Mortality risk prediction from dense PICU data from patients with suspected infection: data-derived physiological trajectories outperform expert assessments when temporal resolution is high

Tom Velez, Oluwakemi Badaki-Makun, Danielle Hirsch, Danielle Claire Mercurio, Holly Depinet, Maya Dewan, Rishikesan Kamaleswaran, Jocelyn Grunwell, Maria Triantafyllou, Nora Wolff, Fehima Abdelrahman, Charles Macias, Ioannis Koutroulis · 2026

Read Mortality risk prediction from dense PICU data from patients with suspected infection: data-derived physiological trajectories outperform expert assessments when temporal resolution is high by Tom Velez, Oluwakemi Badaki-Makun, Danielle Hirsch, Danielle Claire Mercurio, Holly Depinet, Maya Dewan, Rishikesan Kamaleswaran, Jocelyn Grunwell, Maria Triantafyllou, Nora Wolff, Fehima Abdelrahman, Charles Macias, Ioannis Koutroulis (2026) 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/PMC13523156.txt.

Description

Tom Velez leads a team building models to predict death in very ill children with suspected infection in a 2026 modeling study. Data from more than three thousand intensive care encounters, including one hundred seven deaths, covered the first twenty-four hours. Machine learning models were built in stages, adding vital sign extremes, variability and trends, treatment intensity, and laboratory patterns, and were tested against expert severity scores.

Simple vital sign extremes already discriminated well. Adding variability, slope, and the timing of breathing support and blood pressure drugs improved performance substantially, with laboratory trends adding more. Established severity scores, including two pediatric systems, added nothing beyond the data-derived features. The authors conclude that dense hourly data alone can reveal deterioration patterns strongly tied to early mortality.

Catalog record

CREATORTom Velez, Oluwakemi Badaki-Makun, Danielle Hirsch, Danielle Claire Mercurio, Holly Depinet, Maya Dewan, Rishikesan Kamaleswaran, Jocelyn Grunwell, Maria Triantafyllou, Nora Wolff, Fehima Abdelrahman, Charles Macias, Ioannis Koutroulis
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE35809
LICENSEhttps://creativecommons.org/licenses/by/4.0
RIGHTSCC BY
SOURCEEurope PMC

Source: https://europepmc.org/article/MED/PMC13523156

Browse the full library · About the Tower of Babel