Continue discussion on ANF modeling of CDS KNART terminology artifacts.
Discussion items
Time
Item
Who
Notes
60min
ANF Modeling
Group
Keith: The goal is to wrap-up the ANF document and Stephanie has done work to pull it together. For definitions of phenomenon, he said to use what we have for observation. The goal is to get this into DocBook reasonably soon. He would like for Claude to include some diagrams in this document as well. Claude: He can add the diagrams after Stephanie is finished editing the document. We need to ensure all terms are defined.
Claude: "Terminology Model" really refers to the concept model, which is a model defined for terminology, whereas a terminology expression is an instance for that model. He will write a draft paragraph that summarizes this and can be included in the document.
Keith: Regarding observations and findings and "is about," there is a circular problem and we want to avoid it. John: It's important to be clear what we're pointing to, such as this thing is about X.
John: Doesn't see a distinction/boundary between topic and circumstances and how much of a clinical statement goes into each. Keith: If we can't make distinctions, we should not be making them. Each distinction must have a foundation - is there some reason we think we need to make it, and if yes, is the distinction reproducible. Part of the usefulness is around representation and retrieval of data. Putting everything into free text (i.e., clinical statements = paragraphs) and should we just leave it at that? John: We can represent a subset of what free text is, and free text is powerful. Keith: For example, a clinical reminder rule, such as what the KNART project is all about. Do you buy into that project? John: He has a different level of belief for formalism to be usable and obtainable. He doesn't know if the KNARTs are entirely usable and maintainable. Keith: If you're willing to suspend disbelief momentarily, a processable clinical statement is required to go to the next step. If we can't get clinical statements that we can write a rule on, then the whole process is not attainable. John: If we set our site down, it's how much we want to express in a clinical statement. Keith: You've heard me talk in the past that there is a group that likes to build a skyscraper starting at the third floor with no foundation. We are trying not to do that and instead start with the foundation. The distinction between topic and circumstance is in part due to the reality of the notion that we would put all of the information in the post-coordinated blog into the topic (missed the rest of what he said). John: Putting it into one blog wouldn't work, and we need to figure out what to leave out. Keith: There is utility in separating what goes into the topic vs. circumstances. As Stephanie pointed out, she is still struggling because we don't have good definitions. He'd like for us to focus on making definitions clean and clear so they are understandable and for John to be an appropriate critic for them. We want your ability to say here is an example that the discipline is supposed to be use case-driven and architecture-centric. For this document, we need to say is the definition of a clinical statement reproducibile and if yes, then what are the 4 types of clinical statements. That's your first step in looking at this document critically. We tried to pick phenomenon because it has less baggage than condition, observation, finding, etc.
Stephanie: Has concerns about the name of "Phenomenon Measurement" and prefers "Phenomenon" instead since measurement implies we are measuring something.
Keith: He would like for John to take everything that is a definition and which is a criterion for defining one aspect from another and provide examples of where he thinks a clinical statement doesn't fit into 1 of the 4 types of clinical statements. Also, sufficiently describe how we have not made the distinction between topic and circumstances.
Stephanie: Asked if John would forward her his comments on the TSR spreadsheet since he indicated during today's call that he had provided them but she doesn't have them and did not see them annotated on the TSR spreadsheet in Google.docs.
Keith: He would like for Stephanie to show the REDCap survey instrument at our IA Small Group call on Tuesday.
Catherine: She can work with Stephanie on the user perspective. Keith: He would like to know what the impact is on the analysis if we allow the user to copy/paste the part of clinical statement that represents the topic and circumstances. Catherine: She and Stephanie will schedule time next week to meet.
Claude: He is here next week but then out until 10/17.
Action items
John Kilbourne: Provide his comments on the ANF modeling document.
John Kilbourne: Forward to Stephanie his comments on the TSR spreadsheet.
Add Comment