Here is a snippet of the metadata.  For age in months we calculate off of 
BIRTH_DATE which has been shifted +- 30.

C_HLEVEL

C_FULLNAME

C_NAME

C_SYNONYM_CD

C_VISUALATTRIBUTES

C_BASECODE

C_METADATXML

C_FACTTABLECOLUMN

C_TABLENAME

C_COLUMNNAME

C_COLUMNDATATYPE

C_OPERATOR

C_DIMCODE

C_TOOLTIP

SOURCESYSTEM_CD

UPDATE_DATE

M_APPLIED_PATH

4

\Demographics\patient_demographics\patient_age\patient_age_00-09\patient_age_00\

Patient Age < 1 Year

N

FA

DEM|AGE:0

CONCEPT_CD

CONCEPT_DIMENSION

CONCEPT_PATH

T

LIKE

\Demographics\patient_demographics\patient_age\patient_age_00-09\patient_age_00\

Demographics \ patient_demographics \ patient_age \ patient_age_00-09 \ 
patient_age_00

demographics.owl

2014-03-11 00:00:00

@

5

\Demographics\patient_demographics\patient_age\patient_age_00-09\patient_age_00\patient_age_month_00\

Patient Age < 1 Month

N

FA

PATIENT_NUM

PATIENT_DIMENSION

BIRTH_DATE

N

<

GetDate() - (365.25 * 0.085)

Demographics \ patient_demographics \ patient_age \ patient_age_00-09 \ 
patient_age_00 \ patient_age_month_00

demographics.owl

2014-03-11 00:00:00

@

5

\Demographics\patient_demographics\patient_age\patient_age_00-09\patient_age_00\patient_age_month_01\

Patient Age 01 Month

N

LA

PATIENT_NUM

PATIENT_DIMENSION

BIRTH_DATE

N

BETWEEN

GetDate() - (365.25 * 0.162) AND GetDate() - (365.25 * 0.085)

Demographics \ patient_demographics \ patient_age \ patient_age_00-09 \ 
patient_age_00 \ patient_age_month_01

demographics.owl

2014-03-11 00:00:00

@


Currently our pediatricians will be happy that they can create finer 
aggregations.  Some have asked for age in days, but our IRB said no to that 
granular of information in i2b2.

Nathan



From: [email protected] [mailto:[email protected]] On Behalf Of Justin Dale
Sent: Tuesday, March 18, 2014 10:10 AM
To: Wilson Nathan
Cc: [email protected]
Subject: Re: [Gpc-dev] de-identification - patients under 1 year old

Thanks for the replies.

Nathan - for the patients 36 months or less: do you shift the birthdate +/- 30 
days and then provide the user with the shifted birthdate value and/or age 
concept represented in years/months?  Have you found times when even a +/- 30 
day shift causes issues with pediatricians (or other users)?  We currently use 
a date shift between -1 and -365 days but I'm thinking that may be too extreme.

Regards,

Justin Dale
Manager - Research Development and Support
University of Minnesota
Academic Health Center - Office of Information Systems
Phone: 612.624.9747
Fax: 612.624.7458
[email protected]<mailto:[email protected]>

On Tue, Mar 18, 2014 at 9:34 AM, Wilson Nathan 
<[email protected]<mailto:[email protected]>> wrote:
Here at UW-Madison we currently group all patients whose age is less than 1 
years old into a single age category called "0 Years Old".  This however did 
not work well with pediatricians here, so we are in the process of implementing 
an ontology which will calculate a patients age for patients 36 months or less 
based the patients shifted birthdate in addition to the traditional concept 
based age approach.

Nathan Wilson

From: 
[email protected]<mailto:[email protected]> 
[mailto:[email protected]<mailto:[email protected]>]
 On Behalf Of Justin Dale
Sent: Tuesday, March 18, 2014 9:04 AM
To: [email protected]<mailto:[email protected]>

Subject: [Gpc-dev] de-identification - patients under 1 year old

Seeking input on how your site represents a patient's age who is under 1 year 
old? Do you group them in the 1 year old bucket (similar to 90 and over)?  Do 
you provide fractions of a year or represent months (seems like it goes against 
the safe harbor rules)?  Do you shift the birthdate and just provide that?  
Thanks.

Justin

_______________________________________________
Gpc-dev mailing list
[email protected]
http://listserv.kumc.edu/mailman/listinfo/gpc-dev

Reply via email to