Data-Driven Approaches to Improve Healthcare Supply Chain Management

Abstract

Healthcare supply chain management entails timely, reliable and cost effective delivery of medical consumables that faces several management challenges such as a short time from planning, to ordering, receiving and inventorying the goods, high fluctuation in the demand for these goods and shortage of inventories, inefficiency of selecting suppliers, etc. An artificial intelligence based data driven approach to Healthcare Supply Chain optimization is presented in this research. Long Short Term Memory (LSTM) is implemented for demand forecasting, with accuracy of 94.2% which surpasses traditional ARIMA models. Supply chain optimization was carried out by applying the Genetic Algorithm in order to reduce the overall costs by 18.6 % with 22.3 % increase in delivery efficiency. Supplier segmentation is done though K-Means clustering to classify suppliers into high performance, moderate, and low performance groups giving 16.8% improvement in supplier reliability. Inventory management was handled through the Apriori algorithm, resulting in stockout reduction by 27.5% and overstocking reduction by 21.4%. Experimental results were shown to enhance the efficiency of supply chain compared to that of the conventional methods. We verified by a comparative analysis with existing studies that our approach is superior to the existing studies in terms of the improvement of accuracy rate at forecasting, reduction of cost and optimization of inventory. A major weakness are the infrastructure and data security issues that still exist despite the promising findings. Future research might extend this work towards exploration of hybrid AI and blockchain models to provide more resilience in supply chains. The results of this study provide a scalable framework in which to manage the healthcare supply chain and improve both better resource allocation and better service delivery.