We held our first Relevant and Pertinent meeting in several weeks due to some scheduling challenges, which resulted in some interesting thinking about what our output is going to look like. We've stopped focusing so much on whether a decision that something is relevant affects what is done in the send or on the receive side. It's a judgement, an assessment, a category that can help decide how to do something innovative to make information more useful.
I think people will argue less about the result because they will be less worried about unanticipated consequences. Making an assessment about the relevance of a thing becomes a point of information that can be acted upon within an exchange, but it need not prohibit information from being exchanged, we leave that in the hands of the users of that assessment. In some ways, it's very much similar to how saying "this information is sensitive" helps make access control decisions. The assessment of sensitivity can be agreed upon if we can separate that assessment from access control consequences that we might not agree with.
In drilling down into this, we came up against one of my first deep discussions on problem classifications. I was looking at the "Problem Status" vocabulary of CCR at the time: Active, Inactive, Chronic, Intermittent, Recurrent, Rule Out, Ruled Out, Resolved.
What a basket of stuff, crossing at least three different classification boundaries. Since then, we have probably all come to understand acuity, activity, and temporality as being three distinct axes of classification.
An acute condition is typically active or resolved. A chronic condition is either active, inactive (I think I prefer managed), or in some few cases, resolved. Chronic lower back pain is something that I suffered through for about a 12 month period, but it did eventually get resolved. Other cases (specifically dealing with chronic disease states), rarely occur, but have been known -- often due to radical intervention.
This then gets into the question of relevance. What is relevant about my chronic lower back pain, vs. my acute lower back pain, and how could one automate distinctions. Let me propose that the following is a true statement:
What is relevant about a condition is all data related to that condition over the history where it has been present. The data relevant to the acute lower back pain covers the period from which the problem was discovered to when it was last active and shortly thereafter. The same is true for the chronic lower back pain, it's just that it's "active" history might be much longer. Now, if you link diagnosis and treatment to problems (in a problem oriented medical record), you can now determine what treatments and diagnostics are perhaps "relevant" when considering lower back pain, regardless of chronicity/acutity.
I think we'll next need to get into a discussion of current, recent and historical. Current is thought of in the now, but is actually in the very short term past, as in just happened. Recent is a slightly longer window, perhaps measured in the context appropriate to the situation. When one asks what my recent immunizations are, the context is a little bit different from when was my most recent blood glucose. Both are thinking about "most recent last", but not just now events. The time context window is different for one rather than the other. If you wanted to know when my most recent TDAP was, you might look back 5 years before saying, "That's far enough! You need a new one," whereas for blood glucose, you might go back as far as a year. In both cases, anything before those dates is "historical" because it doesn't matter how far back after that time frame, its relevance is about equal (time for a new one).
The locus of focus may also affect our perception of relevance. At the time of a visit, attention is on the condition that the patient is being treated for, the chief complaint or reason for visit. For another whose attention is on that visit (locus), their attention (focus) may be elsewhere. In communicating forward in time we usually assume that our focus will match that of the receiver, and so would determine relevance based on that focus. We often discover that what appeared to be inconsequential then is more important now.
We have no way to predict the future, which results in the fear that a determination of relevance could prevent something that could/should have been communicated. If, however, we can, as I would, put a protective force bubble around "what to do with the determination" as being subject to use-case appropriate judgement, we might come to dramatic consensus on what is "relevant" as best we are able to determine.
Keith
I think people will argue less about the result because they will be less worried about unanticipated consequences. Making an assessment about the relevance of a thing becomes a point of information that can be acted upon within an exchange, but it need not prohibit information from being exchanged, we leave that in the hands of the users of that assessment. In some ways, it's very much similar to how saying "this information is sensitive" helps make access control decisions. The assessment of sensitivity can be agreed upon if we can separate that assessment from access control consequences that we might not agree with.
In drilling down into this, we came up against one of my first deep discussions on problem classifications. I was looking at the "Problem Status" vocabulary of CCR at the time: Active, Inactive, Chronic, Intermittent, Recurrent, Rule Out, Ruled Out, Resolved.
What a basket of stuff, crossing at least three different classification boundaries. Since then, we have probably all come to understand acuity, activity, and temporality as being three distinct axes of classification.
An acute condition is typically active or resolved. A chronic condition is either active, inactive (I think I prefer managed), or in some few cases, resolved. Chronic lower back pain is something that I suffered through for about a 12 month period, but it did eventually get resolved. Other cases (specifically dealing with chronic disease states), rarely occur, but have been known -- often due to radical intervention.
This then gets into the question of relevance. What is relevant about my chronic lower back pain, vs. my acute lower back pain, and how could one automate distinctions. Let me propose that the following is a true statement:
What is relevant about a condition is all data related to that condition over the history where it has been present. The data relevant to the acute lower back pain covers the period from which the problem was discovered to when it was last active and shortly thereafter. The same is true for the chronic lower back pain, it's just that it's "active" history might be much longer. Now, if you link diagnosis and treatment to problems (in a problem oriented medical record), you can now determine what treatments and diagnostics are perhaps "relevant" when considering lower back pain, regardless of chronicity/acutity.
I think we'll next need to get into a discussion of current, recent and historical. Current is thought of in the now, but is actually in the very short term past, as in just happened. Recent is a slightly longer window, perhaps measured in the context appropriate to the situation. When one asks what my recent immunizations are, the context is a little bit different from when was my most recent blood glucose. Both are thinking about "most recent last", but not just now events. The time context window is different for one rather than the other. If you wanted to know when my most recent TDAP was, you might look back 5 years before saying, "That's far enough! You need a new one," whereas for blood glucose, you might go back as far as a year. In both cases, anything before those dates is "historical" because it doesn't matter how far back after that time frame, its relevance is about equal (time for a new one).
The locus of focus may also affect our perception of relevance. At the time of a visit, attention is on the condition that the patient is being treated for, the chief complaint or reason for visit. For another whose attention is on that visit (locus), their attention (focus) may be elsewhere. In communicating forward in time we usually assume that our focus will match that of the receiver, and so would determine relevance based on that focus. We often discover that what appeared to be inconsequential then is more important now.
We have no way to predict the future, which results in the fear that a determination of relevance could prevent something that could/should have been communicated. If, however, we can, as I would, put a protective force bubble around "what to do with the determination" as being subject to use-case appropriate judgement, we might come to dramatic consensus on what is "relevant" as best we are able to determine.
Keith
No comments:
Post a Comment