Pharmaceutical inventory management is a critical operation in healthcare centres. This is due to the fact that most pharmaceutical products are perishable. Managing the inventory of perishable items can be a complicated process as the healthcare industry needs to maintain a high level of services. In order to manage the inventory of pharmaceutical products, it is important to forecast the demand, which will enable the distribution to be planned and scheduled effectively. In this research, we focus on one fast moving medicine which is paracetamol that commonly used to treat fever and pain across all ages group of patients. Data is obtained from University Malaya Medical Centre (UMMC) for the year 2017-2020. Before applying the forecasting techniques, the data pattern needs to be identified. Among the five forecasting techniques are Additive Decomposition Method, Multiplicative Decomposition Method, Simple Exponential Smoothing and Adaptive Response Rate Exponential Smoothing. The performance of these techniques was evaluated based on four error measurements; (i) Mean Absolute Deviation, (ii) Mean Squares Error, (iii) Mean Percentage Error and Mean Absolute Percentage Error. Multiplicative Decomposition method displays the lowest values of error measurements which indicates the greatest accuracy and implies the suitability for this research. The data predicts that the demand for paracetamol will likely continue to move downwards over the next five years.
Read full abstract