate and content validity index (CVI). After content validation, the instrument was applied by four assisting nutritionists to verify reliability, using Cronbach`s alpha coefficient and the agreement between the evaluators by the Kappa coefficient. Results: All items of the instrument`s content reached the minimum agreement rate (90%) and/or CVI (0.9), except for item three, which in the first cycle obtained CVI 0.77 and 40% of agreement and, in the second cycle, CVI 0.75 and 38% agreement. At the end of the third cycle, all items had CVI values above 0.9. In the instrument application, alpha of 0.96 and Kappa of 0.74 were obtained, which reflect adequate values of internal consistency and agreement between the evaluators. Conclusions: FARNNeo proved to be reliable, clear, relevant, and reproducible for tracking early nutritional risk, systematizing the care of Brazilian newborns admitted to an intensive care unit.

RESUMO Objetivo: Desenvolver e validar o conteúdo e a confiabilidade da Ferramenta de Avaliação do Risco Nutricional Neonatal (FARNNeo). Métodos: Estudo metodológico, convergente assistencial. O instrumento foi construído previamente à revisão da literatura e analisado por oito juízes, durante três ciclos da técnica Delphi. Os juízes avaliaram sua pertinência e clareza com respostas na escala Likert com três níveis, além de sugestões. A validação do instrumento foi calculada pela taxa de concordância e índice de validade de conteúdo (IVC). Após a validação do conteúdo, o instrumento foi aplicado por quatro nutricionistas assistenciais para verificar a fidedignidade, utilizando o coeficiente de alfa de Cronbach e a concordância entre os avaliadores pelo coeficiente Kappa. Resultados: Todos os itens do conteúdo do instrumento alcançaram o valor mínimo da taxa de concordância (90%) e/ou IVC (0,9), com exceção do item três, que no primeiro ciclo obteve IVC 0,77 e 40% de concordância e, no segundo ciclo, IVC 0,75 e 38% de concordância. No fim do terceiro ciclo, todos os itens obtiveram valores de IVC acima de 0,9. Na aplicação da ferramenta, obteve-se alfa de 0,96 e Kappa de 0,74, que refletem valores adequados de consistência interna e concordância entre os avaliadores. Conclusões: A FARNNeo mostrou-se confiável, clara, pertinente e reprodutível para rastreamento do risco nutricional precoce, sistematizando o atendimento de recém-nascidos brasileiros internados em centro de terapia intensiva.

CONSTRUCTION AND VALIDATION OF THE NEONATAL NUTRITIONAL RISK SCREENING TOOL.

OBJECTIVE: To develop and validate both the content and reliability of the Neonatal Nutritional Risk Screening Tool (FARNNeo). METHODS: Methodological study, convergent care. The instrument was built prior to the literature review and was analyzed by eight judges, during three cycles of the Delphi technique. The judges assessed their relevance and clarity with responses on the Likert scale with three levels, in addition to suggestions. The validation of the instrument was calculated using the agreement rate and content validity index (CVI). After content validation, the instrument was applied by four assisting nutritionists to verify reliability, using Cronbach`s alpha coefficient and the agreement between the evaluators by the Kappa coefficient. RESULTS: All items of the instrument`s content reached the minimum agreement rate (90%) and/or CVI (0.9), except for item three, which in the first cycle obtained CVI 0.77 and 40% of agreement and, in the second cycle, CVI 0.75 and 38% agreement. At the end of the third cycle, all items had CVI values above 0.9. In the instrument application, alpha of 0.96 and Kappa of 0.74 were obtained, which reflect adequate values of internal consistency and agreement between the evaluators. CONCLUSIONS: FARNNeo proved to be reliable, clear, relevant, and reproducible for tracking early nutritional risk, systematizing the care of Brazilian newborns admitted to an intensive care unit.