OpenEHR. How to get started!?

I’m a GP in Thailand who interested in IT things and just get interested by the idea of openEHR.
Since I’m really new to this community and may be replicate this topic so I would like to apologize beforehand.

I just wondering how to get start OpenEHR and want to confirm something.

1.We don’t have to change database of EMR software that we already used.
Just mapping it with the library(or archetype) that already existed then we can use openEHR query(or AQL) to query the data in our database.

2.if 1.) is yes, then how to get started. My guess is something with CKM but I don’t actually know how to do it. it would be really useful if there are some videos or some manuals to do it.

3.if i’m wrong, please correct it. Thanks

1 Like

openEHR does involve another database, because it has its own information model. But this doesn’t mean you have to abandon your current EMR, over which you may have no choice. Usually AQL won’t work directly over a legacy EMR DB because the data in that system won’t be archetyped, so the AQL processor cannot find the semantic elements. Whether openEHR is an option for you depends on what you are trying to do - in a hospital or community environment, it could add value by aggregating patient-centric data over which you could run more powerful querying than you may already have.

In any case, we actually have one of the openEHR Specifications Editorial Committee (SEC) members and experienced implementers living in Thailand (Christian Chevalley), and it might be a good idea to contact him and you can discuss details that he will understand better than most of us (i.e. to do with how IT for GPs works there etc). If you message me directly, I’ll connect you to him.


Yes, it will be great if you can connect me with Christian.
Thank you so much for your help!

Hello Pi,

  1. openEHR has different levels that you can implement, for instance archetypes are a good way of documenting your system’s data requirements, and share that with others, without changing anything at the database level. But if your requirement is to use AQL, AQL is implemented as a layer over your DB, which needs to have certain features to make AQL work. If the requirement is not AQL but amybe a domain specific query mechanism (check this paper: it is possible to support openEHR data queries with minimal changes to the database, at least it will need the archetype ID and archetype path associated to each data point you want to query, since openEHR querying needs those elements. This is a small video showing that kind of simpler querying mechanism in the EHRServer
  1. The CKM is not related to how the data is represented in a database to support openEHR querying. CKM is just a repository and version manager for archetypes, shared internationally, and some countries have their own CKM. Also anyone can create a similar tool to store and manage archetypes, using the CKM is not mandatory.

  2. It depends on your specific requirements for querying, one strategy will lead to a lot of changes, and even needing to redesign the database or even change the technology, and other approaches might need minor modifications. Also those querying requirements might depend on how you plan to use the data. Do you have some ideas on that area?


1 Like

Hi Pablo
My hospital have an opensource EMR. it’s also used in many hospitals(around 50) in Thailand.
So I think if we can do modification, other 50 hospitals can also adopted it,too.

I like to start with minor change of database as much as possible and then test it. But I have no idea how to get start. All I know right now seem to be a big change which is not possible for me to do it.

I have an idea of NEWS/PEWS score which is a score calculate by patient vital sign and some other parameters to assess clinical severity. right now we have a google sheet to do it but if we can make an app it would be more convenient to use and record. I also can do it an easy way just to create new app but if it can do something which can contribute to openEHR, I will prefer that way


One option is to use your current software as a frontend, and adopt an openEHR backend. There are some open source options, the EHRServer (the one I created) and Ethercis/EHRBASE (I’m currently working on that one for a German project).

The points of contact with your EMR and the openEHR server would be sending documents using the openEHR REST API. This could be implemented using a middle like Mirth connect to minimize the integration code needed to send data from the EMR to the backend. Then you can use any wiring mechanism available on the server to get your data back. Part of my work is doing such integrations. Hope that helps.

1 Like


You can certainly use the openEHR models as pure design-time artefacts and then implement them in a traditional database but I would strongly suggest that you look instead at using an openEHR-enabled datastore (Clinical Data Repository) , perhaps sitting alongside your current database. Use it for new content/functionality e.g the NEWS score work.

There is actually an open source project in the UK which is working on NEWS, Sepsis etc that you might want to connect with - @David-Jobling can make the intros.

1 Like

That would be really nice.
I hope to dig in more

Such a topic is really helpful! Thanks