Proposal for Improving Inventory Management in a Pharmacy Through Automated Demand Forecasting, Reorder Point, and ABC Analysis

Guevara N.J.R.; Matos F.L.E.; Noriega-Aranibar M.T.; Gutierrez J.A.T. · Advances in Transdisciplinary Engineering (2026) · DOI: 10.3233/atde260720

Resumen

This study proposes an integrated model to improve inventory management in an ambulatory pharmacy through automated demand forecasting, reorder point (ROP) calculation, and ABC classification. The current system presents a 24.9% stockout rate, low forecasting accuracy (68.75%), and long replenishment lead times, generating reactive decisions and service discontinuities. Using historical consumption data from 2023-2024, the Random Forest algorithm achieved the best forecasting performance (MAPE 3.75%), serving as the basis for defining accurate ROP parameters for 12 critical medications. Additionally, the ABC analysis enabled prioritization of high-impact items, supporting the redesign of the internal warehouse layout. The proposed model was validated through discrete-event simulation in Arena, showing a reduction in stockouts to 3.8%, a decrease in replenishment effort by 38.8%, and an improvement in inventory accuracy to approximately 98.7%. The results demonstrate that integrating predictive analytics, structured inventory policies, and layout optimization significantly enhances availability, operational efficiency, and service continuity in ambulatory pharmacy settings.

Instituciones: Universidad de Lima