Loading...
Publication
Electronic Health Record (EHR) : modelo de recomendação nutricional para o Serviço Nacional de Saúde Português
| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 8.36 MB | Adobe PDF |
Authors
Advisor(s)
Abstract(s)
A normalização dos registos eletrónicos de dados clínicos é um fator de melhoria da qualidade dos cuidados de saúde e da eficiência das organizações.
Também a melhoria do conhecimento organizacional e o desenvolvimento de sistemas de apoio à decisão são potenciados com a normalização dos dados clínicos.
Um outro componente dos dados clínicos são as ontologias que lhes dão significado semântico e validam a sua coerência.
Na área nutricional desenvolveu-se um modelo de recomendação nutricional e propõe-se um sistema de gestão de dietas.
O modelo proposto baseia-se numa rede neuronal que adquire conhecimento com base nos casos de prescrições nutricionais. A validação do modelo de recomendação demonstrou um baixo valor de erro relativo médio.
Também se investiga, neste estudo, o método de criação de archetypes de dados nutricionais. O desenvolvimento destes archetypes permitirá incluir dados nutricionais nos registos eletrónicos de dados clínicos. Segundo os investigadores, médicos e dietistas que participaram no estudo, a prescrição de valores nutricionais adequados aos perfis e patologias dos doentes tem uma contribuição muito significativa para a adesão à terapia e diminuição das morbidades.
A aplicação do modelo proposto poderá ainda apoiar a decisão de prescrição nutricional e contribuir para um melhor conhecimento dos efeitos nutricionais nos doentes.
The standardization of electronic records for clinical data is a factor in improving the quality of healthcare and efficiency of organizations. The improvement of organizational data and the development of decision support systems are also enhanced by the normalization of clinical data. Another component of the clinical data are the ontologies that provide semantic meaning to them and validate their coherence. In the nutritional field, it was developed a model of nutritional recommendations and also it is suggested a diet management system. The proposed model is based on a neural network that acquires knowledge which is based on the cases of nutritional prescriptions. The recommendation model validation showed a low value of mean relative error. In this study it is also investigated the method of creating nutritional data archetypes. The development of these archetypes will make it possible to include nutritional data in electronic records of clinical data. According to the researchers, doctors and dietitians who participated in the study, the prescription of nutritional values adequate to the patients’ profiles and pathologies has a very meaningful contribution to be well received by patients to perform the therapy and to decrease morbidity. The application of the proposed could also support the nutritional prescription decision and contribute to a better understanding of the nutritional effects in patients.
The standardization of electronic records for clinical data is a factor in improving the quality of healthcare and efficiency of organizations. The improvement of organizational data and the development of decision support systems are also enhanced by the normalization of clinical data. Another component of the clinical data are the ontologies that provide semantic meaning to them and validate their coherence. In the nutritional field, it was developed a model of nutritional recommendations and also it is suggested a diet management system. The proposed model is based on a neural network that acquires knowledge which is based on the cases of nutritional prescriptions. The recommendation model validation showed a low value of mean relative error. In this study it is also investigated the method of creating nutritional data archetypes. The development of these archetypes will make it possible to include nutritional data in electronic records of clinical data. According to the researchers, doctors and dietitians who participated in the study, the prescription of nutritional values adequate to the patients’ profiles and pathologies has a very meaningful contribution to be well received by patients to perform the therapy and to decrease morbidity. The application of the proposed could also support the nutritional prescription decision and contribute to a better understanding of the nutritional effects in patients.
Description
Tese de Doutoramento em Gestão da Informação na especialidade de Informação apresentada à Universidade Aberta
Keywords
Serviço Nacional de Saúde Sistemas de informação Bases de dados Registos eletrónicos Normalização Registos médicos Nutrição Ontologia Dieta EHR Clinical data electronic record Ontologies Neural network Nutrition Diets Pathologies
Citation
Macedo, Mário José Costa de - Electronic Health Record (EHR) [Em linha] : modelo de recomendação nutricional para o Serviço Nacional de Saúde Português. [Lisboa] : [s.n.], 2011. 327 p.