Predicting Hospitalisation In Older Patients Attended In The Emergency Department Using Classical And Artificial Intelligence Mathematical Strategies
Background: The ageing population poses a significant challenge for health and social care systems. Emergency Departments (EDs) frequently experience overcrowding due to the high volume of patients and the limited availability of hospital beds. From the perspective of bed management planners, knowin...
| Autores: | , , , , , , , , , , , , , , |
|---|---|
| Tipo de recurso: | artículo |
| Estado: | Versión publicada |
| Fecha de publicación: | 2025 |
| País: | España |
| Institución: | Varias* (Consorci de Biblioteques Universitáries de Catalunya, Centre de Serveis Científics i Acadèmics de Catalunya) |
| Repositorio: | Recercat. Dipósit de la Recerca de Catalunya |
| OAI Identifier: | oai:recercat.cat:2445/222386 |
| Acceso en línea: | https://hdl.handle.net/2445/222386 |
| Access Level: | acceso abierto |
| Palabra clave: | Persones grans Intel·ligència artificial en medicina Older people Medical artificial intelligence |
| Sumario: | Background: The ageing population poses a significant challenge for health and social care systems. Emergency Departments (EDs) frequently experience overcrowding due to the high volume of patients and the limited availability of hospital beds. From the perspective of bed management planners, knowing the likelihood of a patient's admission at the earliest stage of care can be highly beneficial for effective resource planning. The goal of our study was to develop a prediction model to identify patients with a high probability of being admitted to the hospital. Methods: We included all patients aged 65 or older who were treated over the course of one week in 52 Spanish Emergency Departments. The data collected included socio-demographic characteristics, baseline functional status, comorbidities, vital signs, chronic treatments, and laboratory test results. The primary outcome variable was hospital admission. We applied several mathematical strategies to develop the most accurate model for identifying high-risk patients likely to require hospitalisation. Results: The most effective model was developed using a random forest algorithm, incorporating various variables available during patient care in the ED. The probability of admission was categorised into four risk groups: 2.19 %, 15.65 %, 25.09 %, and 57.08 %. The resulting model had a sensitivity of 0.88. Conclusion: We developed a high-sensitivity score for hospital admission in older patients treated in the ED to enhance the management of patient flow by bed planners. This score will help prevent ED overcrowding, which compromises patient safety and disrupts the healthcare system. |
|---|