Guest Column | March 1, 2018

3 Keys To Successful Blockchain Adoption In Clinical Research

By Bhargav Adagarla and Brian McCourt, Duke Clinical Research Institute (DCRI)

3 Keys To Successful Blockchain Adoption In Clinical Research

Excitement around the potential for blockchain platforms continues to build. This parallels the surge in popularity of blockchain-based cryptocurrencies such as Bitcoin. Meanwhile, “critics argue it’s all hype — a technological hammer looking for a nail — and that the complexities of health information prevent practical use of blockchain technology.”1 Nonetheless, several initiatives that promise to leverage blockchain are underway in healthcare, clinical research, and other fields:

  • IBM Watson Health and the FDA are developing an initiative that will define how blockchain can be used to exchange healthcare data from varied sources including electronic health records, clinical trials, and wearables.2
  • DeepMind, Alphabet's AI division, announced it will build a “verifiable data audit” for health records using a blockchain-like ledger. The company claims it will make what is being done with healthcare data both verifiable and auditable in real time.3
  • Healthcare startup Patientory aims to develop a blockchain-based medical record system that is decentralized and gives patients control over their data.4,5
  • Sweden is piloting the management of real estate transactions on a blockchain. This registry aims to secure and verify the authenticity of a process/file, the identity of the signatories, and the provenance of a property.6
  • A partnership between Australia Post, a postal service, and Alibaba will explore storing entire supply chain transactions on a blockchain to fight the rise of counterfeit food in China.7

What Are Blockchains?

Blockchains are distributed ledgers where data is bundled into blocks.

They can provide immutability and, consequently, data integrity; eliminate the need for a central intermediary, allowing for trustless interactions; and bestow fault tolerance and robustness due to their distributed nature8.

Each block contains a hash — which takes one input and calculates one unique output — of the preceding block. This hash acts as a pointer to the predecessor, chaining the two blocks together. If data in one block in this resulting chain of blocks is changed, the succeeding blocks are invalid.

When implemented as a network, new blocks are added to the chain and broadcast to other participating entities (nodes) in the network. As the new blocks are propagated across the network and accepted by the nodes into their chains, the data in the blockchain becomes immutable.

Depending on the logic that determines which blocks the nodes accept as valid, the blockchain network can be used to eliminate the middle man or a central authority that validates the data being exchanged within the network.

What’s Needed For Blockchain Adoption In Clinical Research?

As interest in leveraging blockchain for clinical research surges, we have identified three key needs:

1. Design Model

A model is needed to demystify blockchain and help researchers and investigators develop blockchain applications. We proposed a simple, conceptual model for design of applications for clinical research based on implementations of blockchain.9 This model can be used to map entities and processes in clinical research to blockchain elements to facilitate the design of blockchain applications.

The four elements of this conceptual model are:

  • Nodes: Nodes are entities participating in a blockchain implementation. Nodes can represent research organizations or their subunits, such as investigational review boards (IRBs), clinical data systems, and regulatory bodies such as the FDA. They can also represent people involved in the research such as participants and investigators.
  • Transactions: Transactions represent the clinical research processes that involve interactions among the various entities involved.
  • Data: This layer represents the clinical research data generated by the transactions and the associated audit trails.
  • Logic: This is the business logic that binds the above three layers, determining the source code underpinning the system.

2. Evaluation Model

An evaluation model would be useful in helping researchers break through the hype of blockchain to identify and prioritize projects that could provide real value.

A simple evaluation model would involve two axes: ease of implementation and the extent to which the innovation uses blockchain’s strengths. Ideal projects would leverage the core strengths of blockchain and have high implementation ease (and therefore low implementation costs). More complex models could also include evaluation of how the innovation solves a particular problem in clinical research and whether the problem can be solved using existing information systems in which the organization has invested resources.

3. Implementation Road Map

An implementation road map describing the various components of a blockchain solution for a clinical research problem could both inform the evaluation model and learn from it. A review of blockchain applications such as Bitcoin and a land registry,6 and platforms such as Hyperledger,10 provides an understanding of the major components of an implementation road map for a blockchain solution.

Basic components of such a road map could include:

  • Permissioning: Blockchains can be permissioned to be public, as with Bitcoin, or private. Private blockchains, where nodes can participate by invitation only, are better-suited to clinical research applications than public ones. The alternative is a consortium blockchain where a certain number of nodes control the consensus and thereby the validity of transactions, while the other nodes are limited to reading the blockchain.11
  • Identity management: Identity management is the process of credentialing and maintaining the identities/roles of each node in a blockchain. In healthcare and clinical research, there is a need to identify nodes and link them to unique individuals or entities/organizations and their identification documents in the real world.
  • Consensus logic: This describes how the nodes will reach consensus when the implementation lacks a central authority.
  • Governance: The governance component represents the regulations and guidelines that are intrinsic to clinical research. This deserves special attention during implementation due to its potential to influence both feasibility and cost.
  • Interfaces: Interfaces allow blockchain applications to interact with other systems in the complex healthcare and clinical research ecosystems. For example, some or most of the data might be stored off the blockchain and in traditional systems to avoid storing big data on blockchains, creating the need for interfaces. Also, interface design can be complicated, given the interoperability problem inherent to the domain.

Looking Forward

Throughout drug development, data is exchanged among various stakeholders. Monitoring and validation — which are major cost drivers — are essential to the integrity of this data and the resulting credibility of the scientific findings. Establishing trust among the various actors in clinical research is also essential, in the form of contracts, consents, and agreements.

With their proven ability to underpin decentralized cryptocurrencies, blockchains could ultimately be used to tackle the cost of establishing trust in clinical research. However, in the immediate future, a more realistic goal might be to address issues around data integrity, reproducibility, and provenance. While leaders slowly adopt blockchain within the clinical research ecosystem, the most immediate need is arguably for resources to help them see past the hype to identify real value.


  8. Tanenbaum AS, Van Steen M. Distributed systems: Principles and paradigms, Second Edition, December 2009.
  9. AMIA 2018 Informatics Summit Oral Presentation, March 13, 2018. Blockchain Applications in Clinical Research .B. Adagarla, A. Tasneem, T. Reece, Duke Clinical Research Institute; S. Balu, Duke Institute for Health Innovation; A. Krishnamurthy, University of North Carolina at Chapel Hill; B. McCourt, Duke Clinical Research Institute.

About The Authors:

Bhargav Adagarla is a senior informaticist within Clinical Research Informatics at DCRI. He has several years of experience providing data management and software engineering support to clinical and basic science research and managing research information systems supporting translational research. He has an M.S. in computer engineering from the University of Kansas and a B.E. from Osmania University, India. He can be reached at


Brian McCourt is the director of data solutions group at the DCRI and is an active leader in Clinical Research Informatics with experience in clinical research supporting both academic investigator-initiated clinical research and industry-sponsored pharmaceutical and device projects. He has led a variety of infrastructure initiatives supporting large research programs and organizations and has unique experience representing the breadth of clinical research informatics. McCourt received his bachelor’s degree from Saint Anselm College and worked at Massachusetts General Hospital before joining the DCRI.