Predicting cardiovascular outcomes in elderly patients with acute coronary syndrome: a nomogram approach

TL;DR

This study creates a tool to predict heart risks in elderly patients with a specific heart condition, using clinical data to identify high-risk individuals.

Contribution

The novel contribution is a validated nomogram for predicting MACE in elderly STEMI patients using clinical and procedural factors.

Findings

A nomogram with 8 selected factors predicted MACE with 71% AUC accuracy.

Calibration plots confirmed the model's alignment with observed outcomes.

The model showed discriminative power for MACE prediction via decision curve analysis.

Abstract

Although ST Elevation Myocardial Infarction (STEMI) diagnosis and therapy have improved, high-risk categories like elderly persons still have a significant chance of MACE despite treatment.

This study attempts to construct a predictive nomogram for MACE incidence using clinical data from a STEMI registry.

Tehran Heart Center’s computerized record recognized all 65-year-old STEMI primary PCI patients consecutively. This retrospective study examined demographic, laboratory, clinical, and intra-procedural factors. Post-PCI univariate and multivariate analyses identified MACE risk variables. Decision curve analysis, ROC, and calibration plots validated predictive nomograms. R Studio and R used “tidyverse” and “rms” packages for all analyses.

The 1946 study included 70% training and 30% testing patients. Basic demographic and clinical variables were identical for both groups. The average follow-up was 17 months. 8 factors were selected for the nomogram after univariate and multivariate analysis: left-ventricular ejection fraction (LVEF), serum creatinine, hemoglobin, and fasting blood glucose levels, presence of valvular heart disease, post-PCI TIMI flow grade, diameter of the culprit lesion stent, and presence or absence of shock after PCI. The post-PCI MACE prediction AUC was 71%. Calibration plots showed that the nomogram model was well-calibrated and close to observed outcomes. Decision curve analysis also revealed that the model predicted MACE discriminatively.

A nomogram successfully predicts MACE risk in older STEMI patients using laboratory, clinical, and procedural parameters. This algorithm may identify vulnerable high-risk patients for more aggressive preventative interventions.

not applicable.

Figures