PMID- 36312291 OWN - NLM STAT- PubMed-not-MEDLINE LR - 20221102 IS - 2297-055X (Print) IS - 2297-055X (Electronic) IS - 2297-055X (Linking) VI - 9 DP - 2022 TI - 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. PG - 994359 LID - 10.3389/fcvm.2022.994359 [doi] LID - 994359 AB - 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 (SpO(2)), 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. CI - Copyright (c) 2022 Peng, Huang, Liu, Deng, Sun, Tang, Chen, Cao, Wang, Duan, Luo and Peng. FAU - Peng, Shengxian AU - Peng S AD - Scientific Research Department, First People's Hospital of Zigong City, Zigong, China. FAU - Huang, Jian AU - Huang J AD - Graduate School, Guangxi University of Chinese Medicine, Nanning, China. FAU - Liu, Xiaozhu AU - Liu X AD - Department of Cardiology, The Second Affiliated Hospital of Chongqing Medical University, Chongqing, China. FAU - Deng, Jiewen AU - Deng J AD - Department of Neurosurgery, The First Affiliated Hospital of Chongqing Medical University, Chongqing, China. FAU - Sun, Chenyu AU - Sun C AD - AMITA Health Saint Joseph Hospital Chicago, Chicago, IL, United States. FAU - Tang, Juan AU - Tang J AD - Scientific Research Department, First People's Hospital of Zigong City, Zigong, China. FAU - Chen, Huaqiao AU - Chen H AD - Department of Cardiology, The First Affiliated Hospital of Chongqing Medical University, Chongqing, China. FAU - Cao, Wenzhai AU - Cao W AD - Department of Cardiology, First People's Hospital of Zigong City, Zigong, China. FAU - Wang, Wei AU - Wang W AD - Department of Cardiology, First People's Hospital of Zigong City, Zigong, China. AD - Information Department, First People's Hospital of Zigong City, Zigong, China. FAU - Duan, Xiangjie AU - Duan X AD - Department of Infectious Diseases, The First People's Hospital of Changde City, Changde, China. FAU - Luo, Xianglin AU - Luo X AD - Information Department, First People's Hospital of Zigong City, Zigong, China. FAU - Peng, Shuang AU - Peng S AD - General Affairs Section, The People's Hospital of Tongnan District, Chongqing, China. LA - eng PT - Journal Article DEP - 20221012 PL - Switzerland TA - Front Cardiovasc Med JT - Frontiers in cardiovascular medicine JID - 101653388 PMC - PMC9597462 OTO - NOTNLM OT - MIMIC-IV OT - heart failure OT - hypertension OT - interpretable machine learning OT - neural networks COIS- The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest. EDAT- 2022/11/01 06:00 MHDA- 2022/11/01 06:01 PMCR- 2022/01/01 CRDT- 2022/10/31 04:37 PHST- 2022/07/14 00:00 [received] PHST- 2022/09/15 00:00 [accepted] PHST- 2022/10/31 04:37 [entrez] PHST- 2022/11/01 06:00 [pubmed] PHST- 2022/11/01 06:01 [medline] PHST- 2022/01/01 00:00 [pmc-release] AID - 10.3389/fcvm.2022.994359 [doi] PST - epublish SO - Front Cardiovasc Med. 2022 Oct 12;9:994359. doi: 10.3389/fcvm.2022.994359. eCollection 2022.