Quantifying immune dysregulation in pneumonia and sepsis with a parsimonious machine-learning model: a multicohort analysis across care settings and reanalysis of a hydrocortisone randomised controlle
3 hours ago
- #immunomodulation
- #sepsis
- #machine-learning
- The study aimed to quantify immune dysregulation in pneumonia and sepsis using a parsimonious machine-learning model.
- A multicohort analysis was conducted across different care settings, including emergency departments, general wards, and intensive care units.
- The research involved reanalyzing a hydrocortisone randomized controlled trial (CAPE COD trial) to assess treatment effects based on immune dysregulation stages.
- A three-biomarker machine-learning framework (procalcitonin, soluble TREM-1, and IL-6) accurately predicted immune dysregulation stages and continuous scores (cDIP).
- Increased immune dysregulation was associated with higher mortality and secondary infections, independent of clinical severity.
- Hydrocortisone showed a survival benefit only in participants classified as severely dysregulated by the model, with faster immune recovery observed in this group.
- The study provides a publicly available three-biomarker framework for precision-guided immunomodulatory therapy in sepsis and pneumonia.