search

logicahealth / vha-kbs-knarts

VHA Knowledge-Based Systems clinical knowledge artifact content development program.
Apache License 2.0
6 stars 3 forks source link

incomplete representation of concept in the KNART - family history example #116

Open catherinestaes opened 6 years ago

catherinestaes commented 6 years ago

https://github.com/preston/vha-kbs-knarts/blob/44bb353f460ce3502673efd2889a49b10c1df751/content/cardiology_ep_arrhythmias_devices/b36/20180507REVISED-CDSK_KRprt_CRDT_B36EP.xml#L1313

the KNART does not include the full representation of family history as documented in the TSR. the KNART is missing the fact that this is related to a family member, not the subject of the information, and it doesn't include the Observation procedure.

as an aside, I noticed this issue because I was wondering why the Validation Report only included one UUID in the Table 4.1. It is missing the UUID shown in columns AA and AB in the TSR (a997cc03-3e99-40eb-833a-6374c7750a3a).

**the KNART says this: - see 1313 in the KNART

Examples from the TSR: Family History of Sudden Unexplained Death subjectOf Information: 303071001 |Person in the family (person)| topic: [a997cc03-3e99-40eb-833a-6374c7750a3a |Observation procedure (procedure)]- ->(363702006 |Has focus (attribute))->[418362005 |Dead - sudden death (finding)]

Family History of Drowning subjectOf Information: 303071001 |Person in the family (person)| topic: [a997cc03-3e99-40eb-833a-6374c7750a3a |Observation procedure (procedure)]- ->(363702006 |Has focus (attribute))->[418362005 |Dead - sudden death (finding)]- ->(42752001 |Due to (attribute))->[40947009 |Drowning (event)]

csp4z commented 6 years ago

Assigning to @amsimms

catherinestaes commented 6 years ago

as i look at this further, i realize that this subelement has a 'responseDataType' of 'code' and is structured differently than other subelements.

csp4z commented 6 years ago

The KNART standard for doc templates does not provide a notion of persistence, it is only about collecting information. You can use it to define the questions, and tag options with codes. But unlike CreateActions, there is no actionSentence construct to map to an ANF statement. Thus, it simply does not make sense to use the ANF topic/or to build an ANF statement in a DocTemplate.

This is a valid result, meaning that if the KNART standard is to move forward it should have some mechanism of tying doc templates to some sort of model—much like the actionSentence concept for CreateActions.

Resolution summary

Action

kthlnkeating commented 6 years ago

Returning issue to PO.