Decision Support Software for Forecasting Patient’s Length of Stay

Beyond the Bedside: Machine Learning-Guided Length of Stay (LOS) Prediction for Cardiac Patients in Tertiary Care

Efficient management of hospital resources is essential for providing high-quality healthcare while ensuring sustainability. Length of stay (LOS), measuring the duration from admission to discharge, directly impacts patient outcomes and resource utilization. Accurate LOS prediction offers numerous benefits, including reducing re-admissions, ensuring appropriate staffing, and facilitating informed discharge planning. While conventional methods rely on statistical models and clinical expertise, recent advances in machine learning (ML) present promising avenues for enhancing LOS prediction. This research focuses on developing an ML-based LOS prediction model trained on a comprehensive real-world dataset and discussing the important factors towards practical deployment of trained ML models in clinical settings. This research involves the development of a comprehensive adult cardiac patient dataset (SaudiCardioStay (SCS)) from the King Faisal Specialist Hospital & Research Centre (KFSH&RC) hospital in Saudi Arabia, comprising 4930 patient encounters for 3611 unique patients collected from 2019 to 2022 (excluding 2020). A diverse range of classical ML models (i.e., Random Forest (RF), Extreme Gradient Boosting (XGBoost), Light Gradient Boosting Machine (LGBM), artificial neural networks (ANNs), Average Voting Regression (AvgVotReg)) are implemented for the SCS dataset to explore the potential of existing ML models in LOS prediction. In addition, this study introduces a novel approach for LOS prediction by incorporating a dedicated LOS classifier within a sophisticated ensemble methodology (i.e., Two-Level Sequential Cascade Generalization (2LSCG), Three-Level Sequential Cascade Generalization (3LSCG), Parallel Cascade Generalization (PCG)), aiming to enhance prediction accuracy and capture nuanced patterns in healthcare data. The experimental results indicate the best mean absolute error (MAE) of 0.1700 for the 3LSCG model. Relatively comparable performance was observed for the AvgVotReg model, with a MAE of 0.1703. In the end, a detailed analysis of the practical implications, limitations, and recommendations concerning the deployment of ML approaches in actual clinical settings is presented.

Enhancing length of stay prediction by learning similarity-aware representations for hospitalized patients

This paper focuses on predicting the length of stay for patients on the first day of admission and propose a predictive model named DGLoS. In order to capture the influence of various complex factors on the length of stay as well as the dependencies among various factors, DGLoS uses a deep neural network to model both the patient information and diagnostic information. Targeting at different attribution types, we utilize different coding methods to convert raw data to the input features. Besides, we find that similar patients have closer lengths of stay. Therefore, we further design a module based on graph representation learning to generate patients' similarity-aware representations, capturing the similarity between patients and therefore enhancing predictions. These similarity-aware representations are incorporated into the output of the deep neural network to jointly perform the prediction. We have conducted comprehensive experiments on a real-world hospitalization dataset. The performance comparison shows that our proposed DGLoS model improves predictive performance and the significance test demonstrates the improvement is significant. The ablation study verifies the effectiveness of each of the proposed components and the hyper-parameter investigation shows the robustness of the proposed model.

A deep attention model to forecast the Length Of Stay and the in-hospital mortality right on admission from ICD codes and demographic data

Leveraging the Electronic Health Records (EHR) longitudinal data to produce actionable clinical insights has always been a critical issue for recent studies. Non-forecasted extended hospitalizations account for a disproportionate amount of resource use, the mediocre quality of inpatient care, and avoidable fatalities. The capability to predict the Length of Stay (LoS) and mortality in the early stages of the admission provides opportunities to improve care and prevent many preventable losses. Forecasting the in-hospital mortality is important in providing clinicians with enough insights to make decisions and hospitals to allocate resources, hence predicting the LoS and mortality within the first day of admission is a difficult but a paramount endeavor. The biggest challenge is that few data are available by this time, thus the prediction has to bring in the previous admissions history and free text diagnosis that are recorded immediately on admission. We propose a model that uses the multi-modal EHR structured medical codes and key demographic information to classify the LoS in 3 classes; Short Los (LoS⩽10 days), Medium LoS (10<LoS⩽30 days) and Long LoS (LoS>30 days) as well as mortality as a binary classification of a patient's death during current admission. The prediction has to use data available only within 24 h of admission. The key predictors include previous ICD9 diagnosis codes, ICD9 procedures, key demographic data, and free text diagnosis of the current admission recorded right on admission. We propose a Hierarchical Attention Network (HAN-LoS and HAN-Mor) model and train it to a dataset of over 45321 admissions recorded in the de-identified MIMIC-III dataset. For improved prediction, our attention mechanisms can focus on the most influential past admissions and most influential codes in these admissions. For fair performance evaluation, we implemented and compared the HAN model with previous approaches. With dataset balancing techniques HAN-LoS achieved an AUROC of over 0.82 and a Micro-F1 score of 0.24 and HAN-Mor achieved AUC-ROC of 0.87 hence outperforming the existing baselines that use structured medical codes as well as clinical time series for LoS and Mortality forecasting. By predicting mortality and LoS using the same model, we show that with little tuning the proposed model can be used for other clinical predictive tasks like phenotyping, decompensation,re-admission prediction, and survival analysis.

Machine Learning in the Prediction of Medical Inpatient Length of Stay

BACKGROUND

Length of stay (LOS) estimates are important for patients, doctors and hospital administrators. However, making accurate estimates of LOS can be difficult for medical patients.

AIMS

This review was conducted with the aim of identifying and assessing previous studies on the application of machine learning to the prediction of total hospital inpatient LOS for medical patients.

METHODS

A review of machine learning in the prediction of total hospital LOS for medical inpatients was conducted using the databases PubMed, EMBASE and Web of Science.

RESULTS

Of the 673 publications returned by the initial search, 21 articles met inclusion criteria. Of these articles the most commonly represented medical specialty was cardiology. Studies were also identified that had specifically evaluated machine learning LOS prediction in patients with diabetes and tuberculosis. The performance of the machine learning models in the identified studies varied significantly depending on factors including differing input datasets and different LOS thresholds and outcome metrics. Common methodological shortcomings included a lack of reporting of patient demographics and lack of reporting of clinical details of included patients.

CONCLUSIONS

The variable performance reported by the studies identified in this review supports the need for further research of the utility of machine learning in the prediction of total inpatient LOS in medical patients. Future studies should follow and report a more standardised methodology to better assess performance and to allow replication and validation. In particular, prospective validation studies and studies assessing the clinical impact of such machine learning models would be beneficial. This article is protected by copyright. All rights reserved.

Application of Machine Learning Techniques to Analyze Patient Returns to the Emergency Department

The study of the quality of hospital emergency services is based on analyzing a set of indicators such as the average time of first medical attention, the average time spent in the emergency department, degree of completion of the medical report and others. In this paper, an analysis is presented of one of the quality indicators: the rate of return of patients to the emergency service less than 72 h from their discharge. The objective of the analysis was to know the variables that influence the rate of return and which prediction model is the best. In order to do this, the data of the activity of the emergency service of a hospital of a reference population of 290,000 inhabitants were analyzed, and prediction models were created for the binary objective variable (rate of return to emergencies) using the logistic regression techniques, neural networks, random forest, gradient boosting and assembly models. Each of the models was analyzed and the result shows that the best model is achieved through a neural network with activation function tanh, algorithm levmar and three nodes in the hidden layer. This model obtains the lowest mean squared error (MSE) and the best area under the curve (AUC) with respect to the rest of the models used.