In our article on the difference between data sharing and data availability, we emphasised the importance of data availability for the future of healthcare. But how can this data availability be achieved in a complex sector like healthcare? What are the challenges to be tackled and what can open platforms offer to achieve this?
The state of affairs
Data exchange and data availability are hot topics in the Dutch healthcare sector. When the Electronic Data Exchange in Healthcare Act (Wegiz) was introduced, Minister Kuipers of the Ministry of Health, Welfare and Sport emphasised, "Things have to change to maintain the quality of healthcare and to keep it accessible and affordable."
The attitude of more traditional technology providers regarding data exchange and data availability has not helped so far, however. Many parties, due to commercial interests, prefer to maintain complete control over how their systems can be linked and by whom. The closed nature of many healthcare information systems even led to vendor lock-in at hospitals.
However, this reality did lead to an increasing awareness of the need for a vendor-independent data layer as the foundation for healthcare information systems such as EHRs.To achieve this, healthcare information systems must be developed based on an open architecture and the use of open standards.
However, the practice of this is so far rather fragmented. Vendors currently still have the freedom to use different standards and practices (some of which complement and overlap each other). They can choose for themselves how they apply the concept of an open architecture and what degree of control they want to maintain on integration possibilities. In addition, several promising initiatives are being launched - consider, for example, Cumuluz's regional data hubs, the Babyconnect project for data sharing in birth care and the development of a standard for APIs in healthcare. To date, this often leads to combinations of data exchange and data availability, with data still being duplicated in practice. So important steps are being taken, but we are not there yet.
An open platform:
'An ecosystem of services in which healthcare data is modeled and stored according to a standard and accessible to different parties through a standardised set of API services.'
Open (healthcare data) platforms
In the words of Erik Vermeulen of EY, "If different forms of medical care, long-term care, social and community work, public health and other (local) government services need to work together effectively, it means that thousands of people in a region need access to data for various purposes to coordinate care and meet the health and social needs of residents. These collaborations benefit greatly from a new information architecture based on agile open health platforms."
When the linking possibilities of healthcare information systems are limited and no standard information models are used, this is a barrier to collaboration and innovation. When healthcare information systems are built on the basis of an open architecture in which such integration possibilities are central, from the perspective that within healthcare we are always dealing with a landscape of applications, we can arrive at an information architecture that is future-proof.
In addition, legacy systems must be considered at all times - this is the reality of the healthcare IT ecosystem. To effectively look forward, one must also look back to achieve a gradual transition from the current reality.
In an open architecture, the aforementioned vendor-independent data layer as a foundation for healthcare information systems is effectively an 'open platform'. We define an open platform as follows: 'an ecosystem of services in which healthcare data is modeled and stored according to a standard and accessible to different parties via a standardized set of API services'. This ecosystem contains in any case structured healthcare data, but can also contain a combination of services that may differ per open platform provider. The most important thing is obviously being able to read and write this healthcare data, but the range of services may also include, for example, authentication, authorization, terminology and/or imaging.
Whereas traditional healthcare information systems restrict linking with external applications (often from commercial interests), this linking capability is central to the open architecture in which the open platform and applications play their role.
Open standards
Open platforms are built based on open standards, so that healthcare data can be shared without unnecessary translation. We previously told you about our use of openEHR - this is one such standard focused on healthcare data storage, where there is an international definition of what data models in the medical field should contain, what terminology is used and what structure is needed. Standardising data models based on the expertise of healthcare professionals creates a healthcare information model that fits the practice, can be used consistently and truly supports the healthcare provider.
Federation
What might the future look like with such open platforms? In simple terms, the open platforms will then contain the data and can communicate with the applications of various healthcare systems through open APIs. This is also what is presented from the National Vision and Strategy drawn up from Nictiz, VWS, VZVZ and ZN as a "guiding principle" to realise the presented vision of the health information system: "Data is separate from functionality. The open platforms facilitate data availability and services that healthcare systems can invoke.
But, we might hear you think, then there will be multiple open platforms, each offering certain services? (Right!) What if those different open platforms then contain different patient data? Then you still won't have a complete picture of your patient, will you? The solution to this is seen in so-called 'federation': a collaboration layer between the independent open platforms and the healthcare systems that use the data. Because the data from the open platforms come together in this federated layer, this can still form an overall picture. The origin of data is fully available, including the necessary authorisations (data is only available with permission - the word "open" sometimes suggests otherwise, but this is so and will remain so). The federation of open platforms then also solves the vendor lock-in problem.
All this is certainly not a utopia - EY already has a federated sandbox environment running in which multiple open platforms from different vendors are working. For more information, check out the EY Connected Health Cloud.
Where are we now?
At the time of writing, there are plenty of developments - both in terms of the development of open platforms and the further development of technical standards. The industry seems to have come to an agreement on the destination, but the way there still contains question marks. But there is movement.
Certain vendors are also taking their role in this. Whereas traditional EHR vendors have a data foundation that is closely intertwined with the application, the more progressive vendors already separated this and the data layer and application layer are already linked through an API. The data layer can then already operate as an open platform, with the application of the EHR linked to it. There are still many hurdles to take, but we are moving in the right direction.
This article is part of the blog series "The Future of Data Availability in Healthcare," in which we discuss the possibilities of open platforms for data availability, the challenges of that solution direction, and CODE24's vision on this topic.
Also read: