A Novel Predictive Model for In-Hospital Mortality Based on a Combination of Multiple Blood Variables in Patients with ST-Segment-Elevation Myocardial Infarction

Predictive value of machine learning for in-hospital mortality risk in acute myocardial infarction: A systematic review and meta-analysis

BACKGROUND

Machine learning (ML) models have been constructed to predict the risk of in-hospital mortality in patients with myocardial infarction (MI). Due to diverse ML models and modeling variables, along with the significant imbalance in data, the predictive accuracy of these models remains controversial.

OBJECTIVE

This study aimed to review the accuracy of ML in predicting in-hospital mortality risk in MI patients and to provide evidence-based advices for the development or updating of clinical tools.

METHODS

PubMed, Embase, Cochrane, and Web of Science databases were searched, up to June 4, 2024. PROBAST and ChAMAI checklist are utilized to assess the risk of bias in the included studies. Since the included studies constructed models based on severely unbalanced datasets, subgroup analyses were conducted by the type of dataset (balanced data, unbalanced data, model type).

RESULTS

This meta-analysis included 32 studies. In the validation set, the pooled C-index, sensitivity, and specificity of prediction models based on balanced data were 0.83 (95 % CI: 0.795-0.866), 0.81 (95 % CI: 0.79-0.84), and 0.82 (95 % CI: 0.78-0.86), respectively. In the validation set, the pooled C-index, sensitivity, and specificity of ML models based on imbalanced data were 0.815 (95 % CI: 0.789-0.842), 0.66 (95 % CI: 0.60-0.72), and 0.84 (95 % CI: 0.83-0.85), respectively.

CONCLUSIONS

ML models such as LR, SVM, and RF exhibit high sensitivity and specificity in predicting in-hospital mortality in MI patients. However, their sensitivity is not superior to well-established scoring tools. Mitigating the impact of imbalanced data on ML models remains challenging.

Predicting In-Hospital Mortality in Patients With Acute Myocardial Infarction: A Comparison of Machine Learning Approaches

BACKGROUND

Acute myocardial infarction (AMI) remains a leading global cause of mortality. This study explores predictors of in-hospital mortality among AMI patients using advanced machine learning (ML) techniques.

METHODS

Data from 7422 AMI patients treated with percutaneous coronary intervention (PCI) at Tehran Heart Center (2015-2021) were analyzed. Fifty-eight clinical, demographic, and laboratory variables were evaluated. Seven ML algorithms, including Random Forest (RF), logistic regression with LASSO, and XGBoost, were implemented. The data set was divided into training (70%) and testing (30%) subsets, with fivefold cross-validation. The class imbalance was addressed using the synthetic minority oversampling technique (SMOTE). Model predictions were interpreted using SHapley Additive exPlanations (SHAP).

RESULTS

In-hospital mortality occurred in 129 patients (1.74%). RF achieved the highest predictive performance, with an area under the curve (AUC) of 0.924 (95% CI 0.893-0.954), followed by XGBoost (AUC 0.905) and logistic regression with LASSO (AUC 0.893). Sensitivity analysis in STEMI patients confirmed RF's robust performance (AUC 0.900). SHAP analysis identified key predictors, including lower left ventricular ejection fraction (LVEF; 33.24% vs. 43.46% in survivors, p < 0.001), higher fasting blood glucose (190.38 vs. 132.29 mg/dL, p < 0.001), elevated serum creatinine, advanced age (70.92 vs. 61.88 years, p < 0.001), and lower LDL-C levels. Conversely, BMI showed no significant association (p = 0.456).

CONCLUSION

ML algorithms, particularly RF, effectively predict in-hospital mortality in AMI patients, highlighting critical predictors such as LVEF and biochemical markers. These insights offer valuable tools for enhancing clinical decision-making and improving patient outcomes.

Inflammatory and Metabolic Predictors of Mortality in Pulmonary Thromboembolism: A Focus on the Triglyceride-Glucose Index and Pan-Immune Inflammation Value

Objectives: We aimed to evaluate the importance of metabolic and inflammatory markers, specifically the Triglyceride-Glucose Index (TGI) and pan-immune inflammation value (PIV), in predicting mortality among patients diagnosed with pulmonary thromboembolism (PTE). Materials and Methods: A total of 450 patients diagnosed with PTE between December 2018 and December 2023 were included in his study. The diagnosis of PTE was confirmed by clinical presentation, laboratory tests, and imaging studies such as computed tomography pulmonary angiography (CTPA). Data were obtained from medical records, including demographic information, medical history, laboratory results, and clinical outcomes. Results: In terms of age, non-survivors were older on average (66.1 ± 11.8 years) compared to survivors (58.3 ± 12.4 years) (p = 0.03). In terms of gender, 55% of non-survivors and 45% of survivors were male (p = 0.111). Non-survivors had higher BMIs (28.3 ± 5.1) than survivors (25.7 ± 4.5) (p = 0.04). In terms of hypertension, 40% of non-survivors and 30% of survivors had hypertension (p = 0.041). In terms of diabetes, 35% of those who did not survive and 20% of those who survived had diabetes (p = 0.001). In terms of smoking, 25% of non-survivors and 15% of survivors smoke (p = 0.022). In terms of TGI, non-survivors had higher TGI values (12.1 ± 1.5) than survivors (5.9 ± 1.2) (p < 0.001). In terms of PIV, non-survivors had significantly higher PIV (878.2 ± 85.4) than survivors (254.5 ± 61.1) (p < 0.001). The risk factors found to be significantly associated with differentiation in the multiple logistic regression analysis included age, BMI, TGI, and PIV (p = 0.005, p = 0.002, p = 0.013, and 0.022, respectively). As a result, according to ROC analysis for patients who are non-survivors, age, BMI, TGI, and PIV were significant prognostic factors. The cut-off points for these values were >60, >27, >10, and >500, respectively. Conclusions: the TGI and PIV are strong markers for predicting mortality in PTE patients. The independent predictive value of age and BMI further demonstrates their role in risk stratification. We think that high TGI values and PIVs reflect underlying metabolic and inflammatory disorders that may contribute to worse outcomes in these patients.

The predictive value of machine learning for mortality risk in patients with acute coronary syndromes: a systematic review and meta-analysis

BACKGROUND

Acute coronary syndromes (ACS) are the leading cause of global death. Optimizing mortality risk prediction and early identification of high-risk patients is essential for developing targeted prevention strategies. Many researchers have built machine learning (ML) models to predict the mortality risk in ACS patients. Our meta-analysis aimed to evaluate the predictive value of various ML models in predicting death in ACS patients at different times.

METHODS

PubMed, Embase, Web of Science, and Cochrane Library were searched systematically from database establishment to March 12, 2022 for studies developing or validating at least one ML predictive model for death in ACS patients. We used PROBAST to assess the risk of bias in the reported predictive models and a random-effects model to assess the pooled C-index and accuracy of these models.

RESULTS

Fifty papers were included, involving 216 ML prediction models, 119 of which were externally validated. The combined C-index of the ML models in the validation cohort predicting the in-hospital mortality, 30-day mortality, 3- or 6-month mortality, and 1 year or above mortality in ACS patients were 0.8633 (95% CI 0.8467-0.8802), 0.8296 (95% CI 0.8134-0.8462), 0.8205 (95% CI 0.7881-0.8541), and 0.8197 (95% CI 0.8042-0.8354), respectively, with the corresponding combined accuracy of 0.8569 (95% CI 0.8411-0.8715), 0.8282 (95% CI 0.7922-0.8591), 0.7303 (95% CI 0.7184-0.7418), and 0.7837 (95% CI 0.7455-0.8175), indicating that the ML models were relatively excellent in predicting ACS mortality at different times. Furthermore, common predictors of death in ML models included age, sex, systolic blood pressure, serum creatinine, Killip class, heart rate, diastolic blood pressure, blood glucose, and hemoglobin.

CONCLUSIONS

The ML models had excellent predictive power for mortality in ACS, and the methodologies may need to be addressed before they can be used in clinical practice.

FIB-4 Index and Neutrophil-to-Lymphocyte-Ratio as Death Predictor in Coronary Artery Disease Patients

Nonalcoholic fatty liver disease (NAFLD)-associated liver fibrosis is likely related to coronary artery disease (CAD) by the mediation of systemic inflammation. This study aimed at evaluating the predictive value of neutrophil-to-lymphocyte-ratio (NLR) and fibrosis-4 index (FIB-4), indices of inflammation and fibrosis, respectively, on CAD mortality. Data from 1460 CAD patients (1151 males, age: 68 ± 10 years, mean ± SD) were retrospectively analyzed. Over a median follow-up of 26 months (interquartile range (IQR) 12−45), 94 deaths were recorded. Kaplan−Meier survival analysis revealed worse outcomes in patients with elevation of one or both biomarkers (FIB-4 > 3.25 or/and NLR > 2.04, log-rank p-value < 0.001). In multivariate Cox regression analysis, the elevation of one biomarker (NLR or FIB-4) still confers a significant independent risk for mortality (hazard ratio (HR) = 1.7, 95% confidence interval (95% CI): 1.1−2.7, p = 0.023), whereas an increase in both biomarkers confers a risk corresponding to HR = 3.5 (95% CI: 1.6−7.8, p = 0.002). Categorization of patients with elevated FIB-4/NLR could provide valuable information for risk stratification and reduction of residual risk in CAD patients.

Prognostic impact of incident left ventricular systolic dysfunction after myocardial infarction

Introduction

We sought to investigate the prognostic impact of incident left ventricular (LV) systolic dysfunction at the chronic phase of acute myocardial infarction (AMI).

Materials and methods

Among 2,266 consecutive patients admitted for AMI, 1,330 patients with LV ejection fraction (LVEF) ≥ 40% during hospitalization who had LVEF data at 6 months after AMI were analyzed. Patients were divided into three subgroups based on LVEF at 6 months: reduced-LVEF (&lt;40%), mid-range-LVEF (≥ 40% and &lt; 50%) and preserved-LVEF (≥ 50%). Occurrence of a composite of hospitalization for heart failure or cardiovascular death after 6 months of AMI was the primary endpoint. The prognostic impact of LVEF at 6 months was assessed with a multivariate-adjusted Cox model.

Results

Overall, the mean patient age was 67.5 ± 11.9 years, and LVEF during initial hospitalization was 59.4 ± 9.1%. The median (interquartile range) duration of follow-up was 3.0 (1.5–4.8) years, and the primary endpoint occurred in 35/1330 (2.6%) patients (13/69 [18.8%] in the reduced-LVEF, 9/265 [3.4%] in the mid-range-LVEF, and 13/996 [1.3%] in the preserved-LVEF category). The adjusted hazard ratio for the primary endpoint in the reduced-LVEF vs. mid-range-LVEF category and in the reduced-LVEF vs. preserved-LVEF category was 4.71 (95% confidence interval [CI], 1.83 to 12.13; p &lt; 0.001) and 14.37 (95% CI, 5.38 to 38.36; p &lt; 0.001), respectively.

Conclusion

Incident LV systolic dysfunction at the chronic phase after AMI was significantly associated with long-term adverse outcomes. Even in AMI survivors without LV systolic dysfunction at the time of AMI, post-AMI reassessment and careful monitoring of LVEF are required to identify patients at risk.

Development of a Laboratory Risk-Score Model to Predict One-Year Mortality in Acute Myocardial Infarction Survivors

The high post-discharge mortality rate of acute myocardial infarction (AMI) survivors is concerning, indicating a need for reliable, easy-to-use risk prediction tools. We aimed to examine if a combined pre-procedural blood testing risk model predicts one-year mortality in AMI survivors. Overall, 1355 consecutive AMI patients who received primary coronary revascularization were divided into derivation (n = 949) and validation (n = 406) cohorts. A risk-score model of parameters from pre-procedural routine blood testing on admission was generated. In the derivation cohort, multivariable analysis demonstrated that hemoglobin < 11 g/dL (odds ratio (OR) 4.01), estimated glomerular filtration rate < 30 mL/min/1.73 m2 (OR 3.75), albumin < 3.8 mg/dL (OR 3.37), and high-sensitivity troponin I > 2560 ng/L (OR 3.78) were significantly associated with one-year mortality after discharge. An increased risk score, assigned from 0 to 4 points according to the counts of selected variables, was significantly associated with higher one-year mortality in both cohorts (p < 0.001). Receiver-operating characteristics curve analyses of risk models demonstrated adequate discrimination between patients with and without one-year death (area under the curve (95% confidence interval) 0.850 (0.756−0.912) in the derivation cohort; 0.820 (0.664−0.913) in the validation cohort). Our laboratory risk-score model can be useful for predicting one-year mortality in AMI survivors.

Baseline platelet count in percutaneous coronary intervention: a dose-response meta-analysis

OBJECTIVES

The nature of the relationship between baseline platelet count and clinical outcomes following percutaneous coronary intervention (PCI) is unclear. We undertook dose-response and pairwise meta-analyses to better describe the prognostic value of the initial platelet count and clinical endpoints in patients after PCI.

METHODS

A search of PubMed, Scopus and Web of Science (up to 9 October 2021) was performed to identify studies that evaluated the association between platelet count and clinical outcomes following PCI. The primary outcomes of interest were all-cause mortality, major adverse cardiovascular events (MACE) and major bleeding. We performed random-effects pairwise and one-stage dose-response meta-analyses by calculating HRs and 95% CIs.

RESULTS

The meta-analysis included 19 studies with 217 459 patients. We report a J-shaped relationship between baseline thrombocyte counts and all-cause death, MACE and major bleeding at follow-up. The risk of haemorrhagic events exceeded the risk of thrombotic events at low platelet counts (<175×109/L), while a predominant ischaemic risk was observed at high platelet counts (>250×109/L). Pairwise meta-analyses revealed a robust link between initial platelet counts and the risk of postdischarge all-cause mortality, major bleeding (for thrombocytopenia: HR 1.39, 95% CI 1.30 to 1.49; HR 1.51, 95% CI 1.15 to 2.00, respectively) and future death from any cause and MACE (thrombocytosis: HR 1.60, 95% CI 1.29 to 1.98; HR 1.47, 95% CI 1.22 to 1.78, respectively).

CONCLUSION

Low platelet counts were associated with the predominant bleeding risk, while high platelet counts were only associated with the ischaemic events.

PROSPERO REGISTRATION NUMBER

CRD42021283270.

Serum Albumin and Bleeding Events After Percutaneous Coronary Intervention in Patients With Acute Myocardial Infarction (from the HAGAKURE-ACS Registry)

Low serum albumin (SA) on admission in patients with acute myocardial infarction (AMI) has been reported to be associated with adverse cardiovascular events. The relation between low SA and post-AMI bleeding events is presently unknown. We analyzed 1,724 patients with AMI enrolled in the HAGAKURE-ACS registry who underwent primary percutaneous coronary intervention from January 2014 to December 2018. To assess the influence of low SA at admission, patients were divided into 3 groups according to the albumin tertiles: the low SA group (<3.8 g/100 ml), the middle SA (MSA) group (3.8 to 4.1 g/100 ml), and the normal SA (NSA) group (≥4.2 g/100 ml). The primary end point was the incidence of Global Utilization of Streptokinase and Tissue Plasminogen Activator for Occluded Coronary Arteries moderate/severe bleeding. The cumulative 3-year incidence of the primary end point was significantly higher in the low SA group than in the MSA and NSA groups (30.8% and 11.9% vs 7.7%; p <0.001). In the landmark analysis at 30 days, the cumulative incidences of the primary end point were also significantly higher in the low SA group than in the MSA and NSA groups, both within and beyond 30 days (20.1% and 6.1% vs 3.5%; p <0.001, and 12.4% and 6.2% vs 4.5%; p <0.001, respectively). After adjusting for confounders, the low SA group showed excess risk of bleeding events relative to NSA (hazard ratio 1.56; 95% confidence interval 1.06 to 2.30; p = 0.026), whereas risk of bleeding was neutral in MSA relative to NSA (hazard ratio 0.94; 95% confidence interval 0.63 to 1.34; p = 0.752). In conclusion, low SA at admission was independently associated with higher risk for bleeding events in patients with AMI undergoing percutaneous coronary intervention.

Prognostic value of D-dimer to fibrinogen ratio for patients with acute myocardial infarction

Objective

Myocardial infarction is a severe disease with high in-hospital mortality without aggressive clinical treatment. The study aims to evaluate prognostic worth of D-dimer-to-fibrinogen (FIB) ratio (DFR) for patients with acute myocardial infarction (AMI).

Methods

133 patients (65 (37, 93) years old) from our hospital (China) with AMI were enlisted from January 2017 to December 2019. Patients were assigned into the survivor and nonsurvivor group based on in-hospital outcomes. Receiver operating characteristics (ROC) and multivariate analysis were fulfilled to analyze the prognostic value of DFR.

Results

The degree of DFR in the nonsurvivor group was significantly higher than that in the survivor group ( p < 0.05). Logistic regression analysis presented that DFR (hazard ratio (HR), 2.207; 95% confidence interval (CI), 1.050–4.640; p = 0.037) was independently related with in-hospital death. ROC demonstrated that the area under the curve (AUC) of DFR was = 0.808 (0.725–0.892) (sensitivity, 85.3%; specificity, 69.7%).

Conclusion

DFR might be a new independent predictor of in-hospital death for AMI patients. Further studies are needed to validate this preliminary finding.

Application of machine learning to predict the occurrence of arrhythmia after acute myocardial infarction

Background

Early identification of the occurrence of arrhythmia in patients with acute myocardial infarction plays an essential role in clinical decision-making. The present study attempted to use machine learning (ML) methods to build predictive models of arrhythmia after acute myocardial infarction (AMI).

Methods

A total of 2084 patients with acute myocardial infarction were enrolled in this study. (All data is available on Github: https://github.com/wangsuhuai/AMI-database1.git). The primary outcome is whether tachyarrhythmia occurred during admission containing atrial arrhythmia, ventricular arrhythmia, and supraventricular tachycardia. All data is randomly divided into a training set (80%) and an internal testing set (20%). Apply three machine learning algorithms: decision tree, random forest (RF), and artificial neural network (ANN) to learn the training set to build a model, then use the testing set to evaluate the prediction performance, and compare it with the model built by the Global Registry of Acute Coronary Events (GRACE) risk variable set.

Results

Three ML models predict the occurrence of tachyarrhythmias after AMI. After variable selection, the artificial neural network (ANN) model has reached the highest accuracy rate, which is better than the model constructed using the Grace variable set. After applying SHapley Additive exPlanations (SHAP) to make the model interpretable, the most important features are abnormal wall motion, lesion location, bundle branch block, age, and heart rate. Among them, RBBB (odds ratio [OR]: 4.21; 95% confidence interval [CI]: 2.42–7.02), ≥ 2 ventricular walls motion abnormal (OR: 3.26; 95% CI: 2.01–4.36) and right coronary artery occlusion (OR: 3.00; 95% CI: 1.98–4.56) are significant factors related to arrhythmia after AMI.

Conclusions

We used advanced machine learning methods to build prediction models for tachyarrhythmia after AMI for the first time (especially the ANN model that has the best performance). The current study can supplement the current AMI risk score, provide a reliable evaluation method for the clinic, and broaden the new horizons of ML and clinical research.

Trial registration Clinical Trial Registry No.: ChiCTR2100041960.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12911-021-01667-8.

Basic types of the first-day glycemia in acute myocardial infarction: Prognostic, diagnostic, threshold and target glycemia

We described the importance of stress hyperglycemia (SH) in critical illnesses and their evaluation in the emergency department (ED) and coronary care unit (CCU). Hyperglycemia is found in over half of the patients with suspected acute myocardial infarction (AMI). SH can be used for several purposes in AMI. Receiver operating characteristic curves are needed to find optimal cut-offs to divide blood glucose levels associated with good from bad prognosis in AMI. There is a need for a consensus for pragmatic classification of first day glycemia in order to be useful in a busy ED and CCU.

Single prognostic cut-off value for admission glycemia in acute myocardial infarction has been used although high-risk stems from hyperglycemia as well as from hypoglycemia (a narrative review)

All original articles and meta-analysis use the single cut-off value to distinguish high-risk hyperglycemic from other acute myocardial infarction (AMI) patients. The mortality rate is 3.9 times higher in non-diabetic AMI patients with admission glycemia ≥6.1mmol compared to normoglycemic non-diabetic AMI patients. On the other hand, admission hypoglycemia in AMI is an important predictor of mortality. Because both admission hypo- and hyperglycemia correspond to higher in-hospital mortality, this graph is recognized as "J or U shaped curve". The review suggests two cut-off values for admission glycemia for risk assessment in AMI instead of single one because hypoglycemia as well as hyperglycemia represents a high-risk factor.