Demand-Driven Inventory Optimization for Regional Pharmaceutical Distribution: Reducing Stockouts, Waste, and Cost Through Data-Driven Forecasting
Year: 2026
Type: preprint
Abstract: Two problems dominate regional pharmaceutical distribution: stockouts, where patients cannot get the medicine they need, and overstock, where medicine expires unused and is written off as waste. Both trace back to the same root cause: inaccurate demand forecasts. This paper presents a data-driven demand forecasting framework using five years of real daily pharmacy sales data (2014 to 2019, eight drug categories). A machine learning forecasting model is compared against a naive seasonal baseline, achieving a 23.9% to 26.5% reduction in forecast error (MAE and RMSE). This error reduction is translated into business impact through the standard safety-stock relationship, and combined with a published total-cost-reduction benchmark (13.4%) from risk-based supply chain optimization literature. Applied to illustrative regional inventory and logistics spend scenarios (USD 20M to USD 150M), the framework suggests estimated annual savings of USD 2.68M to USD 20.1M. Limitations and an implementation plan for a regional pharmaceutical operations context are discussed.
Author: Prakasa, Nezar Abdilah
Language: English
Field: Decision Sciences
Domain: Social Sciences
Open Access status: green
OpenAlex ID: W7235153014
