Desenvolvimento de um programa de apoio à decisão para análise do crescimento, nutrição e maturação sexual para uso na atenção primária

• Ano de defesa: 2010
• Autor(a) principal: Succi, Ernesto [UNIFESP]
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• dARK ID: ark:/48912/0013000024h2q
• Idioma: por
• Instituição de defesa: Universidade Federal de São Paulo (UNIFESP)
• Palavras-chave em Português: Diagnóstico por computador; Crescimento e desenvolvimento; Tomada de decisões assistida por computador; Maturidade sexual; Bases de conhecimento; Diagnosis, Computer-Assisted; Growth and Development; Decision Making, Computer-Assisted; Sexual Maturation; Knowledge Bases
• Link de acesso: http://repositorio.unifesp.br/handle/11600/9309

The main drive to develop a Health Decision Support System (HDSS) lies on creating a digital system to help medical practitioners exposed to an ever increasing quantity of data, information and knowledge which they have to resort to in order to conduct their duties properly. In a practice field such as adolescent healthcare, in which physicians diagnose deviant states, not always based in sufficient scientific evidence, the HDSS becomes a musthave tool for those who are not specialized in this field. Objective: Survey information related to growth, nutrition, and sexual maturity, compile a knowledge base, develop the HDSS tool to provide non-specialist medical practitioners with information on puberty and sexual maturity, validate the compiled knowledge base and validate its user interface. Design: The collection of information has been conducted by querying related bibliography at Medline, PubMed, Google Scholar databases and UNIFESP’s library. The knowledge base surveyed has been translated into computer language by application of production rules. To validate the knowledge base, all medical diagnoses generated by the HDSS for 30 test cases were compared with diagnoses reported by 9 specialist medical practitioners. The system interface was validated by 55 physicians who answered standardized forms on system satisfaction. Results: We developed an open-source, Java™-based, decision-making system related to puberty and sexual maturity capable of running on Mac™, GNU Linux and Windows™ operating systems. The HDSS comprises a Java™-based inference engine with 128 production rules and graphical user interface. Based on the construction of the knowledge base we generated a 591-entry glossary to help users. Information was evaluated by comparing the results generated by the system and those reported by specialist medical practitioners. The evaluation rendered by specialists and the system was statistically significant by 1% with 77% of the specialists and their respective groups. The user interface was positively evaluated regarding all aspects. Conclusions: We have collected information on puberty and sexual maturity and used that to develop an HDSS comprised of 128 production rules. The system was evaluated and displayed adequate performance when compared with diagnoses reported by specialist medical practitioners, and its user interface was validated by physicians. Due to the fact this is an open-source, Java™-based system, we believe it is suitable for deployment in the public healthcare system.