Machine learning model for healthcare investments predicting the length of stay in a hospital & mortality rate

The demand for healthcare workers and infrastructure from an alarmingly growing patient population may contribute to the increased Length of Stay (LOS) in Hospital and Mortality rate. The shortage of doctors, nurses, and hospital beds may be blamed for this increase. As Constant patient monitoring is essential and the better hospital management and administration are necessary, therefore this research aimed foremost, to develop a machine learning model to predict long-term outcomes like Length of Stay (LOS), mortality rate of a patient admitted into the hospital. We used Machine Learning (ML) in the National Hospital Care Research Database (NHCRD) to create minimum feature-based predictive modeling with adequate performance. Unlike other approaches, ours requires the patient's profile, tests reports at the time of admission and treatment history to accurately predict outcomes like the length of stay and mortality rate, making our technique novel with 98% accuracy, 98% precision, 95% AUROC Score, 94% F1 Score, 0.97 Recall, 0.95 Train Accuracy, and 0.90 Test Accuracy with the Support Vector Machine Algorithm. The ratio of training data to testing data was divided in the ratio 8:2 then the Machine Learning methods were applied. Descriptive statistical graphs, feature significance, precision-recall curve, accuracy plots, and Area Under the Curve (AUC), Accuracy, Precision, Recall, F1-Score, Mean Squared Error, Mean Absolute Error and Root Mean Squared Error were used to evaluate different machine learning methods like Random Forests (RF), Logistic Regression (LR), Gradient Boosting (GB), Decision Tree (DT), Naive Bayes (NB), Artificial Neural Network (ANN), and Ensemble Learning Techniques (EL), etc. Adopting the proposed framework, which considers the imbalanced dataset for classification-based methods based on electronic healthcare records, may allow us to apply Machine Learning to forecast patient length of stay and mortality rate in the hospital's clinical information system. This Prediction Model will help hospitals and healthcare professionals better manage these resources and save lives by proving the utility of ML algorithms in aiding with data-driven decision-making and allowing early treatments, resource planning and finances.