Abstract:
The Lebanese pharmaceutical supply chain (PSC) suffers from two main challenges: drug shortages and the presence of illicit drugs in the market. To address these challenges, this research describes the creation of a blockchain application, integrated with a predictive model, to enhance the traceability of drugs in Lebanon and forecast pharmaceutical demand. This research represents the first study that integrates blockchain technology with predictive modeling for drug traceability in Lebanon. Solidity smart contracts have been employed on the Ethereum blockchain to create a traceable ledger of drugs in efforts to combat illicit drugs entering the market. Moreover, two predictive models – ARIMA and LSTM – have been implemented to forecast pharmaceutical demand and mitigate against drug shortages. While the ARIMA model performed poorly in demand prediction, the LSTM model showed great potential: it scored 21.63 RMSE on the test data and 103.3 RMSE on the train data. Further expansion of datasets and including external factors that affect demand should improve predictive accuracy in the future. This work seeks to provide a holistic approach toward improving the PSC in Lebanon to surmount drug shortages and illicit drugs, hence ensuring patient safety, improved inventory management and operational efficiency in the pharmaceutical industry.
