Pharmaceutical companies have struggled for years to track their products throughout their supply chains, making it possible for counterfeiters to introduce fake drugs into the market.
To counteract this problem, a new system to trace and track drugs is needed. Researchers think that blockchain may provide the technological foundation for such a system because it can successfully track drugs and help prevent the circulation of fake ones.
Before diving into how blockchain could become a solution, we must examine the source of the problem.
Counterfeit Medicine
Counterfeit drugs are defined by the World Health Organization (WHO) as “drugs that are manufactured fraudulently, mislabeled, with low quality, hiding the source detail or identity and do not follow the defined standard”.
While these drugs may contain some genuine ingredients, they may also contain toxic ingredients at the production level. If consumed, these drugs can cause serious health problems.
Sometimes, the manufacturers of counterfeit medications use the logos of reputable companies to allow their product to enter the market. Although this is a worldwide problem, developing countries are disproportionately affected by this issue.
Counterfeit drugs are distributed throughout a very complex network, making them difficult to detect and remove. To prevent counterfeit drugs from being distributed, a system is needed that can trace and track drug delivery at every stage. Blockchain is the latest innovation that can handle the complex supply chain and track the product at every stage.
Blockchain Applications
Blockchain was designed to store the transactions log for Bitcoin. It provides a rich ledger software to store data records, which are transaction logs arranged in several blocks. A blockchain system collects data related to the time, date, price, and participants involved in each transaction.
When one block of information changes, all the pieces update accordingly, giving up to date information for every transaction.
When applied to the pharmaceutical supply chain, blockchain offers an electronic ledger where everyone in the network can see and validate transaction information.
To create a secure drug supply-chain management system, the use of blockchain with Hyperledger fabric was proposed. This ledger software is capable of monitoring and tracking all parts of the drug delivery process. The Hyperledger fabric can configure multiple world state databases to maintain the set of current values, and, when applied to the pharmaceutical world, it enables medicine to be accurately traced regardless of where it is in the world.
The DSCMR System
Researchers have proposed a blockchain and machine learning-based drug supply-chain management and recommendation system (DSCMR) to combat counterfeit drugs. First, we must distinguish between the two complementary components of this system — the drug supply management system and the recommendation system.
In the drug management system, users can track the drug at every step, make orders, update orders, and more.
The recommendation system works to recommend the best medicine to pharmaceutical company customers through the use of the Ngram and LightGBM machine learning-based modules.
The Drug Management System
With the drug supply-chain management system, users can perform various transactions such as checking drug information, tracking and tracing orders, updating records, etc.
The system’s users can include patients, doctors, manufacturers, distributors, pharmacies, and hospitals, and more. Data related to the various users are stored within the blockchain system.
Each user is provided with a web application where they can perform their transaction and communicate with the blockchain network, and each user can track the status of drug delivery. Furthermore, all users must receive permission to check the complete details of drugs. They will use the client application to log in and perform their respective transactions. This works to increase the levels of security along the pharmaceutical supply chain.
To submit a transaction proposal, users must submit a request using their credential. The request, if valid, is sent to peer nodes in the system to be reviewed and approved. These peer nodes work to examine the proposal and give approval if it is valid and fulfills the smart contract criteria. They also work to validate the results of the transactions and record them in the ledger. Once this has been complete, the ledger updates all data for everyone to see.
The proposed procedure can be seen here.
The Drug Recommendation System
The drug recommendation system, which was created with both natural language processing and machine learning techniques, works to recommend the best medicine to pharmaceutical company customers through the use of the Ngram and LightGBM machine learning-based modules.
A public drug review dataset based on the reviews of drug users was used to train the models of this system and to predict the best and most effective drugs in the pharmaceutical industry. The review data include information about any side effects, benefits, and comments from customers received in the client application mentioned earlier.
The recommendation system is also able to train itself through the N-gram model by recognizing new reviews and updating the recommendation results accordingly. This is shown in this figure.
The use of smart contracts is also implemented to limit the number of individuals involved in a certain transaction. A smart contract aims to provide the parties involved a way to exchange information, property, or even money without the use of a third-party agent or broker. It is typically lines of computer code that enforce the agreement made without using a middleman. In the proposed network, Java and Node.JS were used to write smart contracts. These smart contracts are stored in the distributed ledger of the blockchain network, where they are protected from tampering or possible deletion.
Experimental Results
Several conclusions emerged from testing the proposed machine learning-based drug recommendation and management (DSCMR) system.
In the drug management system, the client application mentioned earlier can communicate with the blockchain system effectively. Once each user’s identity is validated by the peer nodes, they are able to start transactions. As shown in this figure, manufacturers can add, update, or delete drug details in the blockchain network. Other participants like doctors, hospitals, and pharmacies can view these records and make changes and updates as needed to the drug information, as seen here.
The REST server composer aids in the communication between the client application and the blockchain network. Any requests to validate transactions are sent through the server, then stored in the blockchain network. (This process is demonstrated here.)
As for the drug recommendation system, four conditions were tested: acne, high blood pressure, anxiety and alcohol dependence. Using the client application, the system worked to accurately suggest the best medicine to customers in a secure and transparent way. Patients could also track the drug source to determine whether it is real or fake.
Overall, the system offers four main advantages in combating drug counterfeiting. First, the DSCMR system can successfully recommend the top-rated and best medicine to the customers of the system. Second, it allows patients to access the system directly through a client application, and pharmaceutical customers such as doctors can access the system using their credentials. Third, the DSCMR system can provide customers with a secure and transparent experience.
Finally, through the DSCMR system, patients are able to check the source of their drug by scanning a barcode, to which the system will provide the entire source of the drug.
Whether it is doctors, pharmacists, hospitals, or manufacturers, everyone involved in the chain will be able to track the relevant drug along the supply chain. This helps prevent anyone from entering fake drugs into the chain, hopefully reducing the problem of counterfeit drugs.
Daniel Browning is business development coordinator at Do Supply Inc.