Patient Treatment Classification using Machine Learning

Clinicians and specialists often face the challenge of determining whether a patient requires inpatient or outpatient care by analyzing the results of laboratory tests. But it is a time consuming process as well as demanding, as it needs significant effort from doctors to make accurate decisions. Moreover, there is a high risk of making incorrect judgments, which can jeopardize the patient's life. This research article intends to utilize a range of Machine Learning (ML) algorithms in order to categorize patient care as either outpatient or inpatient, with the goal of alleviating the workload on doctors and enhancing the quality of services delivered. This study engaged machine learning to construct six models like Logistic Regression, Random Forest, Support Vector Machine, K-Nearest Neighbors (KNN), Decision Tree and Naive Bay's. These models utilized patients' conditions and laboratory test results to predict whether they should be categorized as inpatient or outpatient. The selection of the best model was based accuracy. For this purpose, a dataset from Indonesia's private hospital was used. The dataset consisted of patients' laboratory test results, extracted from their electronic health records (EHR). The models were built using this dataset and subsequently tested. The findings indicated that among the different algorithms tested, the Random Forest algorithm exhibited the good accuracy rate of 92.44%. These findings highlight the possibility of utilizing machine learning to enhance the accuracy and efficiency of patient care classification, thereby reducing the occurrence of human errors that could pose risks to patients' lives and financial burdens to their wallets.