Desenvolvimento e validação de questionário de triagem nutricional para pacientes oncológicos adultos e idosos em atendimento ambulatorial
Data
2023-09-04
Autores
Magalhães, Lidiane Pereira 
Orientadores
Silva, Celso Arrais Rodrigues da
Tipo
Tese de doutorado
Título da Revista
ISSN da Revista
Título de Volume
Resumo
Introdução: A desnutrição é fator presente frequentemente em pacientes oncológicos, que é multifatorial, mas está relacionada ao aumento da toxicidade e redução das taxas de resposta ao tratamento. Por esta razão, detectá-la precocemente, proporciona oportunidade de intervenção nutricional imediata, impedindo ou mitigando o risco ou a desnutrição em si. Atualmente, não há um instrumento de triagem validado para este grupo de pacientes. Objetivos: Elaborar e validar um instrumento de triagem nutricional para pacientes oncológicos adultos e idosos em atendimento ambulatorial. Métodos: Desenvolvimento de um questionário para ser aplicado em indivíduos adultos e idosos, com diagnóstico oncológico ou onco-hematológico confirmado, em atendimento ambulatorial, estando em tratamento ou não, contendo as principais situações que podem induzir ao risco nutricional. Para a definição do primeiro gabarito, dados foram simulados, e analisados determinando o ponto de corte inicial. O instrumento foi desenvolvido e submetido a avaliação semântica e técnica para ser encaminhado para aplicação nas instituições convidadas. A análise estatística contou com estimativa do tamanho da amostra, estatística descritiva, aplicação dos testes do Qui-quadrado, de Kolmogorov–Smirnov, de Diagnóstico, curva ROC e concordância Kappa. Resultados: O questionário foi denominado TriNOAI – Triagem Nutricional em Oncologia para Adultos e Idosos. Foram 632 indivíduos entrevistados, sendo 66,3% do sexo feminino e 51,4% adultos, de instituições públicas e privadas em 4 regiões do Brasil. A independência dos grupos etários foi verificada através do teste Qui-quadrado, atribuindo-se peso às pessoas idosas nas questões de maior relevância. Foi aplicado em comparação aos questionários ASG-PPP short form (padrão ouro), NRS-2002 e NUTRISCORE, e resultado, demonstrado através do gráfico Box-plot e Curva Roc. Em análise subjacente dos critérios de acompanhante e fonte de renda, verificou-se que, daqueles que não tem cuidador bem como, fonte de renda própria, 81,2% são mulheres, 56,2% são adultos e 92,2% provenientes do SUS. Nesta mesma condição, os que apresentam redução de ingestão alimentar, perda de peso e alteração de função física, também são do sexo feminino, representando, 77,8%, 100% e 84,8% respectivamente. Discussão: Os instrumentos de triagem, recomendados atualmente, não foram desenvolvidos para pacientes oncológicos, sendo, portanto, generalistas, provenientes de casuística estrangeira, e elaborados para ambiente intra-hospitalar. A análise estatística destes instrumentos, em comparação ao proposto, demonstrou incerteza na precisão diagnóstica dos mesmos, enquanto o desempenho do instrumento de triagem deste trabalho foi notadamente superior por uma AUC de aproximadamente 0,800 para os dois grupos etários, quando comparados ao ASG-PPP short form, demonstrada também através de outras medidas trazendo maior assertividade ao diagnóstico de risco nutricional. Presença de cuidador e fonte de renda, não são questionamentos realizados em outros materiais, e demonstraram ser fator importante, associado aos outros, no perfil nutricional destes indivíduos. Conclusão: Após avaliação semântica, técnica, aplicação em pacientes, comparando-o aos questionários recomendados, realizado ajuste da pontuação final e determinação do ponto de corte, o instrumento demonstrou-se seguro para o diagnóstico de risco de desnutrição, apresentando desempenho superior aos outros, apresentando como diferenciais a variabilidade da amostra, n estatisticamente significante, inclusão de informação sobre presença de acompanhante e renda, estando, dessa forma, devidamente validado para aplicação na pratica clínica.
Introduction: Malnutrition is a frequently present factor in oncology patients and is multifactorial, being associated with increased treatment toxicity and reduced response rates. Early detection of malnutrition provides an opportunity for immediate nutritional intervention, preventing or mitigating the risk of malnutrition itself. Currently, there is no screening tool developed specifically for this group of patients. Objectives: To develop and validate a nutritional screening tool for adult and elderly oncology patients in outpatient care. Methods: A questionnaire was developed to be administered to adult and elderly individuals with confirmed oncological or onco-hematological diagnoses, receiving outpatient care, whether in active treatment or not, containing key situations that could induce nutritional risk. For the determination of the initial cutoff point, simulated data were analyzed. The instrument underwent semantic and technical evaluation before being applied in invited institutions. The statistical analysis included sample size estimation, descriptive statistics, application of the Chi-square test, Kolmogorov–Smirnov test, Diagnostic test, ROC curve, and Kappa concordance. Results: A total of 632 individuals were interviewed, with 66.3% being female and 51.4% adults, from public and private institutions in 4 regions of Brazil. Age group independence was verified using the Chi-square test, with weight assigned to elderly individuals in questions of greater relevance. The instrument was compared to ASG-PPP short form (gold standard), NRS-2002, and NUTRISCORE questionnaires, and the results were demonstrated through Box-plot and ROC curve. Underlying analysis of caregiver and income source criteria showed that among those without a caregiver and with their own source of income, 81.2% were women, 56.2% were adults, and 92.2% were from the public health system (SUS). In the same condition, those who presented reduced food intake, weight loss, and altered physical function were also predominantly female, representing 77.8%, 100%, and 84.8%, respectively. Discussion: Currently recommended screening tools were not developed specifically for oncology patients; therefore, they are generalist and based on foreign casuistry, designed for in-hospital settings. The statistical analysis of these tools, when compared to the proposed instrument, showed uncertainty in their diagnostic accuracy, whereas the performance of the screening instrument in this study was notably superior with an AUC of approximately 0.800 for both age groups, compared to the ASG-PPP short form. This was also demonstrated through other measures, providing greater accuracy in the diagnosis of nutritional risk. The presence of a caregiver and income source are not questioned in other materials but were shown to be important factors, when associated with others, in the nutritional profile of these individuals. Conclusion: After semantic and technical evaluation, application to patients, comparison with recommended questionnaires, adjustment of the final scoring, and determination of the cutoff point, the instrument proved to be reliable for diagnosing the risk of malnutrition, outperforming others. Its strengths include sample variability, statistically significant sample size (n), inclusion of information about the presence of a caregiver and income source, thus being properly validated for application in clinical practice.
Introduction: Malnutrition is a frequently present factor in oncology patients and is multifactorial, being associated with increased treatment toxicity and reduced response rates. Early detection of malnutrition provides an opportunity for immediate nutritional intervention, preventing or mitigating the risk of malnutrition itself. Currently, there is no screening tool developed specifically for this group of patients. Objectives: To develop and validate a nutritional screening tool for adult and elderly oncology patients in outpatient care. Methods: A questionnaire was developed to be administered to adult and elderly individuals with confirmed oncological or onco-hematological diagnoses, receiving outpatient care, whether in active treatment or not, containing key situations that could induce nutritional risk. For the determination of the initial cutoff point, simulated data were analyzed. The instrument underwent semantic and technical evaluation before being applied in invited institutions. The statistical analysis included sample size estimation, descriptive statistics, application of the Chi-square test, Kolmogorov–Smirnov test, Diagnostic test, ROC curve, and Kappa concordance. Results: A total of 632 individuals were interviewed, with 66.3% being female and 51.4% adults, from public and private institutions in 4 regions of Brazil. Age group independence was verified using the Chi-square test, with weight assigned to elderly individuals in questions of greater relevance. The instrument was compared to ASG-PPP short form (gold standard), NRS-2002, and NUTRISCORE questionnaires, and the results were demonstrated through Box-plot and ROC curve. Underlying analysis of caregiver and income source criteria showed that among those without a caregiver and with their own source of income, 81.2% were women, 56.2% were adults, and 92.2% were from the public health system (SUS). In the same condition, those who presented reduced food intake, weight loss, and altered physical function were also predominantly female, representing 77.8%, 100%, and 84.8%, respectively. Discussion: Currently recommended screening tools were not developed specifically for oncology patients; therefore, they are generalist and based on foreign casuistry, designed for in-hospital settings. The statistical analysis of these tools, when compared to the proposed instrument, showed uncertainty in their diagnostic accuracy, whereas the performance of the screening instrument in this study was notably superior with an AUC of approximately 0.800 for both age groups, compared to the ASG-PPP short form. This was also demonstrated through other measures, providing greater accuracy in the diagnosis of nutritional risk. The presence of a caregiver and income source are not questioned in other materials but were shown to be important factors, when associated with others, in the nutritional profile of these individuals. Conclusion: After semantic and technical evaluation, application to patients, comparison with recommended questionnaires, adjustment of the final scoring, and determination of the cutoff point, the instrument proved to be reliable for diagnosing the risk of malnutrition, outperforming others. Its strengths include sample variability, statistically significant sample size (n), inclusion of information about the presence of a caregiver and income source, thus being properly validated for application in clinical practice.
Descrição
Citação
MAGALHÃES, Lidiane Pereira. Desenvolvimento e validação de questionário de triagem nutricional para pacientes oncológicos adultos e idosos em atendimento ambulatorial. 2023. 219 f. Tese (Doutorado em Hematologia) - Escola Paulista de Medicina, Universidade Federal de São Paulo,(UNIFESP). São Paulo, 2023.