Interpretable Predictive Models for Knowledge Discovery from Home-Care Electronic Health Records

This article is part of the Focus Theme of Methods of Information in Medicine on "Big Data and Analytics in Healthcare". Hospital readmissions raise healthcare costs and cause significant distress to providers and patients. It is, therefore, of great interest to healthcare organizations to predict what patients are at risk to be readmitted to their hospitals. However, current logistic regression based risk prediction models have limited prediction power when applied to hospital administrative data. Meanwhile, although decision trees and random forests have been applied, they tend to be too complex to understand among the hospital practitioners. Explore the use of conditional logistic regression to increase the prediction accuracy. We analyzed an HCUP statewide inpatient discharge record dataset, which includes patient demographics, clinical and care utilization data from California. We extracted records of heart failure Medicare beneficiaries who had inpatient experience during an 11-month period. We corrected the data imbalance issue with under-sampling. In our study, we first applied standard logistic regression and decision tree to obtain influential variables and derive practically meaning decision rules. We then stratified the original data set accordingly and applied logistic regression on each data stratum. We further explored the effect of interacting variables in the logistic regression modeling. We conducted cross validation to assess the overall prediction performance of conditional logistic regression (CLR) and compared it with standard classification models. The developed CLR models outperformed several standard classification models (e.g., straightforward logistic regression, stepwise logistic regression, random forest, support vector machine). For example, the best CLR model improved the classification accuracy by nearly 20% over the straightforward logistic regression model. Furthermore, the developed CLR models tend to achieve better sensitivity of more than 10% over the standard classification models, which can be translated to correct labeling of additional 400 - 500 readmissions for heart failure patients in the state of California over a year. Lastly, several key predictor identified from the HCUP data include the disposition location from discharge, the number of chronic conditions, and the number of acute procedures. It would be beneficial to apply simple decision rules obtained from the decision tree in an ad-hoc manner to guide the cohort stratification. It could be potentially beneficial to explore the effect of pairwise interactions between influential predictors when building the logistic regression models for different data strata. Judicious use of the ad-hoc CLR models developed offers insights into future development of prediction models for hospital readmissions, which can lead to better intuition in identifying high-risk patients and developing effective post-discharge care strategies. Lastly, this paper is expected to raise the awareness of collecting data on additional markers and developing necessary database infrastructure for larger-scale exploratory studies on readmission risk prediction.

Environmental Factors Prediction in Preterm Birth Using Comparison between Logistic Regression and Decision Tree Methods:  An Exploratory Analysis

Newer analytic approaches for developing predictive models provide a method of creating decision support to translate findings into practice. The aim of this study was to develop and validate a clinically interpretable predictive model for 12-month mortality risk among community-dwelling older adults. This is done by using routinely collected nursing assessment data to aide homecare nurses in identifying older adults who are at risk for decline, providing an opportunity to develop care plans that support patient and family goals for care. A retrospective secondary analysis of Medicare and Medicaid data of 635,590 Outcome and Assessment Information Set (OASIS-C) start-of-care assessments from January 1, 2012, to December 31, 2012, was linked to the Master Beneficiary Summary File (2012-2013) for date of death. The decision tree was benchmarked against gold standards for predictive modeling, logistic regression, and artificial neural network (ANN). The models underwent k-fold cross-validation and were compared using area under the curve (AUC) and other data science metrics, including Matthews correlation coefficient (MCC). Decision tree variables associated with 12-month mortality risk included OASIS items: age, (M1034) overall status, (M1800-M1890) activities of daily living total score, cancer, frailty, (M1410) oxygen, and (M2020) oral medication management. The final models had good discrimination: decision tree, AUC = .71, 95% confidence interval (CI) [.705, .712], sensitivity = .73, specificity = .58, MCC = .31; ANN, AUC = .74, 95% CI [.74, .74], sensitivity = .68, specificity = .68, MCC = .35; and logistic regression, AUC = .74, 95% CI [.735, .742], sensitivity = .64, specificity = .70, MCC = .35. The AUC and 95% CI for the decision tree are slightly less accurate than logistic regression and ANN; however, the decision tree was more accurate in detecting mortality. The OASIS data set was useful to predict 12-month mortality risk. The decision tree is an interpretable predictive model developed from routinely collected nursing data that may be incorporated into a decision support tool to identify older adults at risk for death.

The growth and development of children under five (toddlers) affect their health conditions. Each region uniquely identifies the main factors influencing the toddler's health condition. The status of toddlers is generally categorized into two classes, namely normal and abnormal, so it is often found that the condition of toddler status is in the form of multi-response variables. Combining the two binary classes' response variables will form a multiclass response variable requiring different model development techniques and performance measurements. This study aims to determine the main factors that affect toddlers' health conditions in Malang, Indonesia, build multiclass logistic regression and decision tree classification models, and measure the model's performance. The Chi-square test selected predictor features as the input of multiclass logistic regression and decision tree models. From the feature selection, four main factors influence the status of toddlers' health conditions in Malang: the mother's history of diabetes before pregnancy, the father's blood pressure, psychological condition, and drinking water quality. The decision tree model performs better than the logistic regression model on the various performance measures used.

This study examined records of 2566 consecutive COVID-19 patients at five Massachusetts hospitals and sought to predict level-of-care requirements based on clinical and laboratory data. Several classification methods were applied and compared against standard pneumonia severity scores. The need for hospitalization, ICU care, and mechanical ventilation were predicted with a validation accuracy of 88%, 87%, and 86%, respectively. Pneumonia severity scores achieve respective accuracies of 73% and 74% for ICU care and ventilation. When predictions are limited to patients with more complex disease, the accuracy of the ICU and ventilation prediction models achieved accuracy of 83% and 82%, respectively. Vital signs, age, BMI, dyspnea, and comorbidities were the most important predictors of hospitalization. Opacities on chest imaging, age, admission vital signs and symptoms, male gender, admission laboratory results, and diabetes were the most important risk factors for ICU admission and mechanical ventilation. The factors identified collectively form a signature of the novel COVID-19 disease.

A new hybrid classification algorithm for customer churn prediction based on logistic regression and decision trees

Aim: The aim of the work is to evaluate the accuracy and precision in predicting heart disease using Support Vector Machine (SVM), Random forest (RF), Logistic Regression (LR), Decision Tree (DT) Classification algorithms. Materials and Methods: Classification algorithm is appealed on a heart dataset which consists of 180 records. A framework for heart disease prediction in the medical sector comparing Random forest, Logistic Regression, Decision Tree and SVM classifiers has been proposed and developed. The sample size was calculated as 55 in each group using G power 80%. Sample size was calculated using clincalc analysis, with alpha and beta values 0.05 and 0.5, 95% confidence, pretest power 80% and enrolment ratio 1. Results: The Novel Random Forest Algorithm (92.13%), Support Vector Machine (62.51%), Logistic Regression (84.89%), Decision Tree (86.25%) classifiers produce respectively. SVM, RF exists a statistically significant difference between the two groups (p=0.001,p=.004;p<0.05).LR, RF exists a statistically insignificant difference between the two groups (p=.103, P=.080;p>0.05) both with confidence interval 95%. Hence Random forest is better than SVM, RF, DT classifiers. Conclusion: The results show that the performance of RF is better when compared with SVM, LR and DT in terms of both precision and accuracy.

ObjectiveThis study aimed to systematically identify risk factors for urinary catheter-related hematuria in patients with acute myocardial infarction (AMI). By integrating logistic regression and decision tree models, we sought to develop actionable strategies for risk stratification and complication prevention.MethodsA retrospective analysis of 209 AMI patients was conducted to evaluate predictors of hematuria, including demographics, coagulation indices (INR, platelets), and procedural variables. Logistic regression and decision tree (CART algorithm) models were employed to identify risk factors and their interactions. Model performance was assessed using ROC-AUC, sensitivity, and specificity.ResultsThe incidence of catheter-related hematuria was 32.5%. Both models identified persistent agitation during catheter indwelling and PLT ≤ 246 as common predictors. The logistic regression model specifically identified Gender (OR = 0.202), patient awareness of catheter purpose and precautions (OR = 0.470), and emergency catheter placement (OR = 2.257) as significant factors. The decision tree model uniquely identified INR > 0.955 and repeated complaints of urethral pain as predictors.ConclusionHematuria in AMI patients results from coagulation dysfunction, procedural trauma, and behavioral factors. The combined use of logistic regression and decision trees enhances risk stratification. Clinical strategies should prioritize gentle catheterization, dynamic coagulation monitoring, and patient education to reduce complications.

BackgroundClinical determination of patients at high risk of poor surgical outcomes is complex and may be supported by clinical tools to summarize the patient’s own personalized electronic health record (EHR) history and vitals data through predictive risk models. Since prior models were not readily available for EHR-integration, our objective was to develop and validate a risk stratification tool, named the Assessment of Geriatric Emergency Surgery (AGES) score, predicting risk of 30-day major postoperative complications in geriatric patients under consideration for urgent and emergency surgery using pre-surgical existing electronic health record (EHR) data.MethodsPatients 65-years and older undergoing urgent or emergency non-cardiac surgery within 21 hospitals 2017–2021 were used to develop the model (randomly split: 80% training, 20% test). The primary outcome was a 30-day composite outcome including several postoperative major complications and mortality; secondary outcomes included common individual complications included in the primary composite outcome (sepsis, progressive renal insufficiency or renal failure, and mortality). Patients’ EHR-based clinical history, vital signs, labs, and demographics were included in logistic regression, LASSO, decision tree, Random Forest, and XGBoost models. Area under the receiver operating characteristics curve (AUCROC) was used to compare model performance.ResultsOverall, 66,262 patients (median [IQR] age 78 [(70.9–84.0], female 53.9%, White race 68.5%) received urgent or emergency non-cardiac surgery (25.7% orthopedic cases, 21.9% general surgery cases). AUCROC ranged from 0.655 (Decision Tree) – 0.804 (XGBoost) for the primary composite outcome. XGBoost AUCROC was 0.823, 0.781, and 0.839 in predicting outcomes of sepsis, progressive renal insufficiency or renal failure, and mortality, respectively.ConclusionsWe developed a model to accurately predict major postoperative complications in geriatric patients undergoing urgent or emergency surgery using the patient’s own existing EHR data. EHR implementation of this model could efficiently support clinicians’ surgical risk assessment and perioperative decision-making discussions in this vulnerable patient population.

Background Malaria remains a major cause of illness and death among children under five in Nigeria, despite efforts to control transmission. Accurate and reliable prediction of malaria outbreaks is crucial for health authorities to take timely measures. This study aims to identify the most robust machine learning classification algorithms for predicting the status of malaria in children under five (0-59 months). Methods The 2021 Nigeria Malaria Indicator Survey (NMIS) included 10,655 children under five who were tested for malaria using the Rapid Diagnostic Test (RDT). Various machine learning models were explored, including Decision Trees, K-Nearest Neighbor, Naïve Bayes, Random Forest, Support Vector Machines, and Survey Logistic Regression, and their performance was evaluated through metrics such as accuracy, AUC, balanced accuracy, F1-Score, negative predictive value, precision, sensitivity, and specificity. Results Random Forest (RF) is the most robust and balanced classification model due to its superior accuracy (79%), precision (77%), recall (62%), F1-score (69%), and AUC (80%). Support Vector Machine (SVM) also demonstrated strong performance, particularly in accuracy (74%) and AUC (80%). Survey Logistic Regression (SLR) and Decision Tree (DT) offered moderate results but fell short compared to RF and SVM, indicating the need for further optimization. Naive Bayes (NB) and K-Nearest Neighbors (KNN) had limitations, making them less reliable for this task. Conclusion In conclusion, the study reveals that RF and SVM are the best classification models for predicting malaria status in children under five years old. RF is reliable and balanced, while SVM is preferred for recall. SLR and DT show potential but require optimization. NB and KNN have significant performance gaps, making them less suitable. These findings will help policymakers and malaria intervention programs address key factors, enabling more targeted public health interventions to reduce the malaria burden on young children and improve the well-being of vulnerable populations in Nigeria.

The failure of projects remains one of the significant issues in all industries because projects have become more complex, uncertain, and risky. The conventional methods of project risk management are based usually on qualitative measures and fixed risk lists, and thus such methods do not allow them to make early and precise predictions of failure. This study aims at developing an AI-driven risk forecasting predictive model that would be used to mitigate risks of project failure using machine learning tools on past project information. The model development and evaluation was done using a structured dataset of 1, 200 successful projects and several risk-based characteristics (such as cost variance, schedule deviation, team experience, and requirement volatility). Four machine learning algorithms namely, Logistic Regression, Decision Tree, Random Forest, and Support Vector Machine were put to practice and compared. The outcomes of the experiments are able to show that AI-based models can increase predictive accuracy and the early identification of risks by great margins. The Random Forest model demonstrated the best performance with the highest accuracy of 88.9 and the precipitation of 87.5, recall of 86.2 and F1-score of 0.87. Another aspect that the Support Vector Machine performed well with an accuracy of 86.7 was compared to Logistic regression and decision tree whose accuracy of 81.4 and 83.6. The analysis of feature importance has determined cost variance, schedule delay as the most highly influential factors of project failure. The results prove that the AI-based risk prediction models can offer a powerful, data-driven strategy of proactive mitigation of project failures, which can help to make informed decisions and enhance project success rates.

Credit is becoming one of the most important incomes of banking. Past studies indicate that the credit risk scoring model has been better for Logistic Regression and Neural Network. The purpose of this paper is to conduct a comparative study on the accuracy of classification models and reduce the credit risk. In this paper, we use data mining of enterprise software to construct four classification models, namely, decision tree, logistic regression, neural network and support vector machine, for credit scoring in banking. We conduct a systematic comparison and analysis on the accuracy of 17 classification models for credit scoring in banking. The contribution of this paper is that we use different classification methods to construct classification models and compare classification models accuracy, and the evidence demonstrates that the support vector machine models have higher accuracy rates and therefore outperform past classification methods in the context of credit scoring in banking.

This study aimed to develop and validate predictive models using electronic health records (EHR) data to determine whether hospitalized COVID-19-positive patients would be admitted to alternative medical care or discharged home. We conducted a retrospective cohort study using deidentified data from the University of Florida Health Integrated Data Repository. The study included 1,578 adult patients (≥18 years) who tested positive for COVID-19 while hospitalized, comprising 960 (60.8%) female patients with a mean (SD) age of 51.86 (18.49) years and 618 (39.2%) male patients with a mean (SD) age of 54.35 (18.48) years. Machine learning (ML) model training involved cross-validation to assess their performance in predicting patient disposition. We developed and validated six supervised ML-based prediction models (logistic regression, Gaussian Naïve Bayes, k-nearest neighbors, decision trees, random forest, and support vector machine classifier) to predict patient discharge status. The models were evaluated based on the area under the receiver operating characteristic curve (ROC-AUC), precision, accuracy, F1 score, and Brier score. The random forest classifier exhibited the highest performance, achieving an accuracy of 0.84 and an AUC of 0.72. Logistic regression (accuracy: 0.85, AUC: 0.71), k-nearest neighbor (accuracy: 0.84, AUC: 0.63), decision tree (accuracy: 0.84, AUC: 0.61), Gaussian Naïve Bayes (accuracy: 0.84, AUC: 0.66), and support vector machine classifier (accuracy: 0.84, AUC: 0.67) also demonstrated valuable predictive capabilities. This study's findings are crucial for efficiently allocating healthcare resources during pandemics like COVID-19. By harnessing ML techniques and EHR data, we can create predictive tools to identify patients at greater risk of severe symptoms based on their medical histories. The models developed here serve as a foundation for expanding the toolkit available to healthcare professionals and organizations. Additionally, explainable ML methods, such as Shapley Additive Explanations, aid in uncovering underlying data features that inform healthcare decision-making processes.

In recent studies, researchers have focused on using various modalities to recognize emotions for different applications. A major challenge is identifying emotions correctly with only electrocardiograms (ECG) as the modality. The main objective is to reduce costs by using single-modality ECG signals to predict human emotional states. This paper presents an emotion recognition approach utilizing the heart rate variability features obtained from ECG with feature selection techniques (exhaustive feature selection (EFS) and Pearson’s correlation) to train the classification models. Seven machine learning (ML) models: multi-layer perceptrons (MLP), Support Vector Machine (SVM), Decision Tree (DT), Gradient Boosting Decision Tree (GBDT), Logistic Regression, Adaboost and Extra Tree classifier are used to classify emotional state. Two public datasets, DREAMER and SWELL are used for evaluation. The results show that no particular ML works best for all data. For DREAMER with EFS, the best models to predict valence, arousal, and dominance are Extra Tree (74.6%), MLP and DT (74.6%), and GBDT and DT (69.8%), respectively. Extra tree with Pearson’s correlation are the best method for the ECG SWELL dataset and provide 100% accuracy. The usage of Extra tree classifier and feature selection technique contributes to the improvement of the model accuracy. Moreover, the Friedman test proved that ET is as good as other classification models for predicting human emotional state and ranks highest. Doi: 10.28991/ESJ-2023-07-01-011 Full Text: PDF

Purpose This study aims to develop and validate a machine learning-based risk prediction model for 1-year cardiac mortality in acute myocardial infarction (AMI) patients, utilizing a multidimensional dataset that integrates traditional risk factors, environmental determinants, and angiographic findings. Methods Data was sourced from a multicenter prospective cohort study (NCT 03297164) along with electronic medical records from a large tertiary hospital, covering the period from January 1, 2017, to July 28, 2022. The patient population from the tertiary hospital constituted the derivation cohort, while patients from other participating centers served as the external validation cohort. Six machine learning algorithms — logistic regression, XGBoost, LightGBM (LGBM), Gradient Boosting, random forest, and decision tree — were applied, with feature selection performed using the SelectFromModel method. The best-performing model underwent both internal and external validation, with its predictive performance compared to the GRACE score. Result The derivation cohort was divided into training and testing cohorts. In the training cohort, six machine learning algorithms were employed to screen variables and construct models using five-fold cross-validation (Figure 2A). All models achieved an area under the receiver operating characteristic curve (AUC) > 0.8 (Figure 2B). Delong's test revealed no significant differences in AUC among the algorithms. Considering other evaluation metrics and brier score, the LGBM (AUC 0.85) model was selected, incorporating the following variables: age, body mass index, smoking history, temperature, PM2.5, Gensini score, weighted diffuse lesion value, left ventricular end-diastolic diameter, ejection fraction, serum creatinine, cardiac troponin I, glucose, low-density lipoprotein cholesterol, B-type natriuretic peptide, and D-dimmer. External validation of the LGBM model yielded an AUC of 0.80, while the GRACE score achieved AUCs of 0.73 and 0.74 in internal testing and external validation, respectively (Figure 2C). Predicted LGBM scores in the training cohort were stratified into ten groups by decile. Patients were categorized into low-, intermediate-, and high-risk groups based on the actual incidence of cardiac death (Figure 2D). Using these cut-off values, risk stratification was applied to both the internal testing cohort and the external validation cohort, and the actual event rates were calculated (Figure E). Logistic and Cox regression analyses demonstrated significant differences between risk groups (P < 0.001, Figures F-I), with the LGBM-based stratification outperforming that based on the GRACE score. Conclusion Based on recent large-scale multicenter cohort data and incorporating environmental factors along with angiographic findings, this study developed and validated a novel prediction model for 1-year cardiac death in AMI patients, which demonstrated superior performance compared to the GRACE score.Flow chart Main Findings