Comparison of Different Risk Scores for Prediction of In-Hospital Mortality in STEMI Patients Treated with PPCI

Comparative Analysis of the Effectiveness of Riskometer Scales in Predicting the Risk of in-Hospital Mortality in Patients With ST-Segment Elevation Myocardial Infarction After Percutaneous Coronary Intervention

AIM

Comparative evaluation of the effectiveness of riskometer scales in predicting in-hospital death (IHD) in patients with ST-segment elevation myocardial infarction (STEMI) after percutaneous coronary intervention (PCI) and the development of new models based on machine learning methods.

MATERIAL AND METHODS

A single-center cohort retrospective study was conducted using data from 4,675 electronic medical records of patients with STEMI (3,202 men and 1,473 women) with a median age of 63 years who underwent emergency PCI. Two groups of patients were isolated: group 1 included 318 (6.8%) patients who died in hospital; group 2 consisted of 4,359 (93.2%) patients with a favorable outcome. The GRACE, CADILLAC, TIMI-STe, PAMI, and RECORD scales were used to assess the risk of IHD. Prognostic models of IHD predicted by the sums of these scale scores were developed using single- and multivariate logistic regression, stochastic gradient boosting, and artificial neural networks (ANN). Risk of adverse events was stratified based on the ANN model data by calculating the median values of predicted probabilities of IHD in the compared groups.

RESULTS

Comparative analysis of the prognostic value of individual scales for the STEMI patients showed differences in the quality of the risk stratification for IHD after PCI. The GRACE scale had the highest prognostic accuracy, while the PAMI scale had the lowest accuracy. The CADILLAC and TIMI-STe scales had acceptable and comparable prognostic abilities, while the RECORD scale showed a significant proportion of false-positive results. The integrative ANN model, the predictors of which were the scores of 5 scales, was superior in the prediction accuracy to the algorithms of single- and multivariate logistic regression and stochastic gradient boosting. Based on the ANN model data, the probability of IHD was stratified into low (<0.3%), medium (0.3-9%), high (9-17%), and very high (>17%) risk groups.

CONCLUSION

The GRACE, CADILLAC and TIMI-STe scales have advantages in the stratification accuracy of IHD risk in patients with STEMI after PCI compared to the PAMI and RECORD scales. The integrated ANN model that combines the prognostic resource of the five analyzed scales, had better quality criteria, and the stratification algorithm based on the data of this model was characterized by accurate identification of STEMI patients with high and very high risk of IHD after PCI.

Analyzing predictors of in-hospital mortality in patients with acute ST-segment elevation myocardial infarction using an evolved machine learning approach

Acute ST-segment elevation myocardial infarction (STEMI) is a severe cardiac ailment characterized by the sudden complete blockage of a portion of the coronary artery, leading to the interruption of blood supply to the myocardium. This study examines the medical records of 3205 STEMI patients admitted to the coronary care unit of the First Affiliated Hospital of Wenzhou Medical University from January 2014 to December 2021. In this research, a novel predictive framework for STEMI is proposed, incorporating evolutionary computational methods and machine learning techniques. A variant algorithm, AGCOSCA, is introduced by integrating crossover operation and observation bee strategy into the original Sine Cosine Algorithm (SCA). The effectiveness of AGCOSCA is initially validated using IEEE CEC 2017 benchmark functions, demonstrating its ability to mitigate the deficiency in local mining after SCA random perturbation. Building upon this foundation, the AGCOSCA approach has been paired with Support Vector Machine (SVM) to forge the predictive framework referred to as AGCOSCA-SVM. Specifically, AGCOSCA is employed to refine the selection of predictors from a substantial feature set before SVM is utilized to forecast the occurrence of STEMI. In our analysis, we observed that SVM excels at managing nonlinear data relationships, a strength that becomes particularly prominent in smaller datasets of STEMI patients. To assess the effectiveness of AGCOSCA-SVM, diagnostic experiments were conducted based on the STEMI sample data. Results indicate that AGCOSCA-SVM outperforms traditional machine learning methods, achieving superior Accuracy, Sensitivity, and Specificity values of 97.83 %, 93.75 %, and 96.67 %, respectively. The selected features, such as acute kidney injury (AKI) stage, fibrinogen, mean platelet volume (MPV), free triiodothyronine (FT3), diuretics, and Killip class during hospitalization, are identified as crucial for predicting STEMI. In conclusion, AGCOSCA-SVM emerges as a promising model framework for supporting the diagnostic process of STEMI, showcasing potential applications in clinical settings.