Eindhoven Region XDS statistics

As I have shared some aggregated statistics from the national Dutch XDSCloud service in past articles, I will dive into the details of one of the regions in the Netherlands in this one. This region includes 4 hospitals in and around Eindhoven. An area also referred to as the "Brainport of the Netherlands" as it hosts quite a number of innovative companies and startups. 

Another interesting fact is that the hospitals in this region were the first in the Netherlands to start using the XDSCloud service. As true pioneers they decided to no longer host an XDS infrastructure themselves. Instead they chose to purchase "XDS as a service". 

The Health Information Exchange organization in the region is called "RZCC", where 4 hospitals (St. Anna Hospital, Catharina Hospital, Máxima Medical Center and Elkerliek Hospital) and one Diagnostic Center (Diagnostiek4U) participate and share both diagnostic imaging, as well as laboratory data. As such each hospital has connected their EMR, RIS and PACS to the XDS. What further makes regions stand out from others is that the hospitals also publish functional test results, and ECGs. 

Since the Diagnostic Center has its own XDS infrastructure the metrics in this article only relate to the 4 hospitals.

Sharing radiology procedures

As many of the Dutch hospitals do, the 'Eindhoven' hospital also started with "image sharing" as their initial and primary use-case. The graph below shows the growing DICOM study exchange volume since the network started in 2016.

Initially the network had a bit of a "slow" start due to the fact that day to day operational procedures had to be adjusted to include using the XDS network for image sharing opposed to more traditional ways of exchanging imaging data through CDs and DVDs. 

As of 2019 the exchange volumes started to increase somewhat stabilizing to the current day volumes. 

Now let's dive into the KPIs as introduced in previous articles. 

Document Volume and type distribution

Catharina Hospital

Elkerliek Hospital

St. Anna Hospital

Máxima Medical Center

Discussion

The document volume numbers in the above graphs are representative for the different sizes of the 4 hospitals. The document distribution shows that in all hospitals the majority of the registered XDS documents related to diagnostic imaging studies are in line with the "image sharing" use-case the network initially was designed for. The graphs also emphasize the fact that it is only St. Anna hospital that registers laboratory documents. The use-case is for this is data sharing with pharmacies in the area.

Total number of exchanged DICOM Studies

Catharina Hospital

Elkerliek Hospital

St. Anna Hospital

Máxima Medical Center

Discussion

The graphs are a reflection of the existing referral patterns in the region. As can be derived from the "SUM of Success provide DICOM Study" graphs both Máxima Medical Center and St. Anna Hospitals roughly provide twice the number of studies compared to the other two hospitals. This is explained by the fact that the (patient) referral direction is mostly to Catharina Hospital as can be concluded from the "donut" graphs. As these graphs only analyze DICOM study exchanges no referral patterns of laboratory information exchanges are visible. The donut graph of Máxima Medical Center also shows "RNZN" as a referral destination. This used to be a neighboring network that ceased to exist.

While discussing these outcomes with the hospitals it was mentioned that not all hospitals consistently request patient consent. This is becomes clear when looking at the next KPI. 

Patient Volumes

Catharina Hospital

Elkerliek Hospital

St. Anna Hospital

Máxima Medical Center

Discussion

The above graphs show the number of registered patients by their National Patient Identifier (BNS) and their Local Patient Identifier (MRN). The National Identifier is unique for each Dutch citizen. Given that for each unique "BSN" only one "patient consent (BPPC)" document is registered the product of both give an indication how many patient consents have been registered. Although this article doesn't seek to explain these differences it is important to realize that the fact that "data is available" does not automatically mean that "data is available for meaningful use". Given that Dutch privacy laws require a patient's consent before his/her data can be used it is fair to say that the absence of patient consent presents an obstacle to "data availability". 

As previous articles have presented insight into the large volume of registered XDS documents in the national Dutch XDSCloud infrastructure, taking away the obstacles bears an enormous potential for "meaningful use" for relatively low efforts and costs.

Acknowledgements

I sincerely want to thank Catharina Hospital, Elkerliek Hospital, St. Anna Hospital, Máxima Medical Center, and the regional HIE organization RZCC for their consent to share the numbers presented in this article. Their transparency is of great importance to demystify standards-based health information exchange. It helps identify opportunities for improving IHE XDS-based exchanges, and allows for a transition from a "data exchange" to a "data availability" paradigm.