I’m participating in a project owned by the Norwegian Directorate of Health, to publish a small common dataset for a smoking summary. The Tobacco smoking summary archetype is one of the sources informing the project.
One difficult question which has been discussed in the project has been whether the dataset needs to include a repeatable group of data elements for describing an “Episode” or “Period” of smoking behaviour. The Tobacco smoking summary has such a group/cluster, and examples given for its use include “during pregnancy” and “while on holiday”. As part of the project discussion, I’ve been tasked with finding real world, practical examples where the “Episode” group of elements are required. I’ve looked through several recording contexts common in Norway, but haven’t found any examples where the element group is definitely required.
This is why I’m turning to the openEHR community to ask; do you know of any real world, practical examples where you are using or would need to use this kind of repeatable group of elements to detail specific episodes of smoking behaviour? If so, could you point me towards them, preferably with online documentation?
Not sure about any documentation, but it is common to ask women about smoking behaviour during pregnancy in terms of risk to the fetus. This was one of the main drivers in the archetype design.
Yep, I remember this was the background. But do we have any concrete examples of use cases where this is used?
I’ve looked at a lot of Norwegian use cases, but I haven’t found any concrete examples which need the episode structure. So I’m looking for international examples
I think there will be a use case for the Episode Cluster, in the EVALUATIONS for tobacco, substances and alcohol use, but we haven’t yet taken these archetypes in use in highly specialised departments/wards. Especially the ones treating drug and/or alcohol addiction. In addition to pregnancy, we can assume there will be a use case to record periods in an individuals life where they were heavy smokers, different from today’s behaviour.
I think the only time that we have used the per-episode is actually to label it as ‘life-long’ when the client need access to elements that were not available at the per-type of the archetype .
e.g. ’ Pattern’ which is only available per-episode.
I’m not sure if this is a hack or brilliant but it might be worth considering added as a fixed event !!
In Finland, the national data model for tobacco use (Termeta - Tietokomponentit - Tupakka- ja nikotiinituotteiden käyttö) includes data points like “Average use per day during the entire period of use” and “Number of tobacco products per day before quitting”, so we suggested using the Episode label to differentiate between current, life-long and “before quitting” tobacco use details. Interestingly, the Finnish model contains two items relating to pregancy - “Duration of regular use of tobacco or nicotine products before pregnancy” and “Duration of pregnancy after stopping the use of tobacco or nicotine products” - which we felt may be better covered by the Substance use screening questionnaire archetype. We’ve not seen how the national model is actually implemented in practice though so can’t point to any example forms or documentation unfortunately.
For most day to day smoking cessation management, for example in general practice, that level of detail is excessive but for specialist clinics and perhaps researchers I suspect that the episodic detail could be needed.
That said I have had a dig around and failed to come up with anything that would map to this pattern in real life!
This for example - a pharmacy smoking cessation service - needs only the history of the current episode, and not that level of detail.
The survey questions from here however, on brief inspection, would make use of the Episode pattern but only really to record the details of quit attempts for the current episode, not specifically past episodes.
Regardless it looks to me that the modelling approach is the correct.
If you want to record even one episode, that being the current one (which may be the only one) then you can do this as the archetype is currently designed.
OTOH, I have not really found any IRL examples of multiple episodes.
Interestingly, this is a mix of two concepts. The health risks assosiated with smoking tobacco are quite different from other routes of nicotine, so already there is an opportunity to improve, in case anyone working with the national data model in Finland. A qustionnaire is probably the right solution in this case, and the new, very soon to be published, “Substance use screening questionnaire”, should be the one to use.
