Inflammation Indices Predict 3-Year Mortality in Diabetic AMI
This study develops and validates a stacked predictive model using five inflammation-related indices to predict three-year all-cause mortality among patients with severe diabetes and acute myocardial infarction, using 833 MIMIC-IV cases (training/test) and 166 external cases from Zhongnan Hospital (follow-up to July 22, 2025). LMR, NLR, and PIV were independently associated with mortality; the model achieved AUCs of 0.803 (internal) and 0.781 (external).
Key Points
- 1Identify LMR, NLR, and PIV as significant prognostic inflammatory indices for three-year mortality.
- 2Demonstrate stacked model achieves robust discrimination (AUC 0.803 internal, 0.781 external), indicating predictive validity.
- 3Enable clinicians to use routine CBC-derived ratios for cost-effective risk stratification in resource-limited settings.
Scoring Rationale
Strong external validation and practical biomarkers drive score, limited by single-center derivation and need for multicenter generalization.
Sources
Public references used for this report.
Practice interview problems based on real data
1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.
Try 250 free problems
