# automatic generation of templates **Category:** [Implementers (archive)](https://discourse.openehr.org/c/implementers-archive/158) **Created:** 2008-08-27 23:58 UTC **Views:** 4 **Replies:** 3 **URL:** https://discourse.openehr.org/t/automatic-generation-of-templates/14813 --- ## Post #1 by @Hugh_Leslie1 Hi Nancy We don't see the proliferation of Templates as an issue because they should be ALL based on a much smaller number of shared archetypes. That means that if two different labs use a slightly different request format (or even different departments in the same institution), as long as the underlying archetypes are the same, then the data created is shareable and means the same thing. This approach allows for the semantics to be rigidly captured, while giving the flexibility to allow for every clinical situation to be catered for. I don't know if its the same for you but here in Australia, every clinician wants to do things slightly differently! regards Hugh As for your questions about terminologies - yes. Where it is sensible, you can define small sets of terms in the archetype, however for the types of things that you are describing that will change depending on the institution, you would use a generic archetype and constrain the term set in a template. Remember that the templates can be saved and nested, within other templates, so you can make these changes once and then reuse it everywhere. NMazur wrote: --- ## Post #2 by @NMazur Hi Hugh, > Hi Nancy > > We don't see the proliferation of Templates as an issue because they should be ALL based on a much smaller number of shared archetypes. That means that if two different labs use a slightly different request format (or even different departments in the same institution), as long as the underlying archetypes are the same, then the data created is shareable and means the same thing. This approach allows for the semantics to be rigidly captured, while giving the flexibility to allow for every clinical situation to be catered for. > > I don't know if its the same for you but here in Australia, every clinician wants to do things slightly differently! yes, everybody wants to do it slightly differently, but how are you going to manage these differences... Is it really the objective that every clinician makes his/her own template? I guess the idea is that some templates are provided to the customer as a starting point? How do you manage the versioning? Every site, department then goes on making his own derived template (is there a form of inheritance allowed with the templates? I don't remember), yet the core template is being changed... Any ideas? > regards Hugh > > As for your questions about terminologies - yes. Where it is sensible, you can define small sets of terms in the archetype, however for the types of things that you are describing that will change depending on the institution, you would use a generic archetype and constrain the term set in a template. Remember that the templates can be saved and nested, within other templates, so you can make these changes once and then reuse it everywhere. I don't understand this. Ok for the use of the generic archetype. But how is the term set to be constrained in the template? Usually, that kind of data comes from a query in a database. The same thing with all the physicians active in a department. I guess I'm missing a link here between all the openEHR things, and the core data available of a hospital. regards, Nancy --- ## Post #3 by @thomas.beale Nancy Mazur wrote: > > Hi Hugh, > >     Hi Nancy > >     We don't see the proliferation of Templates as an issue because >     they should be ALL based on a much smaller number of shared >     archetypes\. That means that if two different labs use a slightly >     different request format \(or even different departments in the >     same institution\), as long as the underlying archetypes are the >     same, then the data created is shareable and means the same >     thing\. This approach allows for the semantics to be rigidly >     captured, while giving the flexibility to allow for every clinical >     situation to be catered for\. > >     I don't know if its the same for you but here in Australia, every >     clinician wants to do things slightly differently\! > > yes, everybody wants to do it slightly differently, but how are you > going to manage these differences\.\.\. > Is it really the objective that every clinician makes his/her own > template? no, Hugh didn't mean that literally \- but there is likely to be qite localised production of templates\. However \- there are already templates being standardised at a national and this will likely occur at regional levels as well \(depends on the country somewhat \- i\.e\. structure of health delivery\)\. Templates for some basic things like discharge summary will I think end up being standardised nationally \- this allows the generation of standardised message structures for use nationally, directly from the template\. > I guess the idea is that some templates are provided to the customer > as a starting point? How do you manage the versioning? templates are like any other authored resource \- they have to be versioned within a managed repository, which is the current direction wth archetypes and templates\. > Every site, department then goes on making his own derived template > \(is there a form of inheritance allowed with the templates? I don't > remember\), yet the core template is being changed\.\.\. Any ideas? this is handled \(jsut as for software\) with managed releases, i\.e\. collections of versions of templates and archetypes make up a particular release\. Subsequent changes are included in the next release, so that each release is itself a stable baseline\. There is long\-term experience with doing this for software\. > Ok for the use of the generic archetype\. > But how is the term set to be constrained in the template? Usually, > that kind of data comes from a query in a database\. Terminology subsets can be created and managed, and then referenced from a template\. E\.g\. the subset for 'bacterial and viral lung infections' or somesuch\. > The same thing with all the physicians active in a department\. I guess > I'm missing a link here between all the openEHR things, and the core > data available of a hospital\. do you mean core clinical data? \- thomas beale --- ## Post #4 by @system Just want to clarify - if systems can cope with the shared set of archetypes (all of them) then you do not need standardised templates for interoperability. But in the process of introducing standardised content definitions, the template provides a means of saying we will use this, this and this but not this or that. It also provides the means of building simple schemas based on these agreed templates for sharing information via the web. This provides an early and easy means of participation. Take your lab request for example - you can create the template as agreed by the group - then produce a schema which can then share the data (remembering that it can always be transformed to standard openEHR and thence to CEN or CDA. Cheers, Sam Thomas Beale wrote: [details="(attachments)"] ![OceanInformaticsl.JPG|183x82](upload://2lcnRHcC3QqDv6AeaDZuo8M9Qlv.jpeg) [/details] --- **Canonical:** https://discourse.openehr.org/t/automatic-generation-of-templates/14813 **Original content:** https://discourse.openehr.org/t/automatic-generation-of-templates/14813