It’s not often that you come across a company that so comprehensively bridges the gap between strategic data management and specialised patient care, but Ocean Informatics – trading as Ocean Health Systems – stands out by offering a robust foundation for openEHR usage through its Clinical Knowledge Manager, while also providing a highly specialised product – Multiprac – that showcases openEHR in action at the patient level.
This dual approach not only supports the standardisation of electronic health records but also enhances patient care through precise, actionable insights. By integrating these solutions, the company demonstrates the transformative potential of openEHR from a strategic and practical perspective, setting a new benchmark in healthcare technology.
We talked to Dr Sebastian Garde, CKM Product Lead, and Linden Bungey, Ocean’s Sales and Business Development Executive about Ocean’s origin story, and where the company is headed next…
Sebastian: It was really Dr Sam Heard who started me on this path a long time ago, I think around 2004 or 2005, when I was a post-doc and research associate with Professor Evelyn Hovenga at Central Queensland University (CQU). As an Adjunct Professor of CQU, Sam gave a big presentation on openEHR and how he saw it at that time. It was still such early days, with maybe only about 10 people who really believed in it – I may have been like the 11th person. We set up a research group in that area at CQU and also at Austin Health in Melbourne back then. That’s how it all began: that’s where we came up with the first ideas of what a Clinical Knowledge Manager would look like. We didn’t even have that term for it yet.
So, were you instrumental in building the CKM?
Sebastian: The problem we had at that time was that we essentially had zero visibility on what openEHR could really do, and the idea we were really looking for was a set of archetypes or clinical models that could be reused, ideally around the world. That was kind of the revolution – that you don’t have to do it once and then do it over and over again, but try to share these models. But we didn’t have the visibility to enable sharing. We had this Archetype Editor which Sam had hacked together, but not much more. So we started figuring out how to actually share them. If you want real interoperability, it’s not enough to just define archetypes. You have to share them and make sure they are reusable. That was one point, and the other was really giving visibility to the whole openEHR approach. That’s what changed a lot for openEHR, I think – making these archetypes public by just putting them out there.
Did you know you were on to something?
Sebastian: I had a very strong belief in this. And we were very happy. But it also was a hectic time. I had two babies in May 2008: CKM and my first-born son were both launched in the same month! That’s when we made it public and uploaded the first set of archetypes.
The first four people who officially created a CKM account are still involved in openEHR today: Dr Heather Leslie, who really bootstrapped the archetype editorial work, Dr Sam Heard, Thomas Beale and myself. Today CKM has registered users from 114 countries.
Heather would have run the review rounds for the archetypes. If I remember correctly, it took eight review rounds for the first archetype – the blood pressure one – to get published because we were really diligent, trying to figure out the process. We would do another review round for each missing full stop or something, which is an exaggeration, but you get the idea. Now, the editors typically do it in maybe two or three review rounds, depending on the situation and complexity.
But we started really iterating, iterating, iterating. That obviously doesn’t scale too well in the end. Initially, we started off with archetypes and added templates as the next step. We publish archetypes individually. But in the future, it may also make sense to say something like, “Here’s a release of all the approved archetypes we have in CKM; they are published, and we know they work together. This is an official openEHR release.”
That’s what release sets are doing, but they are more geared towards regional CKM customers like CatSalut in Catalonia or HighMed in Germany. CatSalut for example will be using these to create sets of archetypes and templates that all fit together, then use them in their systems. They are going all-in on openEHR, so they just plug them in and say, “This is what we use, it’s all consistent,” and that’s what CKM ensures.
Have you seen a lot of evolution in the CKM since 2008?
Sebastian: It has evolved quite significantly in many ways. You know, we had very lenient governance before; we had teams, but we had no projects, so it was really about collaboration. Now, we have projects and in general the governance has been getting a lot stricter. A lot of the initial work focused on how to get as much input as possible, get the reviews, and get discussions going.
But the more archetypes we publish, the more regional downstream CKMs we have, and the more we federate archetypes for reuse in national and regional programs, the better and stricter we need to get at governing them. This is so that in the Norwegian, CatSalut, HiGHmed or the Apperta UK CKM, they can just pull them in from the international CKM and for example create local templates on top of the internationally published archetypes, ensuring consistency.
So this all requires a lot more governance, especially for the published archetypes. And how to display this, there’s always this kind of – not a mismatch exactly – but there’s some kind of divide between the two essential user groups of CKM. One group just wants to see what’s there, what’s published, what’s actually approved, and then they use that stuff. They use the archetypes, use the templates. And then there are those who want to develop, edit, review and generally manage the archetypes and templates and need to be supported as much as possible. So, the goal is to combine both of these worlds in one tool, more or less.
What are you involved in now? You’re the product lead of the CKM – what does that look like, day-to-day?
Sebastian: A lot of that work is obviously mine, but over the years quite a few people have been involved in various roles, let me just name Dr Heather Leslie, Dr Ian McNicoll and Dr Sam Heard. The most important part is really figuring out what to do next with it. We’ve got a long backlog of ideas. But not all of that can be done. And getting all that together, getting the international community together with all the feature requests and ideas, and just being proactive in that area, which is often the case.
But also, then we not only have our international CKM, but also national or regional customers, they have their own feature requests, including a major Canadian hospital group that is using CKM in slightly different ways. And we need to ensure it all fits together. That’s also where Linden gets involved…
Linden: We do work very much as a team, and because CKM is the beast – it’s not particularly easy to understand. So we work very closely together. I deal with the business, and Sebastian works on the product details, as the product lead.
And what can you tell us about Multiprac? Is it your flagship product?
Linden: Yes and no. I think CKM is the flagship product for Europe and internationally, whereas Multiprac is very much an Australian product at this stage.
Multiprac Surveillance has become the go-to healthcare-associated infection (HAI) surveillance and management solution, originally tailored specifically for the Australian market.
We developed Multiprac right here in Australia to align with the nation’s infection control standards. But the system can absolutely be adapted for other countries’ standards. There’s no reason the language can’t be modified as well, so it’s a very flexible product that could be used internationally. We’ve already had inquiries from other countries who are keen to get on board. So while Multiprac started out as an Aussie innovation, we would love to take it global.
The numbers speak for themselves: Multiprac is deployed across over 130 public hospital sites in Queensland alone, managing a massive 150 million patient records on the openEHR platform. That’s one of the largest openEHR solutions on the market. They collect the same data types and generate scheduled reports similarly. A key benefit of openEHR is that the data is well-contained and linked.
Multiprac isn’t just big, it’s smart too. It pulls in data from multiple streams like pathology via HL7, patient admissions from the hospitals’ PAS systems, surgical procedures… Using rules-based surveillance, it automatically generates notifications for our infection control practitioner (ICP) teams to review each morning on their dashboards.
This streamlined approach is a total game-changer compared to the old paper-based methods. It saves our ICPs hours of tedious work, which means they can focus on their patients.
We now have a new reporting system called Advanced Data Reporting (ADR) that allows the set up of cyclical reporting. Reports can be tailored for the executives or the ICPs. And if someone’s away on vacation, for example, they can set their reports – their replacement just has to press a button, and it’s done.
Multiprac has the ability for expansion and development to also track antibiotic resistance, occupational exposure incidents, contact tracing – you name it. This data aligns with the Australian infection prevention and control guidelines and it integrates with BI tools for flexible visualisation and analysis across single facilities or massive multi-site groups. That’s the true power of standardised, interoperable data.
Apart from Multiprac and the CKM, what else are you working on?
Linden: Within Multiprac, we’re actually looking at expansion, with new products forming a suite of Multiprac products. One of the products that we almost have ready for demo focuses on staff health, so immunisation and vaccinations. It will be called Multiprac Immunize. The current Multiprac on-premises or on the open-air platform does have a staff health module, but we’ve expanded it, and it will be a standalone product.
We’ve got numerous healthcare groups interested in this because there really isn’t much out there that will record and have the capabilities that this product is going to have. So that’s one product, and we’re also looking at a couple of others. One is in the outbreak area, where we’re looking at being able to quickly pick up an outbreak that occurs with, for example, measles, rubella, COVID, within a hospital or a community area. That would be a module that works on that area, called Multiprac Outbreak.
Another one we’re working on is called Multiprac Stewardship, which is looking at Antimicrobial Stewardship (AMS) or antimicrobial resistance. It’s potentially bringing in some of the genomics, and that information that comes through on the HL7 messages from pathology to look more specifically at that area.
Finally, Multiprac Navigator – one which is going to look at more, I suppose, the management side of things with groups. There is a lot of importance in adherence to guidelines, and a lot of international and national groups are collecting data, and the hospitals have to meet that criteria and are assessed on that.
So this is a tool that might help them in that area. It’s pretty much a shell which works on education; collecting the data and ensuring that the staff are educated correctly. This applies a cyclical model for education and adherence to guidelines by healthcare facilities.
Sebastian, what’s next for the CKM?
We’re just preparing the next big release actually, with a lot of new or adapted functionality – that will be ready in June 2024.
One area where we need to do a lot of work is ADL2 – the Archetype Definition Language version 2. Sooner or later we will have to focus on migrating everything from ADL to ADL2. which will involve migrating all the archetypes.
This obviously has a huge impact on CKM as the repository where we keep all the data, all revisions of archetypes from 2008 until now. We have to ensure we have the whole revision history so anything can still be reproduced. So, upgrading all that is actually a major task for CKM’s future.
For the next release, we’ll be starting to enable more capabilities around archetypes and templates to communicate with other standards and platforms. Not everything is openEHR, but there’s a lot of FHIR and OMOP. So we’ll integrate mappings to FHIR and OMOP into CKM. We’ll also move towards being able to add typical exemplar queries expressed in the Archetype Query Language (AQL) and upload them to CKM.
Of course, you have Chunian Ma at Ocean. She’s like the AQL maestro, right?
Yes, she presented AQL at Medinfo 2007 in Brisbane, and it was probably the most fantastic presentation I’ve ever seen there. Everybody was standing and applauding at the end – it was really impressive.
She’s doing very well with AQL. And the more common archetypes you use, the more important it becomes to be able to not only share the archetypes but also share these queries. So that’s what we are picking up on with this release.
The same goes for the mappings – if you want to map to FHIR, we want to make that as easy and consistent as possible. Grahame Grieve and the team are talking a lot about how we can collaborate on this and other areas. This is at least one step we can take for the next CKM release.
From there, we’ll see where this goes. There’s also ADL2, as I mentioned, and making the review process smoother. There will be work in many areas. But I think these kinds of secondary models such as mappings, queries and maybe subset models, attaching them to archetypes and templates, making sure it’s all consistent – this will be one important future direction. I think this will be very important for the future of openEHR.
Leave a Reply