Comparison of bio-inspired algorithms applied to the hospital mortality risk stratification

The construction of patient classification (or risk adjustment) systems allows comparison of the effectiveness and quality of hospitals and hospital services, providing useful information for management decision making and management of hospitals. Risk adjustment systems to stratify patients’ severi...

ver descrição completa

Detalhes bibliográficos
Autores: Silva, Jesús, Herazo-Beltrán, Yaneth, Marín-González, Freddy, Varela Izquierdo, Noel, Pineda, Omar, Palencia-Domínguez, Pablo, Vargas Mercado, Carlos
Formato: artículo
Estado:Versión aceptada para publicación
Fecha de publicación:2020
País:Colombia
Recursos:Corporación Universidad de la Costa
Repositorio:Repositorio REDICUC
Idioma:inglés
OAI Identifier:oai:repositorio.cuc.edu.co:11323/7741
Acesso em linha:https://hdl.handle.net/11323/7741
https://doi.org/10.1007/978-981-15-4875-8_16
https://repositorio.cuc.edu.co/
Access Level:acceso abierto
Palavra-chave:Hospital mortality
Risk stratification
Intensive care unit
Artificial neural networks
Bootstrap
Descrição
Resumo:The construction of patient classification (or risk adjustment) systems allows comparison of the effectiveness and quality of hospitals and hospital services, providing useful information for management decision making and management of hospitals. Risk adjustment systems to stratify patients’ severity in a clinical outcome are generally constructed from care variables and using statistical techniques based on logistic regression (RL). The objective of this investigation is to compare the hospital mortality prediction capacity of an artificial neural network (RNA) with other methods already known.