Forecasting For Bed Capacity In Maternity Hospital Using Machine Learning

Abstract

Hospital Bed Capacity planning affects economic and social sustainability in healthcare through bed capacity efficiency and medical treatment accessibility. Conventionally, this problem is solved using programming or simulation models with assumptions and limits. Forecasting the Hospital Bed Capacity using time series data on bed occupancy has been considered but not with factors such as the Number of Hospitalized Patients Hospital Diagnosis and patient’s length of stay. This study proposes a data-driven methodology to forecast the Hospital Bed Capacity using Machine Learning. The length of stay classification is performed using several Machine Learning techniques, including the support vector machine, decision tree, and Linear regression. Also, linear regression is applied for the Number of Hospitalized Patients forecasting. The forecasting and descriptive analysis outputs based on length of stay classes are directly applied to a simple mathematical model to predict the required bed capacity. This methodology is implemented in a case study at the Maternity Hospital, leveraging a comprehensive data set that includes 51,585 records. Developed Machine Learning algorithms in Python have facilitated this analysis. The findings indicate a critical review of the hospital's current bed capacity is necessary to ensure it meets the demand efficiently. Additionally, several managerial recommendations are formulated to enhance patient care and operational efficiency. The analysis reveals that a minimum capacity of 116 beds per hour is necessary to accommodate the urgent demand of patients. In order to mitigate the uncertainty associated with factors such as overnight stays and fluctuating daily admissions, it is suggested that the bed capacity be increased by 15%, resulting in a total of 133 beds. This modification guarantees readiness for any circumstance, emphasizing the existing deficiency at the Maternity Hospital, which possesses an only 45 beds. Hence, it is crucial to promptly expand the capacity to accommodate a minimum of 116 beds by 2024 in order to meet the anticipated future demand. This would guarantee sufficient treatment for patients and ensure the facility is prepared.