You can read the Dutch version here
In recent months, the term "data availability" has become increasingly prominent in various discussions and forums. Data availability is often presented as the future principle that will enable efficient (re)use of health data. The intended benefit is that more valuable information can be extracted from "data," thereby improving healthcare and keeping it affordable.
When examining the essence of data availability, it involves the separation of data and its use within an application. This may seem obvious, but in practice, it is not so straightforward, especially since the majority of information systems in healthcare were not developed based on this principle. Often, well-known Electronic Patient Record (EPR) and Electronic Health Record (EHR) systems are cited as examples, but it applies to almost all systems with a "client-server" architecture.
Since the 2000s, a gradual change has taken place. The explosive growth of internet usage led to new insights, and technologies were developed to simplify data access. Service Oriented Architectures became gradually prevalent. The HL7v3 standard and the IHE XDS profile emerged from this period. Both are based on the principle that data is stored "at the source" and centrally indexed, allowing others to find and retrieve it.
While the IHE Cross-Enterprise Document Sharing (XDS) profile essentially focuses on making data available, it initially aimed at the trend of regional collaboration between healthcare organizations. This resulted in the need to exchange (patient) information, often within the context of an existing "referral pattern."
The implementation of the first XDS-based Health Information Exchange (HIE) networks, where all shared patient information was included in a central index, revealed practical challenges hindering data availability. Patient consent, agreements regarding data security and ownership, legal regulations, limited the exchange of indexed information.
Fast forward to the present. In the Netherlands, there is relatively high adoption of data exchange based on IHE XDS and/or XCA profiles. Out of approximately 70 hospitals, 58 somehow utilize XDS. This provides a foundational "national" infrastructure based on the principle of data availability.
In the Netherlands, critical minds in the healthcare field claim that XDS is not suitable for building a national infrastructure for data availability. The facts tell a different story. While various discussions about building data platforms are ongoing, all XDS networks in the Netherlands collectively have indexed over 180 million, and thus available, documents. Of these, 61 million are radiological (DICOM) studies. The rest include various forms of medical documents such as lab results (56 million), radiology and other reports (46 million), and patient consents (11 million).
Suppliers such as Founda/Forcare, Enovation, Epic, Nexus, Chipsoft, Openline, and others are linking their platforms together at the request, and in collaboration with their shared clients to exchange available documents (medical images, reports, lab results, medical letters, and patient summaries).
Unfortunately, the reality is also that a large portion of the 180 million documents are not used to support healthcare processes due to a lack of patient consent, or a collaboration agreement between healthcare institutions. The "sour" fact is that despite this massive amount of available data, limited value is being created.
To harness this potential, the healthcare sector needs to take the lead. While suppliers also play a role, it is essential to remove the barriers hindering data availability.
And now...
In concrete terms, I call for continuing on the current path. Instead of investing hundreds of millions of euros in designing and building new solutions, we can achieve faster progress for lower costs.
Start thinking in terms of data availability rather than data exchange, and generate value by removing the obstacles blocking data availability. Invest in connecting existing XDS infrastructures to achieve national coverage. Properly manage the registration of patient consent, or contribute to the success of MITZ. Massively enter into individual or regional collaboration agreements, or embrace the nationally covering TWIIN collaboration agreement. And don't forget to involve the "workforce," invest in change management, and adapt existing work processes.
This way, we can fulfil the (political) desire to achieve national data availability in the foreseeable future, give substance to the goals of the Integral Care Agreement (IZA), and achieve results much faster.