Interpretable machine learning for predicting 28-day all-cause in-hospital mortality for hypertensive ischemic or hemorrhagic stroke patients in the ICU: a multi-center retrospective cohort study with internal and external cross-validation.

Background: Timely and accurate outcome prediction plays a critical role in guiding clinical decisions for hypertensive ischemic or hemorrhagic stroke patients admitted to the ICU. However, interpreting and translating the predictive models into clinical applications are as important as the prediction itself. This study aimed to develop an interpretable machine learning (IML) model that accurately predicts 28-day all-cause mortality in hypertensive ischemic or hemorrhagic stroke patients. Methods: A total of 4,274 hypertensive ischemic or hemorrhagic stroke patients admitted to the ICU in the USA from multicenter cohorts were included in this study to develop and validate the IML model. Five machine learning (ML) models were developed, including artificial neural network (ANN), gradient boosting machine (GBM), eXtreme Gradient Boosting (XGBoost), logistic regression (LR), and support vector machine (SVM), to predict mortality using the MIMIC-IV and eICU-CRD database in the USA. Feature selection was performed using the Least Absolute Shrinkage and Selection Operator (LASSO) algorithm. Model performance was evaluated based on the area under the curve (AUC), accuracy, positive predictive value (PPV), and negative predictive value (NPV). The ML model with the best predictive performance was selected for interpretability analysis. Finally, the SHapley Additive exPlanations (SHAP) method was employed to evaluate the risk of all-cause in-hospital mortality among hypertensive ischemic or hemorrhagic stroke patients admitted to the ICU. Results: The XGBoost model demonstrated the best predictive performance, with the AUC values of 0.822, 0.739, and 0.700 in the training, test, and external cohorts, respectively. The analysis of feature importance revealed that age, ethnicity, white blood cell (WBC), hyperlipidemia, mean corpuscular volume (MCV), glucose, pulse oximeter oxygen saturation (SpO2), serum calcium, red blood cell distribution width (RDW), blood urea nitrogen (BUN), and bicarbonate were the 11 most important features. The SHAP plots were employed to interpret the XGBoost model. Conclusions: The XGBoost model accurately predicted 28-day all-cause in-hospital mortality among hypertensive ischemic or hemorrhagic stroke patients admitted to the ICU. The SHAP method can provide explicit explanations of personalized risk prediction, which can aid physicians in understanding the model.

Severity of functional tricuspid regurgitation is associated with mortality in patients with pulmonary hypertension in long-term follow-up.

Pulmonary hypertension (PH) is a hemodynamic and pathophysiologic state present in many cardiovascular, respiratory, and systemic diseases. PH is considered to have a higher risk of cardiovascular events and mortality. The most common type of functional tricuspid regurgitation (FTR) is associated with PH. The aim of this study was to evaluate the association between FTR severity and mortality in PH in western China. This is a retrospective analysis in PH patients and all patients underwent right-heart catheterization (RHC) for hemodynamic measurements. The FTR severity was determined according to the guidelines. Uni- and multivariate analyses were used to identify risk factors for mortality. From 2015 to 2021, 136 patients with PH with a median age of 50 years (interquartile range [IQR]: 35-64 years). During 26-month median follow-up (mean 27.7 ± 15.1 months), 40 (29.2%) patients died (mean after 21.7 ± 14.1 months). In the univariate Cox regression analysis, World Health Organization functional class (WHO FC) III/IV, elevated B-type natriuretic peptide, pulmonary vascular resistance (≥16.2 Wood units), pulmonary artery oxygen saturation, severe FTR and right ventricular diameter/left ventricular diameter (≥0.62) were significantly associated with mortality. In the multivariate Cox regression analysis, severe FTR, WHO FC III/IV, and right ventricular end-diastolic pressure (RVEDP) were risk factors for mortality. Severe FTR at baseline was strongly associated with mortality in both precapillary and postcapillary PH patients, independent of the other risk factors as RVEDP, HO FC III/IV, optimal pulmonary arterial hypertension targeted therapy.

Development and assessment of scoring model for ICU stay and mortality prediction after emergency admissions in ischemic heart disease: a retrospective study of MIMIC-IV databases.

Ischemic heart disease (IHD) is the leading cause of death and emergency department (ED) admission. We aimed to develop more accurate and straightforward scoring models to optimize the triaging of IHD patients in ED. This was a retrospective study based on the MIMIC-IV database. Scoring models were established by AutoScore formwork based on machine learning algorithm. The predictive power was measured by the area under the curve in the receiver operating characteristic analysis, with the prediction of intensive care unit (ICU) stay, 3d-death, 7d-death, and 30d-death after emergency admission. A total of 8381 IHD patients were included (median patient age, 71 years, 95% CI 62-81; 3035 [36%] female), in which 5867 episodes were randomly assigned to the training set, 838 to validation set, and 1676 to testing set. In total cohort, there were 2551 (30%) patients transferred into ICU; the mortality rates were 1% at 3 days, 3% at 7 days, and 7% at 30 days. In the testing cohort, the areas under the curve of scoring models for shorter and longer term outcomes prediction were 0.7551 (95% CI 0.7297-0.7805) for ICU stay, 0.7856 (95% CI 0.7166-0.8545) for 3d-death, 0.7371 (95% CI 0.6665-0.8077) for 7d-death, and 0.7407 (95% CI 0.6972-0.7842) for 30d-death. This newly accurate and parsimonious scoring models present good discriminative performance for predicting the possibility of transferring to ICU, 3d-death, 7d-death, and 30d-death in IHD patients visiting ED.

Interpretable machine learning for 28-day all-cause in-hospital mortality prediction in critically ill patients with heart failure combined with hypertension: A retrospective cohort study based on medical information mart for intensive care database-IV and eICU databases.

Background: Heart failure (HF) combined with hypertension is an extremely important cause of in-hospital mortality, especially for the intensive care unit (ICU) patients. However, under intense working pressure, the medical staff are easily overwhelmed by the large number of clinical signals generated in the ICU, which may lead to treatment delay, sub-optimal care, or even wrong clinical decisions. Individual risk stratification is an essential strategy for managing ICU patients with HF combined with hypertension. Artificial intelligence, especially machine learning (ML), can develop superior models to predict the prognosis of these patients. This study aimed to develop a machine learning method to predict the 28-day mortality for ICU patients with HF combined with hypertension. Methods: We enrolled all critically ill patients with HF combined with hypertension in the Medical Information Mart for IntensiveCare Database-IV (MIMIC-IV, v.1.4) and the eICU Collaborative Research Database (eICU-CRD) from 2008 to 2019. Subsequently, MIMIC-IV was divided into training cohort and testing cohort in an 8:2 ratio, and eICU-CRD was designated as the external validation cohort. The least absolute shrinkage and selection operator (LASSO) Cox regression with internal tenfold cross-validation was used for data dimension reduction and identifying the most valuable predictive features for 28-day mortality. Based on its accuracy and area under the curve (AUC), the best model in the validation cohort was selected. In addition, we utilized the Shapley Additive Explanations (SHAP) method to highlight the importance of model features, analyze the impact of individual features on model output, and visualize an individual's Shapley values. Results: A total of 3,458 and 6582 patients with HF combined with hypertension in MIMIC-IV and eICU-CRD were included. The patients, including 1,756 males, had a median (Q1, Q3) age of 75 (65, 84) years. After selection, 22 out of a total of 58 clinical parameters were extracted to develop the machine-learning models. Among four constructed models, the Neural Networks (NN) model performed the best predictive performance with an AUC of 0.764 and 0.674 in the test cohort and external validation cohort, respectively. In addition, a simplified model including seven variables was built based on NN, which also had good predictive performance (AUC: 0.741). Feature importance analysis showed that age, mechanical ventilation (MECHVENT), chloride, bun, anion gap, paraplegia, rdw (RDW), hyperlipidemia, peripheral capillary oxygen saturation (SpO2), respiratory rate, cerebrovascular disease, heart rate, white blood cell (WBC), international normalized ratio (INR), mean corpuscular hemoglobin concentration (MCHC), glucose, AIDS, mean corpuscular volume (MCV), N-terminal pro-brain natriuretic peptide (Npro. BNP), calcium, renal replacement therapy (RRT), and partial thromboplastin time (PTT) were the top 22 features of the NN model with the greatest impact. Finally, after hyperparameter optimization, SHAP plots were employed to make the NN-based model interpretable with an analytical description of how the constructed model visualizes the prediction of death. Conclusion: We developed a predictive model to predict the 28-day mortality for ICU patients with HF combined with hypertension, which proved superior to the traditional logistic regression analysis. The SHAP method enables machine learning models to be more interpretable, thereby helping clinicians to better understand the reasoning behind the outcome and assess in-hospital outcomes for critically ill patients.

Pulmonary artery enlargement predicts poor survival in patients with COPD: A meta-analysis.

No previous meta-analysis has evaluated the relationship between pulmonary artery enlargement (PAE) measured by computed tomography (CT) and prognosis for patients with chronic obstructive pulmonary disease (COPD). Recently, several studies have suggested poor survival and reduced exercise capacity in COPD patients with PAE on CT scan, but there were conflicting results. We aimed to assess the prognostic value of PAE-CT in patients with COPD. Relevant studies were identified by searching major databases. Pooled outcomes were determined to assess the prognostic value of PAE-CT in COPD patients. Eighteen studies including 5694 participants were included. PAE indicated higher mortality in COPD patients (odds ratio [OR] = 3.06; 95% confidence interval [95% CI]: 1.76-5.32; p < 0.0001), shorter 6-minute walk distance (mean difference [MD] = -67.53 m; 95% CI: -85.98 to -49.08; p < 0.00001), higher pulmonary artery systolic pressure (MD = 15.65 mmHg; 95% CI: 13.20-18.11; p < 0.00001), longer length of hospital stay (MD = 2.92 days; 95% CI: 0.71-5.12; p = 0.009) and more severe symptom such as dyspnea (COPD Assessment Test MD = 3.14; 95% CI: 2.48-3.81; p < 0.00001). We also conducted a subgroup analysis regarding the lung function and blood gas analysis for a stable period and acute exacerbation of COPD patients. In conclusion, PAE is significantly associated with mortality, lower exercise tolerance, and poor quality of life in patients with COPD. PAE may serve as a novel imaging biomarker for risk stratification in patients with COPD in the future.

Risk of incident atrial fibrillation with low-to-moderate alcohol consumption is associated with gender, region, alcohol category: a systematic review and meta-analysis.

AIMS: The association between low-to-moderate alcohol consumption and atrial fibrillation (AF) has yet to be fully elucidated. The main purpose of this meta-analysis was to estimate the risk of incident AF related to low-to-moderate alcohol consumption. METHODS AND RESULTS: A meta-analysis was performed on 13 publications discussing the estimated risk for AF with habitual low-to-moderate alcohol intake in 10 266 315 participants. Graphical augmentations to the funnel plots were used to illustrate the potential impact of additional evidence on the current meta-analysis. Thirteen eligible studies were included in this meta-analysis. We found that moderate alcohol consumption was associated with an increased risk of incident AF in males [hazard ratio (HR) 1.09, 95% confidence interval (CI): 1.07-1.11, P < 0.00001], Europeans (HR 1.32, 95% CI: 1.23-1.42, P < 0.00001), and Asians (HR 1.09, 95% CI: 1.07-1.11, P < 0.00001). Moderate beer consumption was associated with an increased risk of developing AF (HR 1.11, 95% CI: 1.02-1.21, P = 0.01). Low alcohol consumption conferred an increased risk of AF in males (HR 1.14, 95% CI: 1.01-1.28, P = 0.04) and Europeans (HR 1.12, 95% CI: 1.07-1.17, P < 0.00001). CONCLUSIONS: This analysis represents the increased risk of incident AF in males, Europeans, and Asians at moderate alcohol consumption levels and in males and Europeans at low alcohol consumption levels. Those who drink any preferred alcohol beverage at moderate levels should be cautious for incident AF. More studies are warranted to find those factors that influence alcohol's effect on predisposing AF.

Twenty-eight-day in-hospital mortality prediction for elderly patients with ischemic stroke in the intensive care unit: Interpretable machine learning models.

Background: Risk stratification of elderly patients with ischemic stroke (IS) who are admitted to the intensive care unit (ICU) remains a challenging task. This study aims to establish and validate predictive models that are based on novel machine learning (ML) algorithms for 28-day in-hospital mortality in elderly patients with IS who were admitted to the ICU. Methods: Data of elderly patients with IS were extracted from the electronic intensive care unit (eICU) Collaborative Research Database (eICU-CRD) records of those elderly patients admitted between 2014 and 2015. All selected participants were randomly divided into two sets: a training set and a validation set in the ratio of 8:2. ML algorithms, such as Naïve Bayes (NB), eXtreme Gradient Boosting (xgboost), and logistic regression (LR), were applied for model construction utilizing 10-fold cross-validation. The performance of models was measured by the area under the receiver operating characteristic curve (AUC) analysis and accuracy. The present study uses interpretable ML methods to provide insight into the model's prediction and outcome using the SHapley Additive exPlanations (SHAP) method. Results: As regards the population demographics and clinical characteristics, the analysis in the present study included 1,236 elderly patients with IS in the ICU, of whom 164 (13.3%) died during hospitalization. As regards feature selection, a total of eight features were selected for model construction. In the training set, both the xgboost and NB models showed specificity values of 0.989 and 0.767, respectively. In the internal validation set, the xgboost model identified patients who died with an AUC value of 0.733 better than the LR model which identified patients who died with an AUC value of 0.627 or the NB model 0.672. Conclusion: The xgboost model shows the best predictive performance that predicts mortality in elderly patients with IS in the ICU. By making the ML model explainable, physicians would be able to understand better the reasoning behind the outcome.

Association of statin use and the risk of recurrent pulmonary embolism in real-world Chinese population.

BACKGROUND: Previous studies have suggested that statins exert protective effects against venous thromboembolism. However, few randomized studies have explicitly concentrated on patients with pulmonary embolism. Thus far, evidence of the effect of statins on the pulmonary embolism recurrence in China remains lacking. METHODS: A retrospective analysis was conducted utilizing our University database. Patients with an International Coding of Diseases-defined diagnosis of pulmonary embolism from 1 January 2017 to 31 December 2019 were included. The patients were divided into two groups, namely, with statin or without statin treatment. Propensity score matching was applied to balance the covariates between the comparison groups. Univariate analysis and multivariable logistic regression were performed to analyze the association between statin use and pulmonary embolism recurrence. RESULTS: A total of 365 patients diagnosed with pulmonary embolism were included in the research. Pulmonary embolism recurrence accounted for 15.1% of the patients and was observed during the entire study period. In the initial population, no significant difference in recurrence was observed between the groups with and without statins treatment (statin 15.6% vs. non-statin 14.9%, p = 0.860). After propensity score matching, multivariate logistic regression analysis revealed that the odds ratio of pulmonary embolism recurrence in the statin users was 0.489 (95% confidence interval 0.190-1.258, p = 0.138). CONCLUSIONS: Our study provides no support for the use of statins as an adjunctive therapy in patients with pulmonary embolism at the initiated time of diagnosis or as a prophylactical plan when anticoagulation is discontinued attempting to reduce the risk of recurrence.

The Efficacy and Safety of Pulmonary Vasodilators in Pediatric Pulmonary Hypertension (PH): A Systematic Review and Meta-analysis.

Background: We performed a meta-analysis to evaluate the efficacy and safety of pulmonary vasodilators in pediatric pulmonary hypertension (PH) patients. Methods: We searched electronic databases including PubMed, EMBASE, and the Cochrane Library up to May 2020, and conducted a subgroup analysis for pulmonary vasodilators or underlying disease. Results: Fifteen studies with 719 pediatric PH patients were included in the meta-analysis. Adverse events did not differ (p = 0.11, I 2 = 15%) between the pulmonary vasodilators group and the control group, neither in the subgroups. In total, compared with the control group treatment, pulmonary vasodilators significantly decreased the mortality (p = 0.002), mean pulmonary artery pressure (mPAP, p = 0.02), and mechanical ventilation duration (p = 0.03), also improved the oxygenation index (OI, p = 0.01). In the persistent pulmonary hypertension of the newborn (PPHN) subgroup, phosphodiesterase type 5 inhibitors (PDE5i) significantly reduced mortality (p = 0.03), OI (p = 0.007) and mechanical ventilation duration (p = 0.004). Administration of endothelin receptor antagonists (ERAs) improved OI (p = 0.04) and mechanical ventilation duration (p < 0.00001) in PPHN. We also found that in the pediatric pulmonary arterial hypertension (PPAH) subgroup, mPAP was pronouncedly declined with ERAs (p = 0.006). Systolic pulmonary artery pressure (sPAP, p < 0.0001) and pulmonary arterial/aortic pressure (PA/AO, p < 0.00001) were significantly relieved with PDE5i, partial pressure of arterial oxygen (PaO2) was improved with prostacyclin in postoperative PH (POPH) subgroup (p = 0.001). Compared with the control group, pulmonary vasodilators could significantly decrease PA/AO pressure (p < 0.00001) and OI (p < 0.00001) in the short-term (duration <7 days) follow-up subgroup, improve mPAP (p = 0.03) and PaO2 (p = 0.01) in the mid-term (7-30 days) follow-up subgroup, also decrease mortality, mPAP (p = 0.0001), PA/AO pressure (p = 0.0007), duration of mechanical ventilation (p = 0.004), and ICU stay (p < 0.00001) in the long-term follow subgroup (>30 days). Conclusion: Pulmonary vasodilators decrease the mortality in pediatric PH patients, improve the respiratory and hemodynamic parameters, reduce the mechanical ventilation duration.

Multidisciplinary Team Managements and Clinical Outcomes in Patients With Pulmonary Arterial Hypertension During the Perinatal Period.

Background: Pulmonary arterial hypertension (PAH) patients with pregnancy have high maternal mortality. This study aimed to provide clinical evidence with multidisciplinary team (MDT) management and to evaluate the clinical outcomes in PAH patients during the perinatal period. Methods: We conducted a retrospective evaluation of PAH patients pregnant at the First Affiliated Hospital of Chongqing Medical University between May 2015 and May 2021. Results: Twenty-two patients (24 pregnancies) were included in this study and received MDT management, and 21 pregnancies chose to continue pregnancy with cesarean section. Nine (37.5%) were first-time pregnancies at 27.78 ± 6.16 years old, and 15 (62.5%) were multiple pregnancies at 30.73 ± 3.71 years old. The average gestational week at hospitalization and delivery were 29.38 ± 8.63 weeks and 32.37 ± 7.20 weeks, individually. Twenty-one (87.5%) pregnancies received single or combined pulmonary vasodilators. The maternal survival rate of PAH patients reached 91.7%. Fifteen (62.5%) pregnancies were complicated with severe adverse events. Patients with complicated adverse events showed lower percutaneous oxygen saturation (SpO2), lower albumin, lower fibrinogen, higher pulmonary artery systolic pressure (PASP), higher blood pressure, longer activated partial thromboplastin time, and longer coagulation time. Fourteen (66.7%) pregnancies with cesarean sections were prematurely delivered and 85.7% newborns who survived after the operation remained alive. Conclusion: The survival rate of parturients with PAH was improved in relation to MDT and pulmonary vasodilator therapy during the perinatal period compared with previous studies. SpO2, albumin, PASP, blood pressure, and coagulation function should be monitored carefully in PAH patients during pregnancy.