| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 867.15 KB | Adobe PDF |
Advisor(s)
Abstract(s)
Introdução: A Inteligência Artificial (IA) assume um papel cada vez mais relevante nas
nossas vidas, tendo potencial para, no futuro, transformar o atual paradigma da
prestação de cuidados de saúde. Uma área na qual esta tecnologia tem demonstrado
avanços promissores é na referenciação de pacientes aos Cuidados Paliativos (CP). Os CP
têm como objetivo fornecer apoio a pacientes com doenças graves, procurando aliviar
sintomas, melhorar a qualidade de vida e atender às necessidades emocionais e
psicológicas dos pacientes e das suas famílias. Ao aproveitar o poder da previsão e
capacidade dos algoritmos de se treinarem a si mesmos, a IA tem o potencial de
revolucionar a identificação e encaminhamento atempados de pacientes que
beneficiariam de CP, por forma a discutir atempadamente com estes e com as suas
famílias, as expetativas e perspetivas de fim de vida.
Objetivo: Analisar os modelos de previsão já construídos para referenciação aos CP,
assim como perceber se a utilização destes como auxiliares de decisão pelos profissionais
de saúde realmente aumenta a qualidade de vida dos pacientes.
Métodos: Esta dissertação é uma revisão sistemática com síntese narrativa. Foram
incluídos artigos originais Peer-Reviewed, os quais foram submetidos a critérios de
avaliação de qualidade. Foi excluída a literatura cinzenta. A população de interesse
focou-se em dados de pacientes com uma patologia com prognóstico limitado, dados de
mortalidade dos pacientes de uma instituição de saúde ou dados dos pacientes de uma
instituição de saúde.
Resultados: Foram incluídos 22 artigos dos 217 obtidos aquando da pesquisa nas
diferentes bases de dados. A maioria dos estudos correspondiam a estudos quantitativos,
nos quais o tamanho das amostras mostrou-se bastante variado, assim como a população
escolhida, adaptando-se esta ao objetivo do estudo. Também estão presentes tanto
estudos retrospetivos como prospetivos. Os algoritmos de IA obtiveram melhores
resultados relativamente aos métodos tradicionais de referenciação, como o recurso a
métodos tradicionais de estatística ou ferramentas de prognóstico com inserção manual
de dados. Também foi verificado que esta abordagem aumenta a qualidade de vida dos
pacientes, verificando-se a antecipação e o aumento no número de consultas de CP,
diminuição da mortalidade intra-hospitalar, das readmissões por qualquer causa em 30
dias e das admissões e tempo de internamento. Aumentou também a referenciação para prestação de CP ao domicílio, estando estes também associados a uma melhor qualidade
de vida para os pacientes.
Discussão/Conclusões: As novas tecnologias são muitas vezes vistas como potenciais
substitutos para os trabalhos e funções que os seres humanos sempre desempenharam,
no entanto, esse não é de todo o objetivo da implementação da IA na referenciação aos
CP, que deve apenas ser usada com uma ferramenta de apoio à decisão, sendo o veredicto
final sempre dado pelos profissionais de saúde. O uso destes algoritmos tem o potencial
de aumentar significativamente a qualidade de vida dos pacientes, ao permitir uma
sinalização e acesso atempados aos CP, focados não na cura, mas na gestão
individualizada de sintomas e emoções, tendo sido obtidos bons resultados a nível de
outcomes de saúde dos pacientes sinalizados por estes algoritmos.
Introduction: Artificial Intelligence (AI) has been playing an increasingly relevant role in our lives, having the potential to transform the current paradigm of healthcare provision in the future. One area in which this technology has demonstrated promising advances is in referring patients to Palliative Care (PC). PC aims to provide support to patients with serious illnesses, seeking to alleviate symptoms, improve quality of life and meet the emotional and psychological needs of patients and their families. By harnessing the power of prediction and the ability of algorithms to train themselves, AI has the potential to revolutionize the timely identification and referral of patients who would benefit from PC, enabling timely discussions with them and their families about end-oflife expectations and perspectives. Aim: Analyze the already existing prediction models for PC referral, as well as understand whether the use of these as decision aids by health professionals, really increases the quality of life of patients. Methods: This is a systematic review with narrative synthesis. Original Peer-Reviewed articles were included, which were subjected to quality assessment criteria. Grey literature was excluded. The population of interest focused on data from patients with a pathology with limited prognosis, mortality data from patients in a healthcare institution, or data from patients in a healthcare institution. Results: 22 studies were included from the 217 found when the databases where searched. Most of the studies were quantitative, with variable sample sizes and populations chosen in accordance with the studies' objectives. Retrospective and prospective studies are present. AI algorithms obtained better results when compared to traditional referral methods, such as traditional statistical methods or prognostic tools with manual data entry. The use of AIto improve PC referral has yielded favorable results in terms of improving patients’ quality of life with the anticipation and increase in the number of PC consultations, reduction of in-hospital mortality, reduction of readmissions for any cause within 30 days and admissions and less days of hospitalization. Referrals for home-based PC services have also increased, which is also associated with a better quality of life for patients. Discussion/Conclusions: New technologies are often seen as potential substitutes for the jobs and functions that humans have always performed. However, this is not the goal of implementing AI in PC referral, since this technology should only be used as a decision support tool, with the final verdict always being given by healthcare professionals. The use of these algorithms has the potential to significantly increase patients' quality of life, by allowing timely signaling and access to PC, focused not on cure, but on the individualized management of symptoms and emotions, with good results being obtained in terms of patient health outcomes signaled by these algorithms.
Introduction: Artificial Intelligence (AI) has been playing an increasingly relevant role in our lives, having the potential to transform the current paradigm of healthcare provision in the future. One area in which this technology has demonstrated promising advances is in referring patients to Palliative Care (PC). PC aims to provide support to patients with serious illnesses, seeking to alleviate symptoms, improve quality of life and meet the emotional and psychological needs of patients and their families. By harnessing the power of prediction and the ability of algorithms to train themselves, AI has the potential to revolutionize the timely identification and referral of patients who would benefit from PC, enabling timely discussions with them and their families about end-oflife expectations and perspectives. Aim: Analyze the already existing prediction models for PC referral, as well as understand whether the use of these as decision aids by health professionals, really increases the quality of life of patients. Methods: This is a systematic review with narrative synthesis. Original Peer-Reviewed articles were included, which were subjected to quality assessment criteria. Grey literature was excluded. The population of interest focused on data from patients with a pathology with limited prognosis, mortality data from patients in a healthcare institution, or data from patients in a healthcare institution. Results: 22 studies were included from the 217 found when the databases where searched. Most of the studies were quantitative, with variable sample sizes and populations chosen in accordance with the studies' objectives. Retrospective and prospective studies are present. AI algorithms obtained better results when compared to traditional referral methods, such as traditional statistical methods or prognostic tools with manual data entry. The use of AIto improve PC referral has yielded favorable results in terms of improving patients’ quality of life with the anticipation and increase in the number of PC consultations, reduction of in-hospital mortality, reduction of readmissions for any cause within 30 days and admissions and less days of hospitalization. Referrals for home-based PC services have also increased, which is also associated with a better quality of life for patients. Discussion/Conclusions: New technologies are often seen as potential substitutes for the jobs and functions that humans have always performed. However, this is not the goal of implementing AI in PC referral, since this technology should only be used as a decision support tool, with the final verdict always being given by healthcare professionals. The use of these algorithms has the potential to significantly increase patients' quality of life, by allowing timely signaling and access to PC, focused not on cure, but on the individualized management of symptoms and emotions, with good results being obtained in terms of patient health outcomes signaled by these algorithms.
Description
Keywords
Cuidados de Fim de Vida Cuidados Paliativos Ferramentas de Suporte de Decisão Inteligência Artificial Machine Learning Referenciação de Pacientes