So yes, every pediatric patient with a BMI above the 5th percentile for
their sex and age is a potential survey respondent. The next panel
(group 2) does narrow it somewhat by excluding the visits where the
/raw/ BMI is above 75.
Please note also that we are no longer filtering for diseases up-front--
disease codes will be included in the de-id EMR data that will accompany
the survey responses. The large set of diagnosis, drug, and other codes
in the "data elements obesity.html" file I sent earlier are not
exclusion or inclusion criteria. They are intended to generate a query
object that can then be dragged onto DataBuilder to create the
share-able de-identified dataset.
This is will be a large number of patients, which is why we have been
spending so much time at our end automating and testing the survey
management workflow.
On 04/01/2015 06:24 PM, Hale, Daniel E wrote:
The exclusion IS less than 5th percentile. The index criteria bins are
5-85. 85-95 and greater than 95. These are standard CDC criteria
On Apr 1, 2015, at 6:08 PM, Nadkarni, Prakash
<[email protected] <mailto:[email protected]>> wrote:
Are the semantics of the “obesity query” correct?
While Ok to test an i2b2 instance’s operation, the query
“ Find everyone with a BMI percentile > 5”
would actually identify **everybody** other than the severely
underweight – not the obese patients who we are expected contact.
(If we are looking for underweight individuals, we would have to
change the criterion to < 5, not >5).
_______________________________________________
Gpc-dev mailing list
[email protected]
http://listserv.kumc.edu/mailman/listinfo/gpc-dev
