Implementation of a Predictive-Analytics Framework for Pharmaceutical Demand Forecasting and Supply-Chain Resilience in U.S. Healthcare Markets

Abstract

The U.S. pharmaceutical industry faces increasing challenges arising from demand volatility, drug shortages, pandemics, regulatory changes, and disruptions in global supply chains. Traditional forecasting approaches often fail to provide accurate predictions and proactive risk management capabilities. This study proposes a predictive-analytics framework integrating machine learning, time-series forecasting, and supply-chain resilience mechanisms to enhance pharmaceutical demand forecasting and ensure continuity of supply. The framework utilizes historical sales records, epidemiological trends, demographic variables, and external market indicators to generate accurate demand predictions while employing resilience metrics to mitigate disruptions. Empirically evaluated across 864 monthly observations covering six pharmaceutical products and four U.S. regions (2023–2025), the proposed Hybrid LSTM-GRU model achieved a mean MAPE of 4.33%, outperforming the XGBoost benchmark (6.95%) by 37.7%. The supply-chain resilience module identified a mean Resilience Index of 47.6, an overall stockout rate of 15.2%, and a High shortage risk classification rate of 51.2% across the portfolio. Hospital Injectables and Antiviral drugs exhibited the lowest resilience scores (32.8 and 37.9, respectively), driven by API source concentration indices exceeding 80 and lead times of up to 37 days. The resilience module recommended 64,821 total reorder units, with 51.2% of records requiring expedited supplier diversification. These findings demonstrate that improved forecasting accuracy alone is insufficient to resolve structural supply-chain vulnerabilities; targeted supplier diversification, dynamic safety stock policies, and regionally differentiated inventory strategies are essential complements. The proposed model supports healthcare organizations, manufacturers, and distributors in optimizing inventory management, reducing shortages, and improving patient outcomes.