Construction and evaluation of nomogram model for individualized prediction of risk of major adverse cardiovascular events during hospitalization after percutaneous coronary intervention in patients with acute ST-segment elevation myocardial infarction

Abstract

BackgroundEmergency percutaneous coronary intervention (PCI) in patients with acute ST-segment elevation myocardial infarction (STEMI) helps to reduce the occurrence of major adverse cardiovascular events (MACEs) such as death, cardiogenic shock, and malignant arrhythmia, but in-hospital MACEs may still occur after emergency PCI, and their mortality is significantly increased once they occur. The aim of this study was to investigate the risk factors associated with MACE during hospitalization after PCI in STEMI patients, construct a nomogram prediction model and evaluate its effectiveness.MethodsA retrospective analysis of 466 STEMI patients admitted to our hospital from January 2018 to June 2022. According to the occurrence of MACE during hospitalization, they were divided into MACE group (n = 127) and non-MACE group (n = 339), and the clinical data of the two groups were compared; least absolute shrinkage and selection operator (LASSO) regression was used to screen out the predictors with non-zero coefficients, and multivariate Logistic regression was used to analyze STEMI Independent risk factors for in-hospital MACE in patients after emergency PCI; a nomogram model for predicting the risk of in-hospital MACE in STEMI patients after PCI was constructed based on predictive factors, and the C-index was used to evaluate the predictive performance of the prediction model; the Bootstrap method was used to repeat sampling 1,000 Internal validation was carried out for the second time, the Hosmer-Lemeshow test was used to evaluate the model fit, and the calibration curve was drawn to evaluate the calibration degree of the model. Receiver operating characteristic (ROC) curves were drawn to evaluate the efficacy of the nomogram model and thrombolysis in myocardial infarction (TIMI) score in predicting in-hospital MACE in STEMI patients after acute PCI.ResultsThe results of LASSO regression showed that systolic blood pressure, diastolic blood pressure, Killip grade II-IV, urea nitrogen and left ventricular ejection fraction (LVEF), IABP, NT-ProBNP were important predictors with non-zero coefficients, and multivariate logistic regression analysis was performed to analyze that Killip grade II-IV, urea nitrogen, LVEF, and NT-ProBNP were independent factors for in-hospital MACE after PCI in STEMI patients; a nomogram model for predicting the risk of in-hospital MACE after PCI in STEMI patients was constructed with the above independent predictors, with a C-index of 0.826 (95% CI: 0.785–0.868) having a good predictive power; the results of H-L goodness of fit test showed χ2 = 1.3328, P = 0.25, the model calibration curve was close to the ideal model, and the internal validation C-index was 0.818; clinical decision analysis also showed that the nomogram model had a good clinical efficacy, especially when the threshold probability was 0.1–0.99, the nomogram model could bring clinical net benefits to patients. The nomogram model predicted a greater AUC (0.826) than the TIMI score (0.696) for in-hospital MACE after PCI in STEMI patients.ConclusionUrea nitrogen, Killip class II-IV, LVEF, and NT-ProBNP are independent factors for in-hospital MACE after PCI in STEMI patients, and nomogram models constructed based on the above factors have high predictive efficacy and feasibility.

Prediction models for major adverse cardiovascular events after percutaneous coronary intervention: a systematic review

“…The models sharing identical predictor variables and stemming from the same study population, yet predicting MACE at varying time intervals, were categorized as one model. A total of 12 studies addressed prognostic models for patients with ST-segment elevation myocardial infarction (STEMI) ( 20 , 26 – 29 , 32 – 34 , 36 – 39 ). Six studies were centered on patients with myocardial infarction (MI) ( 11 , 12 , 21 , 22 , 24 , 35 ).…”

“…All models employed regression analysis, with 15 studies ( 12 , 20 – 25 , 27 , 28 , 30 , 31 , 34 – 36 , 38 ) using logistic regression, while eight studies ( 11 , 13 , 26 , 29 , 32 , 33 , 37 , 39 ) using Cox regression. The models were ultimately presented in various forms, such as formulas ( 22 , 30 , 31 ), risk scores ( 35 , 37 ), nomograms ( 12 , 13 , 20 , 21 , 25 – 29 , 36 , 38 , 39 ), or combinations thereof. Detailed information regarding the modeling methods, variable selecting methods, calibration method, and model presentation can be found in Table 2 .…”

“…Detailed information regarding the modeling methods, variable selecting methods, calibration method, and model presentation can be found in Table 2 . The predicted outcomes spanned from in-hospital MACE ( 20 , 21 , 25 , 27 , 31 , 35 , 36 ) to MACE occurring post-discharge, with follow-up periods extending up to a maximum of 5 years ( 11 ). In terms of calibration, one study ( 29 ) reported Hosmer–Lemeshow test results, while six studies ( 11 , 13 , 20 , 25 , 28 , 32 , 35 ) provided calibration curves.…”

“…In terms of calibration, one study ( 29 ) reported Hosmer–Lemeshow test results, while six studies ( 11 , 13 , 20 , 25 , 28 , 32 , 35 ) provided calibration curves. In addition, nine studies ( 12 , 21 , 26 , 27 , 30 , 31 , 34 , 36 , 38 ) reported both calibration curves and test results. A total of 7 studies ( 22 – 24 , 30 , 33 , 37 , 39 ) did not report calibration information.…”

“…In addition, seven studies (7/23, 30.4%) adhered to the recommendation citing existing models when introducing the medical background in the Introduction section ( 11 – 13 , 29 , 31 , 32 , 35 ) (Item 3a). Nine studies ( 11 – 13 , 32 – 34 , 36 , 37 , 39 ) (9/23, 39.1%) provided supplementary information (Item 21). Four items were incompletely reported (50%–80% adherence).…”