JMIR Publications

ABSTRACT

Background:

Over the past 12 years the international openEHR community has progressively created, peer-reviewed and published the content of multilingual archetypes within an online tool, the Clinical Knowledge Manager. The CKM currently holds a library of more than 500 archetypes but only a small number of templates which have been shared as examples of possible datasets for implementation. In response to the COVID-19 pandemic, several implementers within the openEHR community openly shared templates representing their data set specifications on the CKM. It is the first significant opportunity to investigate modelling approaches and reuse of archetypes across templates by a variety of authors and for a range of clinical use cases.

Objective:

To investigate archetype use and reuse within seven openEHR templates authored by a different modellers for a range of clinical purposes developed in response to the COVID-19 pandemic.

Methods:

Analysis of both the number of occurrences and patterns of occurrence within seven discrete openEHR templates was carried out at the archetype, or clinical concept, level.

Results:

Whilst the trigger for developing the templates was in response to the COVID-19 pandemic, the data sets represented a varied cross section of clinical data requirements including clinical screening, confirmed infection reports, clinical decision support activities and a research data registry: - 231 instances of 59 unique archetypes were used across 7 templates - the most reused clinical archetype occurred 24 times across 4 templates - 2 archetypes were reused across 5 templates; 5 archetypes were reused across 4 templates - rates of archetype reuse varied from 52% to 100% in each template - archetype reuse of 100% occurred in 4 of the 7 templates - templates were deployed in English, Norwegian, Chinese and Italian.

Conclusions:

Despite the variation in size, complexity, scope and language of implementation there is evidence of significant reuse of archetypes across the 7 templates, ranging from 52% to 100%. The increase in reuse in later templates to 80-100% was driven by strategic development of purpose-agnostic archetypes to represent novel clinical concepts. The new archetypes were reused extensively in the later templates and demonstrates the benefits of an agile, strategic archetype development approach. Reuse of high quality, peer reviewed information models in data sets has enormous potential to improve health data quality and consistency.