Integrating discrete-event simulation and artificial intelligence for shortening bed waiting times in hospitalization departments during respiratory disease seasons

Seasonal Respiratory Diseases (SRDs) usually produce a heightened number of Emergency Department (ED) attendances due to their rapid dissemination within the community and the ineffective prevention measures. Such a context requires effective management of the emergency care processes to provide in-...

Full description

Autores:
Ortiz-Barrios, Miguel
Ishizaka c , Alessio
Barbati, Maria
Arias-Fonseca, Sebastián
Khan, Jehangir
Gul, Muhammet
Yücesan, Melih
Alfaro-Saíz, Juan-Jose
Pérez-Aguilar, Armando
Tipo de recurso:
Article of investigation
Fecha de publicación:
2024
Institución:
Corporación Universidad de la Costa
Repositorio:
REDICUC - Repositorio CUC
Idioma:
eng
OAI Identifier:
oai:repositorio.cuc.edu.co:11323/13994
Acceso en línea:
https://hdl.handle.net/11323/13994
https://repositorio.cuc.edu.co/
Palabra clave:
Discrete-Event simulation (DES)
Artificial intelligence (AI)
Random forest (RF)
Hospitalization departments (HDs)
Seasonal respiratory diseases (SRDs)
Bed waiting time
Rights
openAccess
License
Atribución 4.0 Internacional (CC BY 4.0)
Description
Summary:Seasonal Respiratory Diseases (SRDs) usually produce a heightened number of Emergency Department (ED) attendances due to their rapid dissemination within the community and the ineffective prevention measures. Such a context requires effective management of the emergency care processes to provide in-time diagnosis and treatment to infected patients. Nonetheless, EDs have evidenced severe operational deficiencies during these periods, thereby provoking extended bed waiting times in Hospitalization Departments (HDs). Therefore, this paper presents a hybrid approach merging Artificial Intelligence (AI) and Discrete-Event Simulation (DES) to shorten the bed waiting times in HDs considering patient records collated in the first emergency care stages. First, we implemented Random Forest (RF) to estimate the probability of respiratory worsening based on sociodemographic and clinical patient data. Second, we inserted these probabilities into a DES model mimicking the emergency care from the admission to the HD. We then pretested different HD configurations and strategies seeking to reduce the HD bed waiting time. A case study of a European hospital group was used to validate the suggested framework. The AI-DES model enabled decision-makers to identify an improvement proposal with hospitalization bed waiting time lessening, oscillating between 7.93 and 7.98 h.