Source record
TREWS sepsis early warning on the inpatient ward
The source is Adams and colleagues, Nature Medicine, 21 July 2022, a prospective evaluation of TREWS across five hospitals. It reports that, among patients with sepsis identified before antibiotics, confirmation of the alert within 3 hours was associated with an adjusted absolute mortality difference of 3.3 percentage points (relative 18.7%) versus alerts not confirmed in that window. A companion paper reported a 1.85-hour shorter median time to the first antibiotic order for the same comparison. These are associations with how the alert was used, not a randomized trial.
- Source date
- Jul 21, 2022 · Year 2022
- Evidence label
- Published
- Source type
- Paper
Citation
Adams et al., Nature Medicine
Adams et al. Prospective, multi-site study of patient outcomes after implementation of the TREWS machine learning-based early warning system for sepsis. Nature Medicine, 21 July 2022. https://www.nature.com/articles/s41591-022-01894-0
Key public facts
- TREWS was studied prospectively in emergency departments and wards at two academic and three community hospitals.
- A clinician confirms or dismisses the alert. The system does not itself diagnose sepsis.
- Among patients with sepsis identified before antibiotics, confirmation within 3 hours was associated with an adjusted absolute mortality difference of 3.3 percentage points (relative 18.7%) versus alerts not confirmed in that window.
- A companion paper reported a 1.85-hour shorter median time to the first antibiotic order for that same comparison.
- The paper reports associations with how the alert was used. It is not a randomized trial of the model.
Directory entry for education and market scanning. Not medical advice and not an endorsement.