Early prediction of vasopressor initiation in ICU sepsis patients using an interpretable EHR-based ML model
Lucas Duval, Antoine Villié, Fei Zheng, Gabriel Terraz, Sophie Blein, Esther Duperchy, Martin Everett, Johan Frieling, Jean-François Llitjos, Maxime Bodinier · 2025
Read Early prediction of vasopressor initiation in ICU sepsis patients using an interpretable EHR-based ML model by Lucas Duval, Antoine Villié, Fei Zheng, Gabriel Terraz, Sophie Blein, Esther Duperchy, Martin Everett, Johan Frieling, Jean-François Llitjos, Maxime Bodinier (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-12706921.txt.
Description
Sepsis patients who will need blood pressure drugs often show warning signs only when shock is near. Lucas Duval and co-authors built a model that flags them hours early from hospital records. Data came from a large intensive care database covering twenty oh eight to twenty nineteen. Cases starting the drugs six to forty-eight hours after admission were matched to similar patients who never needed them. Seven algorithms trained on data from hours before the event.
A random forest model led the field on the held-out set. Shapley values showed which inputs drove each call, keeping the model readable. The final set held fifteen hundred thirty-nine cases and fourteen hundred thirty-one controls. Early flags could buy time for fluids and closer watching.
Catalog record
| CREATOR | Lucas Duval, Antoine Villié, Fei Zheng, Gabriel Terraz, Sophie Blein, Esther Duperchy, Martin Everett, Johan Frieling, Jean-François Llitjos, Maxime Bodinier |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 51 KB |
| VERSION | 1.0 |
| LICENSE | CC BY 4.0 |
| RIGHTS | CC BY |
| SOURCE | Europe PMC |