Quantification of a qualitative sepsis code: laying the foundations for the automation revolution

To quantify a qualitative screening tool for the early recognition of sepsis in children with fever either visiting the emergency department or already admitted to hospital. Prospective observational study including febrile patients under 18 years of age. Sepsis diagnosis was the main outcome. A mul...

Descripción completa

Detalles Bibliográficos
Autores: Solé-Ribalta A, Balaguer M, Bobillo-Pérez S, Girona-Alarcón M, Guitart C, Esteban E, Jordan-Garcia I
Tipo de recurso: artículo
Estado:Versión publicada
Fecha de publicación:2023
País:España
Institución:Fundació Sant Joan de Déu
Repositorio:r-FSJD. Repositorio Institucional de Producción Científica de la Fundació Sant Joan de Déu
OAI Identifier:oai:fsjd.fundanetsuite.com:p22937
Acceso en línea:https://fsjd.fundanetsuite.com/Publicaciones/ProdCientif/PublicacionFrw.aspx?id=22937
Access Level:acceso abierto
Palabra clave:Sepsis
Paediatric
Screening
Automatic
Descripción
Sumario:To quantify a qualitative screening tool for the early recognition of sepsis in children with fever either visiting the emergency department or already admitted to hospital. Prospective observational study including febrile patients under 18 years of age. Sepsis diagnosis was the main outcome. A multivariable analysis was performed with 4 clinical variables (heart rate, respiratory rate, disability, and poor skin perfusion). The cut-off points, odds ratio, and coefficients of these variables were identified. The quantified tool was then obtained from the coefficients. The area under the curve (AUC) was obtained and internal validation was performed using k-fold cross-validation. Two hundred sixty-six patients were included. The multivariable regression confirmed the independent association of the 4 variables with the outcome. The quantified screening tool yielded an excellent AUC, 0.825 (95%CI 0.772-0.878, p < 0.001), for sepsis prediction.Conclusion: We successfully quantified a sepsis screening tool, and the resulting model has an excellent discriminatory power.