From razman@mindspring.com Thu Apr 15 14:22:16 1999 Received: from mxu3.u.washington.edu (mxu3.u.washington.edu [140.142.33.7]) by lists.u.washington.edu (8.9.3+UW99.02/8.9.3+UW99.01) with ESMTP id OAA09648 for ; Thu, 15 Apr 1999 14:22:15 -0700 From: razman@mindspring.com Received: from smtp5.mindspring.com (smtp5.mindspring.com [207.69.200.82]) by mxu3.u.washington.edu (8.9.3+UW99.02/8.9.3+UW99.01) with ESMTP id OAA03297 for ; Thu, 15 Apr 1999 14:22:14 -0700 Received: from drmemory (user-38lcdrg.dialup.mindspring.com [209.86.55.112]) by smtp5.mindspring.com (8.8.5/8.8.5) with SMTP id RAA03521 for ; Thu, 15 Apr 1999 17:22:12 -0400 (EDT) Message-ID: <002c01be8784$19ec4520$5a90c1a7@drmemory> To: Subject: Re: WAPHGIS: GIS AND CONFIDENTIALITY Date: Thu, 15 Apr 1999 17:08:21 -0400 MIME-Version: 1.0 Content-Type: multipart/alternative; boundary="----=_NextPart_000_0029_01BE8762.908F2E80" X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 4.72.2106.4 X-MimeOLE: Produced By Microsoft MimeOLE V4.72.2106.4 This is a multi-part message in MIME format. ------=_NextPart_000_0029_01BE8762.908F2E80 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable WAPHGIS: The Health Assessment Section of the Georgia Division of Public Health = has been working on geocoding Vital Records birth files for the development = of rates at the census tract level. We exclude personal identifiers (name, address, etc.) from public data requests. (This is not the case = for data requests from legitimate public health partners.) These data = requests now require electronic form, such as Access data bases or flat files for SAS analysis. This change from printed reports makes confidentiality an even more important issue for us. It also raises issues of ownership of derived rates and conflict over published rates attributed to Georgia health data that differ from state published rates. However, to address the thread issue, one possibility for a spatial confidentiality protocol could be to classify spatial location as a derivative of personal identifier data. This would place latitude and longitude in the same data class as name and address identifiers, which could then be excluded under existing data policy from HHS, CDC, NCHS (National Center for Health Statistics), or state health department. A different issue derives from spatial locations of public record information, such as food service, well, and septic tank permits. If spatial data for these locations becomes a standard part of the public record, they becomes subject to Open Record Act public information requests. We have yet to address the consequences of spatial accuracy = on the release of this spatial public information. On the issue of data visualization, whether it be by web page of printed map, when we teach health planners and environmentalists GIS, we stress data confidentiality. In particular, we stress that, if it absolutely required to present point data for individual health events, maps shall = not include significant spatial features (such as roads) which would imply identification. I think that we should not forget that a false identification is just as damaging as a correct identification. This specifically challenges randomization schema. Richard Schreiber GIS Analyst Health Assessment Section Division of Public Health Georgia Department of Human Resources ras@dhr.state.ga.us At 11:02 AM 4/14/99 -0700, you wrote: >Colleagues: Perhaps you can respond to the message below. Please = share=20 >your responses with the listserve if you also contact her directly. If = an=20 >agency were to draw up a set of confidentiality guidelines, what are = the=20 >issues? Why is this important? What problems have you had displaying,=20 >using maps with sensitive data? What guidelines or policies does your=20 >organization have in place generally and how does that impact the use = of=20 >GIS? Have you had an instance where GIS use compromised someone's=20 >privacy? Do you know who really knows all about this? If you have = written=20 >about this, could we have access to your papers or a citation? > >Thanks > >Richard E. Hoskins =20 > eMAIL: REH0303@doh.wa.gov > tel: 360-236-4270 > fax:360-236-4245 > >Washington State Public Health & GIS listserve >WAPHGIS@u.washington.edu >WA DOH & NW Center for Public Health Practice >at Univ WA in Seattle > > >********************* Forwarding Message ********************* >Dick > >We've been discussing this topic in light of the agency work around=20 >confidentiality and in a data release group that is working to develop = an=20 >agency policy around data release. We are looking into feeding this = topic=20 >into a regional conference, and I was interested in knowing if there = are=20 >any "industry" guidelines around GIS and health information in=20 >particular, and if you could point me to those or to knowledgeable=20 >persons that we could discuss with or invite as presenters. Thanks for=20 >the help! > >Vicki Hohner, Data Products Manager >Hospital and Patient Data Systems >WA State Department of Health >Phone: (360) 236-4211 >FAX: (360) 664-8579 >MS: 7811 >Internet: VKH0303@DOH.WA.GOV > Richard A. Schreiber GIS Analyst Division of Public Health Georgia Department of Human Resources ras@dhr.state.ga.us ------=_NextPart_000_0029_01BE8762.908F2E80 Content-Type: text/html; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable
WAPHGIS:

The Health Assessment Section of the Georgia = Division of=20 Public Health has
been working on geocoding Vital Records birth files = for the=20 development of
rates at the census tract = level.     =20 We exclude personal identifiers
(name, address, etc.) from public = data=20 requests. (This is not the case for
data requests from legitimate = public=20 health partners.)  These data requests
now require electronic = form, such=20 as Access data bases or flat files for
SAS analysis.  This = change from=20 printed reports makes confidentiality an
even more important issue = for=20 us.  It also raises issues of ownership of
derived rates and = conflict=20 over published rates attributed to Georgia
health data that differ = from state=20 published rates.

However, to address the thread issue, one = possibility=20 for a spatial
confidentiality protocol could be to classify spatial = location=20 as a
derivative of personal identifier data.  This would place = latitude=20 and
longitude in the same data class as name and address identifiers, = which
could then be excluded under existing data policy from HHS, = CDC,=20 NCHS
(National Center for Health Statistics), or state health=20 department.

A different issue derives from spatial locations of = public=20 record
information, such as food service, well, and septic tank=20 permits.  If
spatial data for these locations becomes a standard = part of=20 the public
record, they becomes subject to Open Record Act public=20 information
requests.  We have yet to address the consequences = of=20 spatial accuracy on
the  release of this spatial public=20 information.

On the issue of data visualization, whether it be by = web=20 page of printed
map, when we teach health planners and = environmentalists GIS,=20 we stress
data confidentiality.  In particular, we stress that, = if it=20 absolutely
required to present point data for individual health = events, maps=20 shall not
include significant spatial features (such as roads) which = would=20 imply
identification.  I think that we should not forget that a=20 false
identification is just as damaging as a correct = identification. =20 This
specifically challenges randomization schema.

Richard=20 Schreiber
GIS Analyst
Health Assessment Section
Division of = Public=20 Health
Georgia Department of Human Resources
ras@dhr.state.ga.us


At= 11:02 AM=20 4/14/99 -0700, you wrote:
>Colleagues:  Perhaps  you can = respond=20 to the message below. Please share
>your responses with the = listserve if=20 you also contact her directly. If an
>agency were to draw up a = set of=20 confidentiality guidelines, what are the
>issues? Why is this = important?=20 What problems have you had displaying,
>using maps with sensitive = data?=20 What  guidelines or policies does your
>organization have in = place=20 generally and how does that impact the use of
>GIS? Have you had = an=20 instance where GIS use compromised someone's
>privacy? Do you = know who=20 really knows all about this? If you have written
>about this, = could we=20 have access to your papers or a=20 citation?
>
>Thanks
>
>Richard E.=20 Hoskins       
> eMAIL: REH0303@doh.wa.gov
> tel:=20 360-236-4270
> fax:360-236-4245
>
>Washington State = Public=20 Health & GIS listserve
>WAPHGIS@u.washington.edu
= >WA=20 DOH & NW Center for Public Health Practice
>at Univ WA in=20 Seattle
>
>
>********************* Forwarding Message=20 *********************
>Dick
>
>We've been discussing = this=20 topic in light of the agency work around
>confidentiality and in = a data=20 release group that is working to develop an
>agency policy around = data=20 release. We are looking into feeding this topic
>into a regional=20 conference, and I was interested in knowing if there are
>any=20 "industry" guidelines around GIS and health information in=20
>particular, and if you could point me to those or to = knowledgeable=20
>persons that we could discuss with or invite as presenters. = Thanks for=20
>the help!
>
>Vicki Hohner, Data Products=20 Manager
>Hospital and Patient Data Systems
>WA State = Department of=20 Health
>Phone: (360) 236-4211
>FAX:   (360)=20 664-8579
>MS:    7811
>Internet: VKH0303@DOH.WA.GOV
>

=
Richard A. Schreiber
GIS = Analyst
Division=20 of Public Health
Georgia Department of Human Resources
ras@dhr.state.ga.us
<= /BODY> ------=_NextPart_000_0029_01BE8762.908F2E80-- .