Dataset Information

Ensemble machine learning prediction and variable importance analysis of 5-year mortality after cardiac valve and CABG operations.

ABSTRACT: Despite having a similar post-operative complication profile, cardiac valve operations are associated with a higher mortality rate compared to coronary artery bypass grafting (CABG) operations. For long-term mortality, few predictors are known. In this study, we applied an ensemble machine learning (ML) algorithm to 88 routinely collected peri-operative variables to predict 5-year mortality after different types of cardiac operations. The Super Learner algorithm was trained using prospectively collected peri-operative data from 8241 patients who underwent cardiac valve, CABG and combined operations. Model performance and calibration were determined for all models, and variable importance analysis was conducted for all peri-operative parameters. Results showed that the predictive accuracy was the highest for solitary mitral (0.846 [95% CI 0.812-0.880]) and solitary aortic (0.838 [0.813-0.864]) valve operations, confirming that ensemble ML using routine data collected perioperatively can predict 5-year mortality after cardiac operations with high accuracy. Additionally, post-operative urea was identified as a novel and strong predictor of mortality for several types of operation, having a seemingly additive effect to better known risk factors such as age and postoperative creatinine.

SUBMITTER: Castela Forte J

PROVIDER: S-EPMC7876023 | biostudies-literature | 2021 Feb

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Ensemble machine learning prediction and variable importance analysis of 5-year mortality after cardiac valve and CABG operations.

Castela Forte José J Mungroop Hubert E HE de Geus Fred F van der Grinten Maureen L ML Bouma Hjalmar R HR Pettilä Ville V Scheeren Thomas W L TWL Nijsten Maarten W N MWN Mariani Massimo A MA van der Horst Iwan C C ICC Henning Robert H RH Wiering Marco A MA Epema Anne H AH

Scientific reports 20210210 1

Despite having a similar post-operative complication profile, cardiac valve operations are associated with a higher mortality rate compared to coronary artery bypass grafting (CABG) operations. For long-term mortality, few predictors are known. In this study, we applied an ensemble machine learning (ML) algorithm to 88 routinely collected peri-operative variables to predict 5-year mortality after different types of cardiac operations. The Super Learner algorithm was trained using prospectively c ...[more]

PMID: 33568739

Similar Datasets

Project description:BackgroundSocial-environmental data obtained from the US Census is an important resource for understanding health disparities, but rarely is the full dataset utilized for analysis. A barrier to incorporating the full data is a lack of solid recommendations for variable selection, with researchers often hand-selecting a few variables. Thus, we evaluated the ability of empirical machine learning approaches to identify social-environmental factors having a true association with a health outcome.MethodsWe compared several popular machine learning methods, including penalized regressions (e.g. lasso, elastic net), and tree ensemble methods. Via simulation, we assessed the methods' ability to identify census variables truly associated with binary and continuous outcomes while minimizing false positive results (10 true associations, 1000 total variables). We applied the most promising method to the full census data (p?=?14,663 variables) linked to prostate cancer registry data (n?=?76,186 cases) to identify social-environmental factors associated with advanced prostate cancer.ResultsIn simulations, we found that elastic net identified many true-positive variables, while lasso provided good control of false positives. Using a combined measure of accuracy, hierarchical clustering based on Spearman's correlation with sparse group lasso regression performed the best overall. Bayesian Adaptive Regression Trees outperformed other tree ensemble methods, but not the sparse group lasso. In the full dataset, the sparse group lasso successfully identified a subset of variables, three of which replicated earlier findings.ConclusionsThis analysis demonstrated the potential of empirical machine learning approaches to identify a small subset of census variables having a true association with the outcome, and that replicate across empiric methods. Sparse clustered regression models performed best, as they identified many true positive variables while controlling false positive discoveries.

Project description:BackgroundDespite recommendations regarding the use of intraoperative transesophageal echocardiography (TEE), there is no randomized evidence to support its use in cardiac valve surgery. The purpose of this study was to compare the clinical outcomes of patients undergoing open cardiac valve repair or replacement surgery with and without transesophageal echocardiographic monitoring. The hypothesis was that transesophageal echocardiographic monitoring would be associated with lower 30-day mortality and shorter length of hospitalization.MethodsIn this observational retrospective cohort study, Medicare claims were used to test the association between perioperative TEE and 30-day all-cause mortality and length of hospitalization among patients undergoing open cardiac valve repair or replacement surgery between January 1, 2010, and October 1, 2015. Baseline characteristics were defined by inpatient and outpatient claims. Medicare death records were used to ascertain 30-day mortality. Statistical analyses included regression models and propensity score matching.ResultsA total of 219,238 patients underwent open cardiac valve surgery, of whom 85% underwent TEE. Patients who underwent TEE were significantly older and had greater comorbidities. After adjusting for patient demographics, clinical comorbidities, surgical characteristics, and hospital factors, including annual surgical volume, the TEE group had a lower adjusted odds of 30-day mortality (odds ratio, 0.77; 95% CI, 0.73 to 0.82; P < .001), with no difference in length of hospitalization (<0.01%; 95% CI, -0.61% to 0.62%; P = .99). Results were similar across all analyses, including a propensity score-matched cohort.ConclusionsTransesophageal echocardiographic monitoring in cardiac valve repair or replacement surgery was associated with lower 30-day risk-adjusted mortality, without a significant increase in length of hospitalization. These findings support the use of TEE as routine practice in open cardiac valve repair or replacement surgery.

Project description:BackgroundCardiac valvular disease affects millions of people worldwide and is a major cause of morbidity and mortality. Female patients have been shown to experience inferior clinical outcomes after nonvalvular cardiac surgery, but recent data are limited regarding open valve surgical cohorts. The primary objective of our study was to assess whether female sex is associated with increased in-hospital mortality after open cardiac valve operations.MethodsUtilizing the Healthcare Cost and Utilization Project (HCUP) State Inpatient Databases (SID), we conducted a retrospective cohort study of patients who underwent open cardiac valve surgery from 2007 to 2018 in Washington, Maryland, Kentucky, and Florida; from 2007 to 2011 in California; and from 2007 to 2016 in New York. The primary objective of this study was to estimate the confounder-adjusted association between sex and in-hospital mortality (as recorded and coded by SID HCUP) after open cardiac valve surgery. We used multilevel multivariable models to account for potential confounders, including intrahospital practice patterns.ResultsA total of 272,954 patients (108,443 women; 39.73% of sample population with mean age of 67.6 ± 14.3 years) were included in our analysis. The overall mortality rates were 3.8% for male patients and 5.1% for female patients. The confounder-adjusted odds ratio (OR) for in-hospital mortality for female patients compared to male patients was 1.41 (95% confidence interval [CI], 1.35-1.47; P < .001). When stratifying by surgical type, female patients were also at increased odds of in-hospital mortality ( P < .001) in populations undergoing aortic valve replacement (adjusted OR [aOR], 1.38; 95% CI, 1.25-1.52); multiple valve surgery (aOR, 1.38; 95% CI, 1.22-1.57); mitral valve replacement (aOR, 1.22; 95% CI, 1.12 - 1.34); and valve surgery with coronary artery bypass grafting (aOR, 1.64; 95% CI, 1.54 - 1.74; all P < .001). Female patients did not have increased odds of in-hospital mortality in populations undergoing mitral valve repair (aOR, 1.26; 95% CI, 0.98 - 1.64; P = .075); aortic valve repair (aOR, 0.87; 95% CI, 0.67 - 1.14; P = .32); or any other single valve repair (aOR, 1.10; 95% CI, 0.82 - 1.46; P = .53).ConclusionsWe found an association between female patients and increased confounder-adjusted odds of in-hospital mortality after open cardiac valve surgery. More research is needed to better understand and categorize these important outcome differences. Future research should include observational analysis containing granular and complete patient- and surgery-specific data.

Project description:BackgroundThe recent use of artificial intelligence (AI) in medical research is noteworthy. However, most research has focused on medical imaging. Although the importance of laboratory tests in the clinical field is acknowledged by clinicians, they are undervalued in medical AI research. Our study aims to develop an early prediction AI model for pneumonia mortality, primarily using laboratory test results.Materials and methodsWe developed a mortality prediction model using initial laboratory results and basic clinical information of patients with pneumonia. Several machine learning (ML) models and a deep learning method-multilayer perceptron (MLP)-were selected for model development. The area under the receiver operating characteristic curve (AUROC) and F1-score were optimized to improve model performance. In addition, an ensemble model was developed by blending several models to improve the prediction performance. We used 80,940 data instances for model development.ResultsAmong the ML models, XGBoost exhibited the best performance (AUROC = 0.8989, accuracy = 0.88, F1-score = 0.80). MLP achieved an AUROC of 0.8498, accuracy of 0.86, and F1-score of 0.75. The performance of the ensemble model was the best among the developed models, with an AUROC of 0.9006, accuracy of 0.90, and F1-score of 0.81. Several laboratory tests were conducted to identify risk factors that affect pneumonia mortality using the "Feature importance" technique and SHapley Additive exPlanations. We identified several laboratory results, including systolic blood pressure, serum glucose level, age, aspartate aminotransferase-to-alanine aminotransferase ratio, and monocyte-to-lymphocyte ratio, as significant predictors of mortality in patients with pneumonia.ConclusionsOur study demonstrates that the ensemble model, incorporating XGBoost, CatBoost, and LGBM techniques, outperforms individual ML and deep learning models in predicting pneumonia mortality. Our findings emphasize the importance of integrating AI techniques to leverage laboratory test data effectively, offering a promising direction for advancing AI applications in medical research and clinical decision-making.

Dataset Information

Ensemble machine learning prediction and variable importance analysis of 5-year mortality after cardiac valve and CABG operations.

Publications

Ensemble machine learning prediction and variable importance analysis of 5-year mortality after cardiac valve and CABG operations.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets