Is Blockchain the future of Research Data Integrity?

Is Blockchain the future of Research Data Integrity?

Fifteen years ago, British computer scientist Michael Smith and his longtime collaborator, the world-renowned child psychiatrist Robert Goodman, digitized two psychiatric tests and made them available online, free of charge.

Clinicians all over the world began using these tests in their own private practices or as part of research studies, with the raw data stored and catalogued on servers that Smith controlled.

Once data were recorded, Smith had the power to decide how to manage it. A self-described data "fanatic," he opted to keep original records of each survey, although he knew that others in his position might instead choose to overwrite aberrant data to eliminate statistical noise.

"Data smoothing, as many would call it, is unfortunately a fairly common thing," Smith told MedPage Today. "People for various reasons – not [necessarily] nefarious reasons – may process the data, smooth it out, to make it look better. In a way, it's a lie, from a data point of view."

Because Smith and Goodman alone controlled the servers, it wasn't long before one of his competitors accused them of tampering with patient data after they had been recorded. Smith, who declined to name his accuser, called the allegation "absurd."

To prove his innocence, he turned to an elegant new technology called blockchain. Smith shut himself away for 18 months, meticulously programming multiple iterations of a so-called blockchain system. Blockchain is the name for the technology that underpins digital "cryptocurrencies" such as Bitcoin, but has a much wider range of actual and potential uses.

When he finally deployed his system, his critics "retracted the complaint and went away," said Smith. "It cleared the whole thing up. It solved the issue."

Internal Workings

Like most people, Smith first became aware of blockchain as the backbone for Bitcoin, but quickly realized its potential as a steward of health and medical data. Blockchain provides an unprecedented ability to audit everything that happens within a network. Once a transaction is recorded, it cannot be clandestinely altered or discarded.

"What you have, in effect, is something like an aircraft's black box recorder in every piece of software," Smith said. Like a black box recorder, blockchain allows auditors to reconstruct exactly what happened within a machine, and when.

Now, anyone Smith invites to monitor his database has a unique window into the flow of information across his servers. But the window is translucent rather than fully transparent; it reveals each timestamped data point without revealing the underlying patient data.

Within a blockchain system, identical copies of the same database are interlinked across multiple computers, and all transactions are bundled into time-stamped records called "blocks." Any change to one database will be reflected throughout the network, making it nearly impossible to tamper with the data.

Blockchain is also very secure, encrypting information so it's revealed only to the intended recipient. Also, users are identified by numbers, not names, so the system also maintains each individual's privacy.

Smith would like to see his custom-built blockchain model or one like it, applied to clinical research, "where the temptation is very great to alter your results," he said.

And he's not alone. A 2017 IBM survey found that about 16% of healthcare leaders planned to start experimenting with blockchain by the end of the year, and BIS Research said the global market for blockchain health solutions will reach $5.61 billion by 2025.

Within healthcare, much of the blockchain hype has focused on fixing the gnarled mess of electronic medical records. But clinical research is another promising application, with leading computer scientists, healthcare executives and regulatory agencies convinced that this technology will transform the business of human experimentation, not only reducing hanky-panky but also speeding up drug development timelines.

Typically competitive pharmaceutical giants Pfizer, Amgen, and Sanofi have teamed up to pilot a blockchain-enabled clinical research platform. Their goal is to cut R&D costs by improving the way data are stored and shared across research sites, and potentially automating certain aspects of research.

And the FDA -- which sets ground rules for clinical research -- has partnered with IBM to explore how blockchain might be used to democratize health records, giving patients direct control over their health data while also unlocking that data for research purposes. The goal, according to an FDA spokesperson, is to create "a scalable data exchange ecosystem that can support high quality research while safeguarding against breaches of sensitive patient-level data."

The use of blockchain in clinical research is converging around three main trends, said Michelle Longmire, CEO of Medable, a clinical research and technology company. The first application is preventing data tampering and improving research reproducibility. Second is using blockchain to promote data sharing among institutions that otherwise have little incentive to share. And the third is giving patients direct control over data that's collected on them in a clinical trial.

Blockchain is particularly well-suited to these applications because of its "ability to negotiate the tension between data privacy and data sharing," said Maria Palombini, who helps set global standards for emerging technologies as part of her work at the Institute of Electrical and Electronics Engineers, better known by its acronym IEEE.

But it's early days for blockchain, with Palombini noting it will probably take years, even decades, before large-scale clinical research programs are conducted using the technology. She compared today's blockchain ecosystem to the internet's dawning – early adopters can see its potential, but the future is still difficult to predict.

As with all new technologies, "there are going to be small advances and people are going to fail along the way," she said.

Building Utopia

Blockchain can't prevent human error, like a doctor jotting down a blood pressure reading incorrectly, but a blockchain-enabled clinical research system could go a long way in preventing mistakes and deception after data are recorded.

In theory, here's how a blockchain-enabled clinical trial would work: First, a group of research centers, hospitals, life science companies and their contract research organizations would come together to form the "nodes" of a trial network. Each would volunteer server space, and all the data collected and stored would be verified simultaneously across the network.

Observers can see each timestamped data point but not the underlying patient data.

Throughout the trial, each component -- from informed consent to dissemination of results -- would be managed with "smart contracts," or software designed in such a way that the completion of each step unlocks the next, like a chain reaction.

Smart contracts are a perfect tech solution to help automatically execute the steps of a study protocol, which is a detailed document describing the study design, explains Mehdi Benchoufi, MD, assistant professor of epidemiology at Hôtel Dieu Hospital and the University of Paris Descartes, and a blockchain expert.

"It's impossible to cheat without everyone witnessing," said Benchoufi. Today, for example, a pharmaceutical company might promise to publish results at the end of a trial, but later decide to keep those results secret. With smart contracts, blockchain could prevent this obfuscation because the completion of one event (e.g., tallying of results) automatically unlocks the next (publishing results online).

Better yet, throughout the course of a study, patients could be able to on check a trial's progress themselves using an online portal or mobile app, giving them more control over personal data that's collected during the course of a study. Eventually, "that data becomes licensable by biopharma ... and they [the patients] get revenue from that licensing," said Longmire, who is working to create a clinical trial network for multiple sclerosis that would operate under a personal data licensing model.

That's the vision. Today, the industry is still very much in beta-testing mode, and there are no hard-and-fast rules governing how blockchain research must be designed.

Different people have different ideas about how this system should work. Some, like Benchoufi, argue research should be conducted on an "open" network, modeled after Bitcoin, while others argue researchers should use a "permissioned" network, in which each member must be authorized to join. Smith's blockchain system, now in its third iteration called ChainZY, runs as a private and permissioned network, though it can be configured to run publicly.

The next step is for all these groups to get together and agree on a set of ground rules – a process led by Palombini's team at the IEEE standards group.

"It can take 8 to 12 months before a standard becomes a reality," Palombini said. "When we get into actual pilot, that's when they're playing with real patient data. Then you've got to justify the big cost, which is partial implementation."

Ultimately, the pharmaceutical industry will need to drive the blockchain revolution in research because it's the only group with the resources to do so, said Lucila Ohno-Machado, MD, PhD, chair of UC San Diego Health's biomedical informatics department.

"The other [groups] can talk about it," Ohno-Machado said. "Regulators and sponsors are the ones who can make a change."

Improving Trust

One of the more abstract benefits of deploying blockchain in clinical research is that it might improve the public's trust of clinical research.

This idea was popularized by a 2016 proof-of-concept study in F1000Research, "How blockchain-timestamped protocols could improve the trustworthiness of medical science," which was trumpeted by The Economist and FierceBiotech as a blueprint for clinical research in the future.

But the paper sparked a heated debate over potential flaws in the author's methodology, and was soon retracted. The whiplash is perhaps symbolic of blockchain's hotly-debated potential.

Academics like Benchoufi are full believers in the idea that blockchain will improve public trust of research. With blockchain, even industry-funded studies would be deemed trustworthy and the reproducibility crisis would be solved, he argues.

And for pharma, improving trust isn't just an abstract benefit -- it could significantly impact the bottom line of any research program, with approximately one-third of study budgets devoted to finding and recruiting patients.

"That's the hype of blockchain," said Palombini. "Any time you have a trust issue? Boom. Blockchain."

But Cindy Geoghegan, patient advocate and principal of Patient and Partners LLC, isn't buying it. The average person doesn't associate blockchain with greater trustworthiness, she said. For Geoghegan, patients won't trust blockchain as a new research tool unless they see a direct, tangible benefit.

"It's not going to happen overnight," said Michel Goossens, an executive at pharma contract research firm ICON. But he said blockchain could eventually lead to a "paradigm shift" shift in which people expect their digital data -- including health data -- to be easily accessible, private and secure.

Bitcoin, the cryptocurrency that made blockchain famous, is a strong example that this cultural shift is possible, said Goossens. Over the past decade Bitcoin changed the way people think about currency markets by removing the imperative to trust a central bank.

Ultimately, patients will stop asking whether or not they trust the pharmaceutical industry -- or even if they trust their doctor -- instead placing trust in a bigger, more secure data management system, Goossens predicts.

"I definitely think it's a revolution," he said, "but I think the change is going to be a lot slower than people think."

Courtesy: MEDPAGE TODAY