From: Glen McCallum <address@hidden>
Subject: Re: [Egadss-devel] variance in test units with repeated
measures (was A1C Lab Test Value)
To: address@hidden
Date: Mon, 23 Jan 2006 12:40:15 -0800
Hi Jim,
<response in line>
On 22-Jan-06, at 9:47 AM, J Busser wrote:
At 11:49 AM -0800 1/20/06, Glen McCallum wrote:
...COMPETE returned 0.07 as the value of the A1C test whereas we
return value=7 and unit=% (hence that what I look for in the Arden
Syntax).
Would 0.07 have a unit value of "unity", else "/1", else a null
value for units? Not sure if there is a standard expression for
this.
Likewise, are any of the characters/expressions "%" or "/100" or or
"100^-1" considered "standard"?
I'll defer this one Iryna. She's more familiar with CDA templates
and the interface with the EMR.
I know that within a single EMR, is it possible for a single
patient to have a series of values that have been expressed in
different units. AFAIK, few EMRs provide the means to process (e.g.
translate / express in a equivalent form) such a series. So it's
interesting to consider how such a dilemma could be resolved (if it
all) via EGADSS. For example, I don't know whether a change in
units in which a test is reported will result in a change in its
LOINC code.
Definitely interesting. We're getting more and more into LOINC as we
work with lab test data.
I suppose the use cases for EGADSS would include:
- different EMRs each accessing EGADDS with the same test of
interest but whose units are different
- any one EMR accessing an EGADSS guideline that employs/requires a
series of repeated measures on a single test, where within that
series, the results had been stored using different units.
I'm certainly interested in how or if our CDA Template restricts
these conditions. Iryna?
Is there so far within EGADSS a guideline that must handle a
repeated measure?
No. The guidelines that we've been encoding recently always
reference the value of the most recent test.
A specific use case --- challenging despite adherence to a single
test unit --- would be ongoing blood thinning aka anticoagulation.
A "normal" result for this test, for a patient not receiving
anticoagulation, is around 1.0. A target on-therapy range is
typically 2.0 - 3.0 or 2.5 - 3.5, or could even be locally-defined.
There is value to being able to flag a single result as whether it
lies inside or outside the desired range, kind of like a single
glucose value for a diabetic. But to express treatment results over
time, repeated measures on the test result known as the INR should
be processed to achieve the % of time that the result is within the
target range. Where a consecutive pair of results crosses the
target range, one method applies linear interpolation to the time
interval between them, to apportion the segments as in- or out-of
range. A variant method could determine how often a patient is
"close" to range, e.g. for a target range of 2.0 - 3.0 it might
determine the percent of time that a patient is between 1.7 and 3.5.
This would be a nice application for fuzzy logic.
I suspect the time-based interpolation is important. Is it already
part of the design or the plan, or something to consider, perhaps
for EGADSS v2 or 3?
Definitely.
thnx,
~Glen
-------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc. Do you grep through log files
for problems? Stop! Download the new AJAX search engine that makes
searching your log files as easy as surfing the web. DOWNLOAD SPLUNK!
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=103432&bid=230486&dat=121642
_______________________________________________
Egadss-devel mailing list
address@hidden
https://lists.sourceforge.net/lists/listinfo/egadss-devel