That is my understanding as well.  But if you have IRB approval for a
limited dataset, I think the tract could be part of a limited dataset
since it would not be considered a primary identifier of the patient.

Phillip


On 7/17/17, 1:54 PM, "Mei Liu" <[email protected]> wrote:

>Based on my limited understanding, ACS variables linked to a census tract
>contain aggregate/average information. But the census tract ID itself
>when linked to a patient record is PHI because it identifies a geographic
>location with <20,000 people. Am I interpreting this correctly?
>
>Mei
>
>-----Original Message-----
>From: Bernard Black [mailto:[email protected]]
>Sent: Monday, July 17, 2017 1:09 PM
>To: [email protected]; Dan Connolly; Mei Liu;
>[email protected]; [email protected]
>Cc: [email protected]; [email protected]; [email protected]; Lindsey Cook;
>Charon Gladfelter; Brennan Connolly; [email protected];
>[email protected]; [email protected];
>[email protected]; Hillary Sandoval; George Kowalski
>Subject: RE: [gpc-informatics] #544: how to convert ACS raw variables to
>CAP variables (e.g. type of health insurance)?
>
>Census tract information *is* averaged, and is not PHI.
>That is why Census is willing to release it.
>
>Bernie
>*************************************************************
>Bernard S. Black
>[email protected]
>Nicholas D. Chabraja Professor, Northwestern University
>     Pritzker School of Law
>     Institute for Policy Research
>     Kellogg School of Management, Finance Dep¹t Law School:  375 East
>Chicago Ave., Chicago IL 60611
>Kellogg:  2211 Campus Drive, Evanston IL 60208
>tel:  law:  312-503-2784; Kellogg:  847-491-5049; cell: 847-807-9599
>papers on SSRN at:  http://ssrn.com/author=16042
>************************************************************
>
>-----Original Message-----
>From: GPC Informatics [mailto:[email protected]]
>Sent: Monday, July 17, 2017 12:11 PM
>To: Dan Connelly <[email protected]>; [email protected];
>[email protected]; [email protected]
>Cc: [email protected]; Bernard Black <[email protected]>;
>[email protected]; [email protected]; Lindsey Cook
><[email protected]>; Charon Gladfelter
><[email protected]>; [email protected];
>[email protected]; [email protected];
>[email protected]; [email protected]; [email protected];
>George Kowalski <[email protected]>
>Subject: Re: [gpc-informatics] #544: how to convert ACS raw variables to
>CAP variables (e.g. type of health insurance)?
>
>#544: how to convert ACS raw variables to CAP variables (e.g. type of
>health insurance)?
>--------------------------+-----------------------
> Reporter:  dconnolly     |       Owner:  meiliu
>     Type:  design-issue  |      Status:  assigned
> Priority:  major         |   Milestone:  next-d
>Component:  data-stds     |  Resolution:
> Keywords:                |  Blocked By:
> Blocking:  598           |
>--------------------------+-----------------------
>
>Comment (by meiliu):
>
> Our most recent IRB protocol is in the attachment.
> - As noted in the protocol, "...variables such as income, educational
>attainment, likelihood of employment, poverty status, owner-occupied
>house  value, and health insurance coverage derived from geocoded address
>data  will be used only as average values, not tied to any specific
>individual  but rather to study populations."
> - I did not find any language on sharing census tract information.
> I've discussed this with Tamara McMahon and agree that since census
>tract  may be PHI, we currently don't have the governance to take
>approach 1.
>
> Welcome any input from other sites.
>
>--
>Ticket URL:
><https://urldefense.proofpoint.com/v2/url?u=http-3A__informatics.gpcnetwor
>k.org_trac_Project_ticket_544-23comment-3A15&d=DwICaQ&c=yHlS04HhBraes5BQ9u
>eu5zKhE7rtNXt_d012z2PA6ws&r=CqPxCu-EywA2wo4iO-6BFqfwPQ2roHdsnjQi7SSLgLE&m=
>wjt2QT-caC0BaQoqA-1jzlQgfOhmfVhDcHV0oV9o_sA&s=SiG0BJXb-ag0DCYa_svdEDeEWNVO
>JVkGOkV0JtvXSAY&e= > gpc-informatics
><https://urldefense.proofpoint.com/v2/url?u=http-3A__informatics.gpcnetwor
>k.org_&d=DwICaQ&c=yHlS04HhBraes5BQ9ueu5zKhE7rtNXt_d012z2PA6ws&r=CqPxCu-Eyw
>A2wo4iO-6BFqfwPQ2roHdsnjQi7SSLgLE&m=wjt2QT-caC0BaQoqA-1jzlQgfOhmfVhDcHV0oV
>9o_sA&s=Mpnaed_RgiVlMRfvyIbhydJzr6UR3yEq9-l7Cbir8co&e= > Greater Plains
>Network - Informatics


________________________________

UT Southwestern


Medical Center



The future of medicine, today.


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

Reply via email to