Is blockchain the solution to mitigate the risk of unreliable clinical trial data?

By Federico Turkheimer, PhD, King's College London and LabTrace, and Eric Wragge, Algorand Foundation

Innovation and trust in that innovation are key to a healthy and active clinical trial ecosystem. This combination encourages the advancement of new treatments into clinical trials and the enthusiastic willingness of patients to participate in those trials. This combination makes the treatment possible in the first place, as the data of scientists from the near and distant past is used by their colleagues in the present.

At the same time, the flood of new data generated by novel medical technologies such as advanced biomarker testing has led to several fundamental discoveries. These discoveries (and the treatments they enable) hold great promise for patient care. But where innovation comes, trust must also come.

Unfortunately, these technologies are increasingly vulnerable to plagiarism, forgery and fabrication. Scientific fraud has existed since the dawn of science, but the latest figures show that the number of retracted papers in Europe has quadrupled in the last 20 years. Worryingly, the reasons for retractions have also changed. While in 2000 most retractions were due to ethical and legal issues or questions of authorship, today the same proportion are due to unreliable data.

Growing concerns have prompted government funding agencies and charitable foundations to take action. There are several possible solutions, such as more intensive training on professional codes of conduct for researchers and discipline-specific guidelines (and consequences). The UK Concordat in Support of Research Integrity is an excellent example of these efforts. Similarly, the British Neuroscience Association has introduced credibility toolkits that rely primarily on pre-registration of projects, where one clearly and openly explains one's experimental rationale, hypotheses and methods (including sample size and statistical analyses to be used) before conducting the experiment.

These are great initiatives that should continue to prevent inaccurate or unethical research in the first place. But they do not serve the treatments and trials currently underway. In order to continue these trials, they need data that can be trusted. And in truth, they need data that a broad audience can trust. It can be complex and time-consuming for a trained researcher to evaluate the work that an experimental treatment is based on. But now that concern about increasing scientific fraud has become mainstream, methodology or data verification must also become mainstream.

Ideally, there would be an easily navigable record of the provenance and authenticity of the data. CROs or regulators could access a “stamp of approval” that the data has been reviewed, confirmed, and not retracted – a sign that it can be trusted. This type of secure record was not possible in the past, but today it is possible with blockchain. Clinical trials could benefit from using blockchain in their experimental workflow as a novel and effective tool to ensure the integrity of the process.

A blockchain is a distributed public ledger that is immutable and maintained by a very large network of participants (nodes). Nodes in the network continually verify the accuracy of the ledger and add to that ledger through a consensus protocol that adds records (also called blocks) that are securely linked to each other via cryptographic hashes. While blockchains are associated with cryptocurrencies in the popular press, they basically function as databases that allow history to be recorded securely and immutably.

For example, once a file is uploaded to a particular blockchain's platform, a file identifier (content hash) is created that is uniquely linked to the file content. The hash is then written to the blockchain along with all relevant information. The user receives a certificate that contains the hash, the additional information, and the link to the block in the ledger. While the certificate remains public, the user has the option to either publish the file or keep it private within their own firewalls (the entire process is GDPR compliant); in the latter case, proof of genuine certification can be provided at any time.

The flexibility of the platform allows for the recording of a unique data identifier for each file type at a specific point in time (timestamp) and the sharing of evidence of its accuracy (certification). The platform also allows for the creation of a chain of evidence using secondary data (data extracted from raw data such as images using software). The products of such processing, which we call secondary files, can then be linked to the primary data and the software used via the blockchain. An auditor can easily verify these chains of evidence.

In summary, while recent technological advances have created fertile ground for new forms of science fraud, novel technologies can also be leveraged within an ethical framework to ensure a transparent and certified process for experimental science. In addition, second-generation “green” blockchain systems (e.g., those with a low carbon footprint) now exist, alleviating concerns about the extreme energy requirements of earlier blockchain systems. Additionally, as we discussed above, it is easier than ever to embed blockchain into proprietary systems. For this reason, we are confident that blockchains represent the best approach to embed trust into notarized scientific systems.

Scientific research requires trust, whether in the laboratory or in clinical trials. Clinical trial planners must rely on the results of their colleagues from years past to develop and seek approval for their new therapeutics and treatments. This trust then carries over into the relationship between clinical trial planners and the patients who volunteer to participate in their trials. And when a trial is successful, that trust extends to all of us – everyone who might one day benefit from a discovery. The reliability of data is relevant to all of us, so its integrity should be accessible to all of us.

About the authors:

Federico Turkheimer PhD is a neuroscientist and Professor of Neuroimaging at the Institute of Psychiatry, Psychology and Neuroscience at King's College London, where he is also Director of the KCL Institute for Human and Synthetic Minds. He is also Director of LabTrace, a technology startup aiming to provide secure and GDPR-compliant data traceability for clinical trials and medical research.

Eric Wragge is the Global Head of Business Development at the Algorand Foundation, an organization dedicated to supporting the developer ecosystem by building high-impact, secure, and scalable real-world blockchain applications on the Algorand blockchain.