Predicting the usage of seasonal drugs by utilizing advanced machine learning: An albuterol case study.

Manual forecasting of seasonal medication demand results in inefficiencies and labor burden. With the advancement of machine learning, there is an opportunity to develop a machine learning model that can predict the expected usage of medication for the next season by identifying patterns in usage data. This study aims to assess the application of a seasonal autoregressive integrated moving average (SARIMA) model to forecast the use of inhaled albuterol. This retrospective study utilizes 5 years of administration data. The primary outcome is the model's ability to predict future usage. The secondary outcome is estimated labor cost savings if the model were used in operational workflows. The model effectively forecasted demand for inhaled albuterol for the upcoming season with high accuracy. There is potential to save thousands of dollars in long-term labor costs. When scaled across multiple seasonal drugs, the savings could be substantial not even just on a financial piece but when translated to hours utilized by employees. Implementing SARIMA modeling for seasonal medication forecasting improves prediction accuracy and has the potential to reduce manual labor in purchasing. Our hope is to add to the small but growing body of pharmacy machine learning literature in America.