Thanks Melvin - you may want to pass this around…
Tom Seabury has raised comments that are very thoughtful - I have included some of his as well.
Melvin’s collation of commentsThe descriptionThe description at present reads:
“the measurement of systemic arterial blood pressure which is deemed to represent the actual systemic blood pressure”
The comment reads:
"Horribly inaccurate!
Maybe: “the non-invasive measurement of systemic arterial blood pressure which is deemed to represent the actual arterial blood pressure at the same body site” – or as a generalised representation of systemic BP?"
First, the reading is a surrogate for systemic blood pressure and does not assume non-invasive (it does not matter if the reading is invasive or non-invasive if it is a measure of systemic arterial blood pressure (an observable entity in SNOMED by the way). It is not specific to site - hence site is in the protocol and is not part of the data. I am not sure what a “generalised representation of systemic BP” means. I think systemic arterial blood pressure is a known clinical entity.
My suggestion is that we add "measurement by any appropriate means (invasive or non-invasive) of " to sort out the problem of invasive or non-invasive.
Protocol/DeviceDevices can easily record the device type - location of measurement is not important for clinical interpretation when it is systemic pressure that is being measured - but people could be asked to record this if it was deemed appropriate in the setting.
Default and Assumed valuesArchetypes do not specify default values as these are often setting specific. Archetypes are the ontological level of specification of what it is sensible to record. Default values (and other use dependent constraints) are set in templates rather than in the archetype. People not familiar with openEHR often want to express everything in the archetype - but it is important to keep these as generic structures with high reuse.
In the OBSERVATION class it is possible to add data elements that are deemed to be ‘state’. The definition of this is that this data needs to be recorded to allow appropriate clinical interpretation of the data. Clearly there is no limit potentially to this set - but generally is set at variables that would not otherwise be in the health record. Examples are laboratory measurements following challenges, fasting or random for blood sugars, the position and exercise level for blood pressure. Here, the archetype will have an assumed value for each state variable (as it will usually not be recorded). This allows the state model for different observations to grow with the archetype but not to place demands on clinical applications to make these variables available unless they are relevant in their setting. It also makes explicit what is presently taken for granted.