The Dutch healthcare system is a distributed network of collaborating healthcare institutions. But for years, attempts have been made to squeeze digital data exchange in healthcare into a single centralised system, without much success. If we want to align healthcare digitally, we will have to adopt a decentralised communication network.
Fragmentation by design
In its report ‘Toward an integrated health information system in the Netherlands’, published this year, the OECD writes: [The Dutch healthcare sector] is built on principles of decentralisation, competition, market mechanisms, and the private (not-for profit) sector to deliver quality care to its citizens. […] in contrast to other health systems where it is a result of either legacy factors or unintended policy consequences, fragmentation in the Netherlands is a design feature. The Dutch healthcare landscape thus (deliberately) consists of separate organisations each making their own contribution to good healthcare and much of that healthcare is distributed across multiple healthcare institutions. It is a complex dialogue between specialists, nurses, GPs, pharmacists, patients, informal care providers and other stakeholders. All of these parties have their own IT systems because the market for IT support in healthcare is fragmented as well.
To facilitate this complex dialogue, systems have been introduced for submitting data about a patient or client from one healthcare provider to the other. But there is no uniform way of exchanging data between these systems. As a result, each region has its own method, being “connected” tends to cost a lot of money and even then a lot of cutting and pasting between systems is involved. This often has the unintended consequence of inhibiting communication instead of stimulating it; either because it is too cumbersome for the healthcare provider or because they are charged per message. This is a shame, because although the healthcare industry remains deliberately fragmented, the profile of the patient shouldn’t be. The fiercely competitive open market for healthcare IT systems is a good thing, but it should not come at the expense of information exchange and therefore of healthcare quality.
A fitting solution
In the past decades, attempts have been made to solve the issue of safe and reliable data exchange in a fragmented healthcare landscape with centrally managed facilities. This began thirteen years ago with the introduction of the “national EPD”, which was rejected by the Senate in 2011. The idea was: if we create a centralised system where data is available to each connected healthcare provider, everyone regains a full picture of the patient. A noble ambition, but also one that ignores the fact that distributed data contributes to the patient’s privacy. Since that time, various ideas have been put forth, each offering a solution to the exchange problem to a greater or lesser extent. But all these ideas shared a common denominator: a centralised body that administers a system for exchanging all data in the Netherlands. Whether that body would be a commercial supplier, a foundation, an association or the government itself, the recipe remains roughly the same.
After thirteen years of going back and forth, we need to accept that this type of centrally administered facility is not an appropriate, all-encompassing solution. We need a different system architecture to support the wealth of use cases for data exchange in a way that balances the availability of data with the privacy of the client. One that is a better match for the decentralised Dutch healthcare system. For several years, we have been advocating a technical data exchange model in healthcare that aligns with this fact: a decentralised communication network.
The right example
What is a decentralised communication network? It is best defined by what it is not. In a decentralised communication network, information is not exchanged via a single hub. Data is not stored in a single location. No single party gets to decide what happens on the network.
The Internet is an example of a decentralised network. The Internet is a collection of standards and protocols that can be used on all layers by all parties (from fibreglass or copper in your home to the website that you want to visit). Need a new router? There are thousands of models made by hundreds of different suppliers to choose from. Internet subscription? Fibreglass, cable or telephone line? Or will you go for 5G or satellite? Every option comes with a wealth of suppliers. Will you go for MS Edge, Google Chrome or Apple’s Safari?
Email is another good example. Back in the eighties, almost nobody in healthcare had heard of email but there were various commercial closed “email” service providers that were unable to communicate with each other. Eventually, the SMTP standard (Simple Mail Transfer Protocol) came out on top, which is why today we get to choose from a wide range of email providers, all able to exchange emails with each other. And if you don’t like any of them (or your organisation is large enough), you always have the option of setting up your own email server.
Similarly, a decentralised communication network for healthcare would be a set of open standards that can be implemented by various suppliers. A type of “SMTP” protocol for healthcare. Specialising in the targeted exchange of structured medical data.
Open standard collaboration
We raised our flag in September 2020. For the first time, GP group ZGWA was able to view the patient files of home care institution Aafje directly and in real-time. The GPs were using the collaboration platform cBoards by CareSharing while the home care institution was using the Ons software package by Nedap. This was possible thanks to the decentralised network we developed combined with available open standards (such as ZIBs and FHIR).
Since that time, we have increased our focus on exchange. We organised a hackathon last September with seven IT suppliers specialised in long-term care and hospital care. They helped us build (a basic first version of) nurse handover into our systems, based on the same decentralised network. We are also talking to BabyConnect for birth care and discussing the possibility of organising evening, night and weekend care between long-term care institutions. In addition, we are involved in the KIK-V programme to start using the same open standards everywhere. We call these applications of the Nuts network “Bolts” because nuts are pointless without a bolt.
Because we employ a decentralised architecture, we can scale to different use cases with our network quickly. The decentralised character forces us to operate on the basis of openness: everyone can contribute to the standards, software and process descriptions in the Bolts. We formulated our principles in a transparent manifesto consisting of eight points and whoever feels compelled can join and help accelerate the initiative.
As the number of successful practical examples grows, so does the need for products and services that meet the Nuts principles among healthcare organisations. Back when nobody had heard of email, nobody was asking for an email service that was able to reach other providers. We are now witnessing how healthcare organisations begin to set requirements to their suppliers and ask more and more pungent questions about openness and interoperability. This makes it appealing for both large and small suppliers to increase their collaboration and catch up, which will lead to a technical network that once again reflects how healthcare is organised.
Digital collaboration in healthcare is becoming increasingly important and the need for systems that enable collaboration while safeguarding the patients’ privacy is growing as a result. The technology for decentralised, secure information exchange is available. Let us work together to build on it!
