The purpose of this study is to examine how predictive analytics can optimize demand forecasting for controlled medications, with the goal of improving cost minimization and pharmaceutical supply chain management in Saudi Arabia’s health-care environment.
This study uses historical consumption data (2015–2020) for five high-demand controlled drugs: Codeine, Tramadol, Morphine, Clonazepam and Lorazepam. Four forecasting techniques – Simple Moving Average, Exponential Smoothing, Double Exponential Smoothing and Linear Regression – are applied and evaluated using error metrics such as Mean Absolute Deviation, Mean Absolute Percentage Error, Root Mean Square Error and Tracking Signal to identify the most effective methods.
No single approach proved universally optimal. Exponential Smoothing minimized costs for Codeine and Lorazepam, while Linear Regression excelled for Tramadol and Clonazepam. Morphine required combining Exponential Smoothing for accuracy with Double Exponential Smoothing for cost control. These outcomes highlight the importance of tailoring forecasting strategies to drug-specific consumption patterns and organizational priorities.
Future studies should incorporate broader variables, such as patient demographics and supply chain disruptions, to enhance predictive accuracy.
The findings of this study offer actionable recommendations for health-care administrators and policymakers that enable data-driven decisions to improve drug availability while minimizing procurement costs.
This study contributes to the growing field of pharmaceutical supply chain management by providing a comparative analysis of forecasting methods tailored to controlled substances. This study underscores the importance of integrating advanced predictive models into centralized procurement systems to enhance cost-efficiency and resource allocation.
