Predicting Patient Arrival Rates in Outpatient Department to Optimize Patient Flow

Objectives: This real-life study explores the application of Machine Learning (ML) models to predict hourly patient arrivals in multi-speciality outpatient departments. Also to find the features which affect the arrival rate of patients in different departments. Methods: The patient arrival data, 21298 patients in four specialities (Dental, General medicine, Orthopaedic, ENT) of the Outpatient Department (OPD) of a Multi-Speciality hospital in Gurugram, during the Nov – Dec 2023 have been extracted and pre-processed. In this study, ML models, including linear regression, K nearest neighbors, random forests, decision tree and gradient boosting machine are applied to identify the most effective approach for predicting queuing behaviour using historical arrival data. Afterwards, random forest feature selection algorithm was also applied to select the best features for each speciality. Findings: Random Forest algorithm produced lowest mean absolute error of 2.70 and 3.29 for orthopaedic and dental department respectively. However, Gradient Boosting Machine is found to be the best predictive ML models with least mean absolute error of 2.53 and 3.25 for the general medicine and ENT department respectively. Best subset for dental and general medicine speciality contains day of the month, lagged arrivals, revisit number and hour of the day. For ENT and orthopaedic speciality, day of the month, lagged arrivals, hour and morning shift appear to be the best predictors. Novelty: This work is one of the first to use ML to predict the number of patients in a multi-speciality OPD, emphasising feature selection and model comparison. By better predicting and controlling patient flow, the results give hospital managers important insights to maximise resource allocation and increase patient satisfaction. With accurate prediction of arrival rate, hospitals can better manage resources and staffing, thereby reducing wait times and improving patient care. Keywords: Arrival Rate, Outpatient Department, Hospital, Multi-speciality, Machine Learning Algorithm