An Optimized Machine Learning Model Accurately Predicts In-Hospital Outcomes at Admission to a Cardiac Unit

Abstract

Risk stratification at the time of hospital admission is of paramount significance in triaging the patients and providing timely care. In the present study, we aim at predicting multiple clinical outcomes using the data recorded during admission to a cardiac care unit via an optimized machine learning method. This study involves a total of 11,498 patients admitted to a cardiac care unit over two years. Patient demographics, admission type (emergency or outpatient), patient history, lab tests, and comorbidities were used to predict various outcomes. We employed a fully connected neural network architecture and optimized the models for various subsets of input features. Using 10-fold cross-validation, our optimized machine learning model predicted mortality with a mean area under the receiver operating characteristic curve (AUC) of 0.967 (95% confidence interval (CI): 0.963–0.972), heart failure AUC of 0.838 (CI: 0.825–0.851), ST-segment elevation myocardial infarction AUC of 0.832 (CI: 0.821–0.842), pulmonary embolism AUC of 0.802 (CI: 0.764–0.84), and estimated the duration of stay (DOS) with a mean absolute error of 2.543 days (CI: 2.499–2.586) of data with a mean and median DOS of 6.35 and 5.0 days, respectively. Further, we objectively quantified the importance of each feature and its correlation with the clinical assessment of the corresponding outcome. The proposed method accurately predicts various cardiac outcomes and can be used as a clinical decision support system to provide timely care and optimize hospital resources.

A Novel Nomogram for prolonged length of stay in older patients with chronic heart failure

“…[24] The admission type has the highest importance and studies have shown that emergency admission was associated with longer LOS in patients in cardiac units. [25] Older patients with Chronic HF are often admitted to the hospital via the emergency department due to poorer disease management leading to exacerbations such as respiratory failure and cardiac arrhythmias. [26] Moreover, they are at greater risk of prolonged LOS due to their life-threatening conditions and have longer clinical recovery periods than patients admitted through outpatient clinics.…”

Predicting Patient Length of Stay in Australian Emergency Departments Using Data Mining

“…Future work can utilize machine learning and deep learning algorithms and frameworks for CDS and predicting ED LOS. This will help to determine predictive accuracy of more advanced models when compared to DM [74][75][76].…”

A comprehensive review on efficient artificial intelligence models for classification of abnormal cardiac rhythms using electrocardiograms

“…The advent of artificial intelligence (AI) and advanced computational techniques has substantially revolutionized healthcare diagnostics [13] . Over the past decade, novel machine learning (ML) and deep learning (DL) modalities have been investigated that utilize ECG data and other vital sign measures to predict cardiovascular status in patients and assist physicians in diagnosis [14] , [15] , [16] , [17] , [18] , [19] , [20] , [21] , [22] , [23] , [24] , [25] , [26] . In particular, there has been a great emphasis on the identification and classification of cardiac rhythm based on the ECG data to aid in prediction of impending arrhythmias and enable timely application of therapy.…”

“…In particular, there has been a great emphasis on the identification and classification of cardiac rhythm based on the ECG data to aid in prediction of impending arrhythmias and enable timely application of therapy. AI based methods incorporating neural networks, Bayesian networks, fuzzy logic systems, as well as machine learning models using linear or logistic regression, decision trees, k-nearest neighbors, random forest, or support vector machines, have been shown to accurately predict cardiovascular outcomes in patients [27] , [28] , [14] , [16] . Furthermore, the 12-lead ECG based classification algorithms have demonstrated a great precision in early diagnosis of cardiac arrhythmias, highlighting the potential of AI based approaches in supporting cardiovascular patient management [29] , [30] , [31] .…”