Stress hyperglycemia ratio and machine learning model for prediction of all-cause mortality in patients undergoing cardiac surgery
Authors' conclusion
In patients undergoing cardiac surgery, higher postoperative SHR index levels were significantly associated with increased risk of in-hospital, 90-day and 360-day all-cause mortality. The SHR index may serve as a valid tool for assessing the severity after cardiac surgery and guiding treatment decisions.
Study context
- Population
- Patients undergoing cardiac surgery, Medical Information Mart for Intensive Care IV (MIMIC-IV) database; mean age 68 ± 12 years, 30.6% female
- Participants
- 3,848 patients
- Exposure
- Postoperative stress hyperglycemia ratio (SHR) index quartiles
- Outcome
- 30-day all-cause mortality
- Follow-up
- 30 days
Methodological detailsMIMIC-IV database, machine learning prediction models, LASSO feature selection
- MIMIC-IV database
- Machine learning prediction models
- LASSO feature selection
Reported results (other outcomes)
As reported in the abstract
The primary endpoint is not reported in this abstract; these are its other reported outcomes.
In-hospital, 90-day and 360-day all-cause mortality by SHR level (Kaplan-Meier) · in-hospital, 90 days, 360 days
- Higher postoperative SHR associated with increased risk of mortality (log-rank)
- P < 0.05
- Highest SHR quartile associated with significantly higher mortality risk (Cox regression)
- P < 0.05
- Nonlinear relationship between SHR and all-cause mortality (RCS)
- P for nonlinear < 0.05
Machine learning prediction of mortality
- Naive Bayes model highest AUC for in-hospital mortality
- 0.7936
- Naive Bayes model highest AUC for 360-day all-cause mortality
- 0.7410
Observational data: associations, not effects.
Original publication
Stress hyperglycemia ratio and machine learning model for prediction of all-cause mortality in patients undergoing cardiac surgery. Cardiovasc Diabetol. Published 15 February 2025. doi:10.
Composed by AI from the published abstract and checked against it. How Briefs are made →