Blockchain technology is also known as “Distributed Ledger” technology. That is, a sequential ledger is shared (distributed) amongst a group of people/entities. This terminology, at least, seems to imply that blockchain would be a good way to represent—in digital form—the transactions in a supply chain, which are distributed geographically and over time. Way back in 2011 I posted an essay aimed at analyzing the viability of three different ways of digitally modeling the pharma supply chain. I called it, “The Viability of Global Track & Trace Models”. I’d like to revisit that essay to see how blockchain technology might affect my conclusions.
THREE BASIC MODELS
In the original essay the three models I looked at were Centralized, Semi-Centralized and Distributed. Today we see a fairly wide variation among the countries of the world that have mandated pharma track & trace (or more accurately, pharma tracing). That variation is due, in part, to the different problems each government is trying to solve, and how they believe those problems are best addressed with the technology available to their citizens and businesses. It’s no surprise that the US would choose different technology and a different model than, say, India. So, like most people, I rejected a global Centralized model as being unviable because it is unrealistic to expect every country in the world to align around a single repository/system.
I defined the distributed track & trace model fairly narrowly as “…one where each trading partner communicates with their immediate upstream and downstream trading partners under normal circumstances, and with other companies in the supply chain who may not be direct trading partners when necessary for certain applications.” The original essay was written well before the Drug Supply Chain Security Act (DSCSA) was enacted so it focused on the supply chain data model that was established by the requirements of the California Pedigree law. We don’t have to worry about that law today because the DSCSA preempted it, so some of the content of that original essay can be ignored (see “Preemption: What Does It Mean?”). But as a model, the California Pedigree law defined a distributed track & trace model. Each member of the supply chain held their own track & trace data; they received data from their suppliers; and they sent data to their customers. When track & trace history reports were necessary, they had to be generated by requesting data from each party who owned the given drug(s) throughout the supply chain. At the time, I felt that, to be viable, a Distributed model would necessitate less rigid regulatory requirements than we had in California.
What I called the Semi-Centralized model was characterized by multiple “central” repositories where data specific to a given country or a given group of companies were stored, depending on the geography, the business models and the regulations in effect. In the US, each manufacturer would store the data about the drugs they market in a single repository and all downstream trading partners would need to find that location when they had data to add, representing their ownership of those drugs.
Let’s take a look at how blockchain might fit into these three models. First of all, the idea of using a blockchain to hold all of the track & trace data has been rejected by multiple independent pilots as unworkable because of the difficulty of controlling data visibility and access to only authorized users. Most companies proposing viable track and trace solutions using blockchain are proposing only storing in the blockchain itself some sort of pointers to the actual data. That allows the target repository to perform checks to confirm authorized access, in one way or another, before it is granted (see “Important New Blockchain Study Launched” and “An Open Letter To Blockchain Vendors: Please Pay More Attention”).
In this way, the “distributed ledger” (the blockchain) would only act as a look-up directory, or, a “directory service”, or, a kind of “discovery service” where one could obtain the location(s) of the actionable track & trace data for (a) given unit(s) (see “Could Blockchain Technology Be Used For DSCSA Compliance?”). This would allow the track & trace data to be stored just about anywhere (outside of the blockchain), whether in a centralized government repository (though, probably not…see below), in an EPCIS server owned and controlled by a member of the supply chain, or a third-party server that offered the repository service for a fee. And it may actually (probably would) be spread across multiple repositories.
It doesn’t make much sense to offer a “directory service” or a “discovery service” when all the data is held in a private or government mandated central repository because there is only one place, and everyone who needs to know where that place is will know it. That leaves blockchain use relevant only in fully distributed or semi-centralized systems.
I don’t think purely distributed systems are ever going to exist because many of the companies in supply chains where a distributed model is allowed will opt to align with a vendor who will serve as the repository for their mandated track & trace data. These vendors will hold the data on behalf of multiple customers, thus resulting in one of multiple “semi-central” repositories. But other members of the supply chain may choose to serve as their own repository, so what was originally planned as a fully distributed model will rapidly transform, from the very beginning, into more of a semi-centralized model. Or we could call it a mixed model.
This is irrespective of blockchain or not, but the result is particularly a natural fit for “directory services” or “discovery services” (see “Will The Pharma Supply Chain Find Any Value In GS1 Discovery Services?”). In these models, no one can reasonably know exactly where the data they need currently resides. A directory or discovery service is necessary. These services could well be implemented with blockchain “under the hood”. In fact, I, and growing number of people, would say that blockchain is probably the best way to implement such a service.
I believe we are going to see multiple FDA pilot proposals that include a blockchain used as a directory or discovery service (see “FDA Announces Start Of DSCSA Voluntary Piloting Program”). It is possible that we will see this as the most common feature of all proposals submitted. Certainly the FDA will choose at least one of these proposals for their formal pilot(s) (this assumes they will actually “select” only certain proposals and not just say “sure, go ahead” to all of them). So far the public pilot proposals have not been published on Regulations.gov…at least I haven’t found them. As soon as they are, I will provide an overview of them.