PUBLIC HEALTH GIS NEWS AND INFORMATIONDecember 1997 (No. 19)

Dedicated to CDC/ATSDR scientific excellence and advancement in disease control and prevention using GIS

**********************************************ANNOUNCING: ATSDR�s 1998 �GIS IN PUBLIC HEALTH� CONFERENCE!

August 17-20, Town and Country Resort and Convention Center, San Diego[Requests for Preliminary Announcement and Call for Papers may be sent by e-mail to GIS@CDC.GOV]

***********************************************

Selected Contents: Conferences (p.1); News from GIS Users (pp.1-6); GIS outreach (pp.6-9); Special reports(pp.9-10); GIS literature (pp.10-12); NCVHS news (pp.12-19); Net site of interest (p.19); Final thought (pp.19-20)

I. Public Health GIS (and related) Events

� HazWaste World/Superfund XVIII Conference, E.J. Krause & Associates and the EnvironmentalIndustry Associations, Washington, D.C., December 2-4, 1997 [Contact: http://http://www.ejkrause.com/enviroshows or Susan Cantor at voice (301) 493-5500]

� Land Satellite Information in the Next Decade II:Sources and Applications, American Society forPhotogrammetry and Remote Sensing and the NorthAmerican Remote Sensing Industries Association,Washington, D.C., December 3-5, 1997 [Contact:http://www.asprs.org/asprs/index2.html or voice (301)493-0290]

�International Conference and Workshop onInteroperating GIS, National Center for GeographicInformation and Analysis and the Open GISConsortium Inc., Santa Barbara, December 3-6, 1997[Contact: e-mail i20@ncgia.ucsb.edu or voice (805)893-8224]

���� NCHS Cartography and GIS Guest LectureSeries, �A Study of Homicide in Washington, D.C.Using Desktop GIS,� [see enclosed abstract] byDeWitt Davis, University of the District of Columbia,at NCHS Hyattsville, January 28, 1998, 2:00-3:00 PM;please make Envision arrangements now at offsitelocations [Contact: Chuck Croner at e-mail cmc2@cdc.gov]

� 1998 ACM (American Congress on Surveying and

Mapping) Annual Convention and Exhibition,Baltimore, MD, February 28-March 4, 1998 [Contact:http://www.landsurveyor.com/acsm/or Denise Calvertat voice (301) 493-0200]

� 4th International Conference on HemorrhagicFevers and Hantaviruses, Centers for Disease Controland Prevention, Atlanta, March 5-7, 1998 [Contact:Amy Corneli, Division of Viral and RickettsialDiseases, National Center for Infectious Diseases ate-mail akc8@ cdc.gov or voice (404) 639-1510 or seehttp://www. cdc.gov/ncidod/diseases/hanta/hant conf.htm]

� International Conference on Emerging InfectiousDiseases, Centers for Disease Control and Prevention,Atlanta, March 8-11, 1998 [Contact: voice (202) 942-9248 or see http://www.cdc.gov/ncidod/eidann.htm]

� 94th Annual Meeting of the Association ofAmerican Geographers, Boston, March 25-29, 1998[Contact: AAG, 1710 16th Street NW, Washington,DC 20009-3198 or voice (202) 234-1450]

II. News from GIS USERS (Please communicate directly with colleagues on any issues)

A. General News (and Training Opportunities)1. From Karl Sieber, NIOSH: Steve Spaeth andmyself from NIOSH attended GIS-LIS '97, Oct. 28-31,1997 in Cincinnati. There were a variety of shortcourses, ranging from an introduction to GIS to moreadvanced topics. (Both of us took the intro

2

course-upon reviewing it and the notes from the courseyou [Chuck Croner] presented at the CDC symposiumon small area statistics in 1995, your course wasdefinitely more useful in giving specifics on use ofTIGER files, etc. and other data useful for GIS). Theconference was more a trade show, and I learned aboutmany products and data sources available on the web.While there, I met several folks who work for the EPAand actively use GIS techniques, as well as folksinvolved in mapping nuclear disposal sites and waterquality.

2. The National Center for Health Statistics (NCHS),CDC, commemorated the 10th annual NCHS/NationalGeography Awareness Week (GAW) celebration witha presentation by Lee De Cola, Research PhysicalScientist, US Geological Survey. His talk,�Visualizing Demographic Statistics through MapAnimation: Implications for Human DiseaseDatabases� (retitled) sent a powerful message aboutthe marketability of dynamic mapping. In less thanone minute, the growth of San Francisco wasportrayed by decade over a 150 year period. Theresponse by television news media in SF, in 1994, wasexceptional with all stations providing coverage (andcreative drama) to their respective presentations.Maryland Governor Glendening has shown similarsupport for Lee�s 1997 animation of population growthover 200 years between Washington, DC andBaltimore. Lee used Mathematica and GIF Animatorfor the animation, though the original data werecompiled in ARC/INFO, and some output was doneusing ERDAS. If you have yet to see this animation,you can visit web site http://geog.gmu.edu/gess/classes/geog590/gis_internet/ldecola/baltwash.Eight CDC locations joined Lee�s presentation throughEnvision. The GAW program is part of the NCHSCartography and GIS Guest Lecture Series. Editor

3. January 28, 1998 NCHS Cartography and GISGuest Lecture Series presentation, "A Study ofHomicide in Washington, DC using Desktop GIS," byDr. Dewitt Davis, University of the District ofColumbia. Abstract: The research is a longitudinalstudy which examines homicides at the street, census

tract, ward, and quadrant levels. These are thegeographic units utilized by the different agencies inthe DC City Government when making political andeconomic decisions. The study also attempts toestablish if there is cause and effect in relation torelevant 1990 US census socioeconomic data andWashington DC Metropolitan Police total homicidesdata for several year periods from 1990 to present.[Editor: Dr. Davis taught geography at Ohio StateUniversity (1973-1980), University of Arkansas PineBluff (1980-1984), and University of District ofColumbia (1984-present). Since arriving at theUniversity of the District of Columbia, he chaired theDepartment of Geography four years, was actingassistant dean for one year, and is currently thecoordinator of the Geography Program in theDepartment of Urban Affairs and Geography]

4. Genetic Speaker Series, CDC: The Office ofGenetics and Disease Prevention, in collaboration withthe Office of the Associate Director for Science andHRMO announces a distinguished speaker series onethical, legal and social issues in genetics and publichealth. "Genetics, Public Health and the Law" is thetopic of a December 3 presentation (1:00-2:00 p.m.,Auditorium A, Clifton Rd) by Ellen Wright Clayton,J.D., M.D., Associate Professor of Law, VanderbiltUniversity. Dr. Wright Clayton is a member of theNational Advisory Council for the National HumanGenome Research Institute and has been writing aboutlegal and ethical issues posed by developments ingenetics for the last 20 years. Dr. Wright Clayton willbegin her presentation by discussing the constitutionalfoundations of the government's power to protect thehealth of its citizens, focusing on the particular issuesraised by public health efforts in genetics. She willthen discuss the limits on this authority as well asgovernmental actions that might give rise to liability.

B. Technical News5. From Bob Wentworth, EPA: I'm part of a teamworking to increase public access to environmentaldata via the EPA's Envirofacts data access system. Iwill share your feedback with the team. We areinterested in knowing more of the responses from the

3

public about their reactions to our new home page andif they may have any issues with the following:loading of the new page, colors, are they getting anykind of error messages, either from Netscape or fromEnvirofacts, or otherwise, and how our newfunctionality works for them; how intuitive is it, e.g.,the 'mouse over' capability. This will be valuableinformation for us to have. The more diverse our targetaudience for these tests, the better. Please feel free topass along this request for feedback! We're hoping todevelop a data access system that will be as useful aspossible to the public. Let us know how we canimprove! The URL for Envirofacts is: http://www.epa.gov/enviro/ html/ef_home.html. My e-mail address forresponse is: Wentworth@merc.rx.uga.edu.

6 . From Loren Hal l , EPA ( throughppgis-scope@igc.org): Here are a few suggestions forplaces where you can use on-line mapping and findout more about applying GIS in the environmentaljustice context: On-line Mapping: Pollution MappingProjects and Toxics Databases (collection of links) athttp:// www.envirolink.org/issues/pollution-map/index.html; EPA's Maps on Demand (not real time,but can produce large and detailed maps with pointsources, population characteristics, drinking waterfacilities, etc, but also be aware that large maps insome formats need lots of PC memory to display) athttp://www.epa.gov/enviro/ html/mod/index.html; Surfyour watershed (interactive display and retrieval ofdata by watershed - portions require a Java-capableweb browser) at http://www. epa.gov/surf/;Demographic Data Viewer (national interactiveCensus data thematic mapping) at http://plue.sedac.ciesin.org/plue/ddviewer/; West Virginiastate interactive mapping site (includes a Java appletthat you will need a recent web browser to use)http://poca.osmre.gov/form/frame-1.html.

Data Sources: Government InformationSharing Project: Oregon State University - collectionof national and state-specific population, housing,economic data, etc. - suitable for download andmapping) at http://govinfo.kerr.orst.edu/; Egrets onlineguide to geographic information about East St. Louis(mostly a map library of products created by the East

St. Louis Action Research Project) at http://www.imlab.uiuc.edu/egrets/; Guide To On-line And MostlyFree U.S. Geospatial and Attribute Data at http://www.cast.uark.edu/local/ hunt/index.html.

7. From John Gardenier, NCHS: On December 15-17at NIH, there will be a Conference on Scientific andTechnical Data Exchange and Integration. It issponsored by the U. S. National Committee onCODATA, which represents U.S. interests at theinternational Committee on Data for Science andTechnology (CODATA), an InterdisciplinaryCommittee of the International Council of ScientificUnions. Additional information is available on-line athttp:/www.cisti.nrc.ca/programs/codata/welcome.html, including the full program of this and otherrelated conferences.

Within the conference, I have been invited asa Panelist at the Breakout Session on Bioinformatics,Tuesday 12/16, 1:15-4:15. [Panel description andparticipant list at the end.] Having only a 15 minutewindow, I would still like to mention a range of thetopics most crucial to DHHS components, tobiostatisticians, and to the statistics profession as awhole. If you would be willing to provide a paragraphor two per topic, I would be happy to "get it on therecord" at the conference - with your name andaffiliation, of course. Whether or not you can do sopersonally, you might (also) refer this request tosomeone else whose input you would like to seeincluded.

Apart from whatever else you may suggest, thetopics I am currently hoping to touch upon briefly are:(1) Concerns about potential adverse impacts on datasharing from proposed treaties and legislationregarding new forms of intellectual property protectionfor databases. Emphasis on the AAAS and AmericanStatistical Association (ASA) positions. (2) Concernsabout statistical ethics and competence in biomedicalresearch and reporting - with special emphasis onp r o b l e m s o f a d v o c a c y , b i a s , a n dincomplete/misleading reporting of experiments andanalyses. Activities of the ASA Committee onProfessional Ethics (John Bailar). (3) Concerns aboutthe criticality of global access to sound data-gathering,

4

analysis, and reporting methods. Mention of at leastthe CDC EPI-INFO epidemiological computer systemas an example of a globally distributed, inexpensive,flexible, and useful product promoting and protectingpublic health (Andy Dean). (4) Concerns about theexplosion of data coming out of molecular biology andgenomic research, its non-statistical andquasi-statistical properties requiring new mathematicaland computer science approaches and thedissemination of both the new knowledge and thetools to take advantage of it (Genome project; ACM?).(5) Concerns about data policy and standards and theirimpact on the comparability and intelligibility of datafrom various national and international data systems(Various sources?). (6) Capabilities and limitations ofgeographic information systems (GIS) in both analysisof data and in its presentation to policy makers,journalists, and the public at large (Charles Croner,others?).

Each topic, of course, is worthy of/has meritedentire conferences so there will be no attempt to getinto technical detail. For interested attendees, it wouldbe useful to provide a handout of (non-exhaustive, butinteresting) bibliographic references. Anything youwould be willing and able to contribute would begreatly appreciated and acknowledged. Here is theChair's description of the core questions for thepanelists: The breakout session on Bioinformatics willinclude presentations by a panel of speakers whorepresent research, government and academicorganizations. The panelists will be asked to addressthe issues of a) identifying areas in which dataexchange and integration in bioinformatics areimportant, including exchange and integration ofbiological data and that of other sciences. Somepossible examples are the interactions betweenmolecular biology information and taxonomy,ecological risk assessment, drug design, epidemiology,biodiversity studies, sustainable agriculture, emissionof greenhouse gases, etc. b) addressing examples ofsuccessful or planned efforts to foster serious andsignificant cooperation among scientific andengineering disciplines and different types oforganizations in science and technology; as well asidentifying factors which foster such cooperation and

factors which impede it. More specifically, the core questions for the

panelists are: 1) Define important and emerging issuesinvolving bioinformatics in which multi-disciplinarydata exchange and integration are or will be necessary.Give brief examples of the types of data that areinvolved. 2)List the driving forces (scientific,information technology, applications, policy,organizational, etc.) that will make this data exchangeand integration happen. Identify barriers (such asscientific, information technology, policy,organizational, etc.) that might impede this process. 3)Given that most, if not all, Federal funding agencieshave biological data components to their programs,what are the roles of the federal funding agencies?Inter-agency groups? Professional and technicalsocieties in the biological and other sciences?Standards organizations? Industry? Scientiststhemselves? Others? 4) What actions can be taken toovercome identified and potential barriers? How canCODATA help? 5) What is needed to educate thecommunities involved in the varied areas of biologicalsciences and associated technologies about existingcomputer science approaches and other informationtechnology developments applicable to the issues ofdata access and dissemination? 6. What is the bestscenario for five years from now? What else is neededto get there?

The participants and references are: I.Introduction: Micah I. Krichevsky, Chair, BionomicsInternational & George Mason U.; II. BriefPresentations by Panelists Lois Blaine, Rapporteur,ATCC, Peter Buneman, U. Penn., John Gardenier,CDC, Chris Overton, U. Penn., Mark Segal, EPA; III.Panel Discussion (with session participants); IV.Conclusions. References: 1) The study on "Bits ofPower: Issues in Global Access to Scientific Data" wasrecently published. It investigated the barriers andother issues in the international transfers of scientificdata with the goal of improving access to scientificdata and services internationally. The study containsrecommendations in the area of bioinformatics bearingon the discussion (http://www.nas.edu/cpsma/bits2.htm); 2) The Committee on a Pilot Study forDatabase Interfaces analyzed the data management

5

problems associated with interfacing different types ofenvironmental data, especially small-scale ecologicaldata with large-scale geophysical data. TheCommittee's report, "Finding the Forest in the Trees:The Challenge of Combining Diverse EnvironmentalData", was published in 1994. The study also containspertinent examples and recommendations germane tothe discussion (http://www.nap.edu/readingroom/enter2.cgi?0309050820.html)

C. Internet News 8. USGS Launches New Electronic Clearinghouse forBiological Data: Dennis B. Fenn, Chief Biologist,today announced that the U. S. Geological Survey'sBiological Resources Division has initiated a newInternet-based clearinghouse that can be used to searchfor and locate existing sources of biological data andinformation from a variety of sources. Thisclearinghouse is part of the National BiologicalInformation Infrastructure (NBII), which is acooperative effort led by the USGS to increase accessto biological data and information maintained by avariety of Federal and State government agencies,universities, museums, libraries, and privateorganizations.

Through the NBII Clearinghouse <http://www.nbii.gov/clearinghouse.html>, Internet users cansearch through an assortment of standardizeddescriptions of different biological databases orinformation products to identify those that meet theirparticular requirements. These descriptions(metadata) concisely convey such things as subjectmatter; how, when, where, and by whom the data werecollected; whom to contact for more information; andhow to access the database or information product.The NBII Clearinghouse includes metadatadescriptions of biological databases and informationproducts developed and maintained by USGSscientists, as well as data and information developedand maintained by other NBII participants, includingFederal and State government agencies, universities,and private organizations. The NBII Clearinghousealso functions as a part of the National Spatial DataInfrastructure (NSDI) Clearinghouse <http://www.fgdc.gov/clearinghouse/index.html>, as many of the

biological databases described in the NBIIClearinghouse employ geospatial references.

"We are very pleased to be offering this newservice to the public, resource managers and scientists,and anyone else interested in locating existing sourcesof biological data and information," said Fenn. "Thisis a great opportunity not only for us to help get theresults of USGS biological science out to those whocan use the data and information, but also for us toprovide a tool that our partners and cooperators inNBII can use to help share their own data andinformation."

Users can search through the NBIIClearinghouse (much as they would use a card catalogin a library) using a variety of criteria, such as thename of the investigator or author who collected thedata or produced the information, subject-matterkeywords, and spatial coordinates for the location ofthe study/project. Special biological search criteria,including the ability to search for data or informationrelating to a particular species or other taxonomicgroup, are also provided. Metadata descriptions in theNBII Clearinghouse are developed according to theNBII's biological metadata standard <http://www.nbii.gov/current.status.html>, which also serves as abiological "enhancement" or "profile" of the FederalGeographic Data Committee's Geospatial MetadataContent Standard. The mission of the USGS/BRD is towork with others to provide the scientificunderstanding and technologies needed to support thesound management and conservation of the Nation'sbiological resources. [Source: Duncan Morrow,NBS-MIB; Contact: Contact: Anne Frondorf 703/648-4205]

9. From H. Gyde Lund, European Forest Institute,Finland (through fwim-l@listserv.vt.edu):On 22October 97, U.S. Sec. of Interior, Bruce Babbitt,approved the Federal Geographic Data Committee's(FGCC) Vegetation Information and ClassificationStandard after being endorsed by the FGDC SteeringCommittee. This will now be the standard vegetationclassification system for use by U.S. Federal Agenciesand their cooperators. The standard has been underdevelopment since 1992 and has undergone public

6

reviews. It may be viewed at <http://www.nbs.gov/fgdc.veg/>. The final format of the standard willchange slightly from what is presented at this sitewhen it is officially published, but the content isessentially the same. In addition to the public review,the FGDC vegetation standard was closely coordinateswith the development of the FAO Land CoverClassificaiton System (see website <http://www.fao.org/waicent/faoinfo/ agricult/AGL/AGLS/FGDCFAO.HTM to provide international links and to worktowards the development of a truly global vegetationand land cover classification system. For informationon the development of other U.S. geospatial standards,see the FGDC home page at <http://www.fgdc.gov>[Contact: Gyde at email gyde.lund@efi. joensuu.fi orweb site http://www. efi.joensuu.fi]

10. From Lois Dean , HUD (throughppgis-scope@igc.org) Announcement of ProposalDeadline concerning Competition for the 1998National Brownfields Assessment DemonstrationPilots; Copies of this document are available via theInternet on the EPA Brownfields Home Page athttp://www.epa. gov/brownfields/applicat.htm.Summary: The United States EnvironmentalProtection Agency (EPA) will begin to acceptproposals for the National Brownfields AssessmentPilots on October 9, 1997. The brownfields assessmentpilots (each funded up to $200,000 over two years) testcleanup and redevelopment planning models, directspecial efforts toward removing regulatory barrierswithout sacrificing protectiveness, and facilitatecoordinated environmental cleanup and redevelopmentefforts at the federal, state, and local levels. EPAexpects to select approximately 100 additionalNational brownfields assessment pilots by May 1998.

Applications will be accepted on a "rollingsubmissions" schedule. The deadlines for newapplications for the 1998 assessment pilots areDecember 15, 1997, and March 23, 1998. Applicationspostmarked after December 15, 1997, will beconsidered in the second round of competition.Previously unsuccessful applicants are advised thatthey must revise and resubmit their applications. TheNational brownfields assessment pilots are

administered on a competitive basis. To ensure a fairselection process, evaluation panels consisting of EPARegional and Headquarters staff and other federalagency representatives will assess how well theproposals meet the selection criteria outlined in thenewly revised application guidelines document entitled"The Brownfields Economic Redevelopment Initiative:Proposal Guidelines for Brownfields AssessmentDemonstration Pilots" (October 1997). This action iseffective as of October 9, 1997, and expires on March23, 1998. All proposals must be postmarked or sent toEPA via registered or tracked mail by the expirationdates cited above. The application guidelinesdocument can be obtained by calling the SuperfundHotline at the following numbers: Washington, DCMetro Area at 703-412-9810, Outside Washington, DCMetro at 1-800-424-9346, TDD for the HearingImpaired at 1-800-553-7672. Copies of this documentare available via the Internet at: http://www.epa.gov/brownfields/ applicat.htm

III. GIS Outreach(Editor: All solutions are welcome and will appear in the nextedition; please note that the use of trade names and commercialsources that may appear in Public Health GIS News andInformation is for identification only and does not implyendorsement by CDC or ATSDR) � From Christopher Hartwell, Harvard University:We're looking to see if there's a GIS database out thereon worldwide disease distributions, including (but notlimited to ) malaria and other tropical diseases, for asfar back as the stats allow. Our research team iscompiling a database on important indicators for theworld, as how these indicators may help or hindereconomic growth. If anyone knows of GIS sites thatwould hold this data, it would be appreciated if theycould contact me [Contact Chris at Harvard Institutefor International Development, e-mail chartwel@hiid.harvard.edu or voice (617) 496-5783]

� October 27 �request for assistance� to GIS Users:Brian McCarthy, Division of Reproductive Health,NCCDPHP just called to say he is leaving in a fewdays for Jordan and has a recently approved GISproject regarding disease outbreaks in Palestinianrefugee camps. Brian will use MapInfo software and a

7

GPS enabling link. If you know of ANY existingsources of digital maps of Jordan (especially), Syria,Lebanon, and Gaza and West Bank, please contactBrian by cob Wednesday or before by voice (770)488-5229 or e-mail: BJM1@CDC.GOV. Thank you inadvance for any help you can provide Brian. Editor

1. Response from Morris Maslia, ATSDR: Inresponse to Chuck Croner�s e-mail about the need fordigital maps of Jordan, Syria, etc., a colleague of minefrom the USGS just recently returned from the areawhere the USGS is being asked to be the "objectivepartner" for looking at the water resources of the areafor the Peace Process. He indicated he had visitedIsrael, Jordan, the west bank area, etc. Knowing theUSGS, they may have "ginned up" some quicky digitalmaps of all the countries of the Middle East area. Trycontacting John Clarke at the USGS (Ga. District,Water Resources Division). His telephone and e-mailare as follows: John Clarke, voice (770) 903-9100 ande-mail jsclarke@usgs.gov. Hope this helps and bestregards.

2. Response from Gerry Rushton, U. of Iowa:I am copying your message to one of my colleagues,Rex Honey, who spent a year in Jordan a few yearsago and had extensive contacts with their basemappeople. I don't recall him ever talking about theavailability of digital maps, though, so I don't knowhow much he may be able to help. Perhaps he cangive a good contact in Amman. I'm sure that will behelpful. I have also received the following note fromArt Getis, San Diego State University: �Thanks forthe information on the Palestinian health project. If myson-in-law were still there (he now lives with mydaughter and two children about one mile from us inTierrasanta), he would likely be the liaison for thehealth care study. Actually, Israel land managementofficials have excellent maps of the region (greatdetail). There is a planning group in Gaza that also hasgood maps.�

3. Response from Grant Thrall, U. of Florida:You might check out Maptitude version 4.0 fromCaliper Corporation (Newton Mass). Caliper is justnow shipping version 4. I understand version 4.0 hasGPS software interlink capability; I have not seenversion 4.0 myself. Again, check out Caliper [for

country files]. I understand that they have upgradedtheir international boundary files, which I believecomes with Maptitude.

4. Response from Dabo Brantley, NCCDPHP:We don't use MapInfo but if you can translate fromArcView GIS, I will scan our disk for the shape filesrequested below. [Dabo sent ArcView files to Brian.Editor]

5. Response from Ron Bijeau, Department ofDefense (referred by Miller Dayton, NCEH): I amresponding to a note from Charles Croner requestingGIS data on your behalf. I work for the Department ofDefense in Atlanta, GA, and have a personal andprofessional interest in seeing that data produced withDoD resources are made available to other FederalAgencies (i.e., CDC). As you may know the primarysource for DoD geospatial data is the National Imageryand Mapping Agency (NIMA). Their charter is toproduce, manage and disseminate worldwide digitalgeospatial data to DoD users and other authorizedFederal Agencies. You may be familiar with a NIMAproduct called Vector Map 0 (VMAP0). It is thesmallest scale component (scale of one to one million)in a series of worldwide multi-layered GIS databases.Obviously at this small scale its application is limitedto regional or broad area analyses. It is distributed byNIMA in Vector Product Format (VPF) orcommercially as Digital Chart of the World inARC/INFO format. (I've not tried these data usingMapInfo yet, but I'm certain it can be done.) Othertypes of geospatial data (i.e., satellite imagery, rasterdigitized maps, digital terrain elevation models, etc.)are available depending on your requirement.

The NIMA office that services non-DoDcustomers can be contacted at (703) 264-3012. Ibelieve Ms. Cindy Burns is the Chief of the FederalAgencies Customer Support Team. I don't know if theycan provide you with the data you need prior to yourdeparture, however, I personally think CDC couldbenefit greatly from these valuable resources in themid to long term. Your project sounds very interestingand I trust you'll find the data you need to make it acomplete success. If there is anything I can do to helpyou further, please feel free to call me at (404)363-5473 or e-mail me at bijeaur@forscom.army.mil.

8

6. Response from Jamie Purvis, ATSDR:Contact Bill Banks at the number below forinformation requested in earlier e-mail from ChuckCroner. From Bill Banks [response to Jamie]: Thanksfor the information Jamie. I believe that I can get all ofthe material that CDC desires regarding these types ofMaps. Have some one call me at (301) 865-4506 or at(717) 593-5266. I can get the Maps and Overnightmail them to you at CDC. No problem!

7. Response from Fred Broome, US CensusBureau: I suggest you contact the Central StatisticsOffice in Jordan. The Census Bureau installed a GISsystem there last year to assist their census activities.Also, they have a Central Geographical Office (notcertain of the name) that has quite a large amount ofdigital data. Maybe the U.S. Embassy can help. They(the AID representatives there) were our contacts.

Editor: Thanks to all for the timely response to Brian�ssituation. He was very appreciative of this assistance.I want to also acknowledge other comments forwardedto Brian from Jonathan Mayer, Harvard U., JohnKelmelis, US Geological Survey and Nina Lam,Louisiana State University.

� From Richard Hoskins, WA State Department ofHealth: We are trying to develop a organizationalmodel for organizing GIS in the WA State Dept. ofHealth. I would appreciate very much any informationwhether an org chart, document you have, or just a fewlines of e-mail which describes how GIS is organizedand funded in your organization. The issues are centralvs decentralized, program funding vs indirect funding,and where it is located in the organization, that is, inInformation Services, off on its own, part of anotherunit or group, etc. Also I would be interested in futureplans, etc., staffing concerns, budget worries, so forth.If you are not involved with GIS organizational issuesI would very much appreciate knowing the e-mail orphone number of the person who is. I guess I could doa survey, but something informal will likely serve myneed. If you prefer, send me a phone number and I willcall you for a brief discussion. Thanks in advance,Richard E. Hoskins, PhD MPH, WA State PublicHealth Geographer, GIS and Spatial Epidemiology

Unit, Office of Epidemiology at e-mailREH0303@hub.doh.wa.gov or voice (360) 705-6050.

1. Response from Tom Richards, PHPPO: Ithink your question is excellent. I do not know theanswer. The most recent GIS state survey report that Iam aware of was in 1992, and so would now be prettymuch out of date. Warnecke L. State GeographicInformation Activities Compendium. Lexington, KY.Council of State Governments. 1992. I have heardyour thought about the need to conduct a state healthdepartment GIS-related survey echoed in severalconversations over the past six months. Examplesinclude: 1) Dr. Mark Oberle, Information Network forPublic Health Officials, CDC, voice (770) 488-2427,has been thinking about doing a survey and/ordeveloping a GIS special focus within the INPHOstates; the INPHO group and NACCHO conducted asurvey about a year ago on electronic computerequipment in state and local health departments, butthis survey did not address GIS capacity; and 2) Dr.Bill Henriques, ATSDR, phone 404-639-6088, hasbeen thinking about doing a survey of state GIS needs.Several other groups also might be interested and/orthinking about doing a GIS related survey . Examplesinclude: the Council of State Governments (potentialcontact: Dr. Keon Chi, phone: 606-244-8247; theAssociation of State and Territorial Health Officials(potential contact: Jacklyn Bryan, phone 202-371-9090, ext 224); and the National Association for PublicHealth Statistics and Information Systems (www.naphsis.com).

If you wanted to do a "quick" informal survey,one additional option that you might consider would beto contact Dr. Littleton Fowler, President of theAssociation for State and Territorial Local HealthLiaison Officials (ASTLHLO) (phone: 405-579-2261;e-mail: Littf@aol.com ). Dr. Fowler has e-mailconnections with liaisons in about 30 states, and(depending on how the question was worded and/or thenumber and types of questions that you have) theliaisons might be able to provide some quick informalinsights. In turn, this raises another question: Shouldeach state health department have a GIS coordinator/contact person -- with e-mail capabilities -- so yourtype of question might be answered more directly?

9

And -- does such a list already exists -- perhapsthrough Chuck Croner and/or the Federal GeographicData Committee (www.fdgc.gov.html) ? I hope thishelps some. Thomas B. Richards, Public HealthPractice Program Office, CDC, at e-mail tbr1@cdc.gov or voice (770) 488-2544.

2. Response from Chet Moore, NCID, CDC:I am forwarding your request to several GIS specialistswho may be able to offer more extensive advice. Mycomments relate only to our local program, as I haveno experience with other programs within CDC orHHS. Also, I am an entomologist/ecologist with astrong interest in GIS, rather than a geographer/GISspecialist, so my experience is limited.

The Division of Vector-Borne InfectiousDiseases (DVBID), NCID, is a small field station, somost of our activities are "stand-alone" in relation toservices available in Atlanta. This is true of our GISprogram, which is now nearly two years old. We haveattempted to build a fairly centralized facility that isshared by the DVBID staff. This makes the mostsense, given the cost of hardware, software, and data.There is no separate, formal "GIS group" or section,but the equipment is located in one room. We do makean effort to coordinate funding requests and relatedissues between the branches. We do not have anyfull-time, dedicated GIS staff at this point, althoughwe would like to move in that direction. At themoment, we have three or four staff with seriousinterest in applying GIS and related methods tovector-borne disease ecology and epidemiology. Toassist us in our work, we draw heavily on the expertiseof faculty at Colorado State University and a USDAlaboratory that has a well-developed GIS program.

We have obtained funding from a variety ofsources: disaster-related activities, EmergingInfections funds, and the regular Division budget. Atleast in the startup phase of a program, I suspect thatcentral funding will be important. As the program(s)grow, decentralized funding may become moreappropriate. As I have never worked with a group aslarge as the Washington Dept. of Health, I don't havea good feel for the issues of organization. I suppose themost important considerations are that it be accessibleto all staff with a need for services, and that the GIS

program should not drift off into its own world.The thing that seems to me to be most

important for successful program development is thatthere should be a firm, long-term commitment at thehighest level. P. A. Burrough, in his introductory texton GIS, makes the comment that "It is simply notsufficient for an organization to purchase a computerand some software and to hire or retrain one or twoenthusiastic individuals and then expect instantsuccess." This brief statement could well serve as amotto for any developing program. Good luck indeveloping your program. If you have additionalquestions, please give me a call or send e-mail. ChesterG. Moore, Ecology/Epidemiology Section, ADB,Division of Vector-Borne Infectious Diseases, NCID,CDC Ft. Collins, at e-mail cgm2@cdc.gov or voice(970) 221-6423.

IV. Special Reports(Submissions are open to all)

� �The National Imagery and Mapping AgencyGlobal Geospatial Information Infrastructure: A

Model and Potential Source for Global GeospatialInformation and Services Support� Ron Bijeau, Department of Defense

The National Imagery and Mapping Agency (NIMA)[formerly the Defense Mapping Agency] is responsiblefor the collection, production, management anddissemination of geospatial information to Departmentof Defense (DoD) customers worldwide. The NationalPerformance Review and some amazing advances intechnology have resulted in some very excitingchanges in how NIMA will perform this mission in thefuture. In order to accomplish its mission, NIMA hasrecognized the need to transition from an "industrialage" production to an "information age" services-basedorganization. One of the most exciting aspect of thistransition has been the apparent broadening of itsmission to potentially include providing geospatialinformation and services support to non-DoDcustomers. The overarching concept for accomplishingthis monumental task is the Global GeospatialInformation Infrastructure (GGII).

Millions of dollars are spent every year toprovide geospatial information and services support to

10

meet the President's national security goals. Althoughthere is no doubt they are required, these data shouldand could potentially be made available to non-DoDagencies to more effectively utilize this valuableresource. The Centers for Disease Control andPrevention (CDC) could benefit from this resource.For the purposes of this discussion let us assume CDC,and its affiliated agencies, centers, and offices, has arequirement for geospatial data worldwide. In somecases the requirement goes down to the census blockor street address level.

The NIMA has been working in concert withthe Open GIS Consortium to develop the GGII. TheGGII is a concept for providing authorized usersaccess to vast repositories of geospatial informationvia existing and future communications networks. It isconceivable that CDC centers or individual projectmanagers, in the US and abroad, could "plug into" theGGII using existing and future informationtechnologies. There are customer support teams atNIMA to coordinate such issues. Centers andindividual project managers in the US could accessgeospatial data repositories from the US component ofthe GGII, the National Spatial Data Infrastructure(NSDI). Due to limitations in NIMA's mission toproduce geospatial data of the US, coordination withother government agencies and commercial sourceswould be required. A liaison or coordinating officewould be required to develop and establish such aprogram. Perhaps such an office already exists withinCDC. This office should possess the followingknowledge:- NIMA organization and policies: NIMA has limitedresources, therefore, a knowledge of what programsand services exist and who to speak to is veryimportant. This knowledge will get CDC's foot in thedoor.- NIMA international cooperative productionagreements and treaties: There is a myriad ofinternational programs that could either help or hinderCDC's efforts to acquire global geospatial dataresources. Knowledge of how to effectively leveragethese agreements and treaties would be required.- Security classification issues: In general there is amove to declassify DoD data sources where

appropriate. Knowledge of these processes andprocedures is very important for gaining access torecently declassified geospatial data sources.- NIMA GGII concepts: NIMA has just released thefirst draft of the GGII Master Plan. This plan is the"blue print" for providing geospatial data directly tothe user. This user may be a center, division, office orindividual in the field with a laptop computer and aGPS receiver. Knowledge of this plan is very importantin building an infrastructure for exploiting the vastresources of the GGII.- USGS organization and policies: For similar reasons,a knowledge of the USGS organization and policieswill ensure successful access to, and employment of,NSDI resources. Knowledge of USGS internationaland cooperative production programs is important forgaining access to these resources.- Meta data: A thorough understanding of nationalstandards for meta data and its value in the successfulimplementation of a geospatial data managementprogram is very important.- Standards and specifications: A commitment tointernational, national, commercial and defactostandards and specifications is very important indeveloping a seamless interface to the GGII.- Developing, designing and implementing computernetworks: Information technology and the effective useof wide area networks (i.e., INTERNET) are the keysto a successful program. Knowledge of what resourcesexist, those planned for in the future and the best toolsto exploit them is critical.- Geospatial data requirement process: Every projectstarts with a clear understanding of the problem to besolved. A clear understanding of the importance ofclearly defined requirements and how they should bearticulated to agencies with the potential to satisfythem is very important.- GIS concepts and technology: A thorough knowledgeof GIS concepts and its application to complexproblems is very important.- Contract management: The general trend ingovernment is to contract out many functions that arenot explicitly the purview of government. Therefore, aknowledge of government contracting and thedevelopment and management of contracts is very

11

important.Data production and management are

considered by many to be the most costly element inthe development of an effective geospatial dataprogram. It would be in keeping with the NPR andmay be well worth the expenditure of resources toexplore this potential resource. With further study itmay be seen that an office established for the purposeof gaining access to valuable DoD and othergovernment-owned geospatial data could potentiallypay for itself many times over. [Contact: Ron Bijeauat e-mail bijeaur@forscom.army.mil or voice (404)363-5473]

V. Public Health GIS Literature(This section may include literature citations, abstracts,

syntheses, etc., and submissions are open to all)GIS research award to Grant Thrall: Grant Thrall,Professor of Geography, University of Florida,president of Spatial Decisions and Analysis, andmember of Geo Info Systems� Editorial AdvisoryBoard received first place for co-authored research ina national competition for a new market analysismodel for senior health care and senior housing. Grantreceived the award at the October annual meetings ofthe National Investment Conference for the SeniorLiving and Long-Term Care Industries (AnnapolisMaryland). "The Development of a Senior HousingExpert Decision System: A Theoretical Model andSystems Framework" (NIC Review, volume 5, 1998,3-16) applies to the senior housing and long term careindustries the technology of GIS, gravity models, anda variation of the seven-step procedure for siteselection introduced in the Geo Info Systems'November 1997 "ShopTalk." Editor [Grant can bereached at e-mail thrall@afn.org or http://www.afn.org/~thrall/]

From Roger Friedman, NIP/CDC (the followingdialogue was picked up from the ai-geostats listserv):Question- I have seen reference to a modeling methodreferred to casually as "ASH." I know next to nothingand have only heard mention of this method as a fastersubstitute to kriging with fewer distortions at theboarders....Response from Gerald Whittaker- TheAveraged Shifted Histogram (ASH) is a kernel density

estimation technique introduced by David Scott,Department of Statistics, Rice University. You mayalso see references to weighted averaging of roundedpoints (WARP), by Scott and Haerdle. The algorithmproceeds in two steps: first the data in binned to arectangular array, then kernel density estimation ornonparametric regression is applied to the binned data.The ASH is discussed in detail in "MultivariateDensity Estimation" by Scott. The ASH densityestimation code is available from STATLIB. The codeis a set FORTRAN routines called by functions inS-PLUS. At the Economic Research Service, we havefound the ASH to be much faster, and generally morerobust than other local regression techniques. A typicalsurface will cover the conterminous U.S. with a largenumber of features. Locfit, Loess and kriging do notdid not do well in application to this data. Anadditional problem is that the data is from stratified,complex design sample surveys. There are no versionsof Loess or Kriging that can deal with stratification,while the ASH deals with this in a simple way. Someresults: 535,000 observations of population at theblock group level from the 1990 census tookapproximately 15 minutes to bin in three dimensions(population, latitude, longitude) to a 600 x 400 grid ona pentium 90. The nonparametric regression then tookabout 5 min. There is a description of this work inScott and Whittaker,"Multivariate Applications of theASH in Regression," Communications in Statistics,v.25, 2551-2530 (1996). [Contact: Gerald Whittaker,Economic Research Service, USDA, at e-mail gerryw2@econ.ag.gov]

From Mike Mungiole, NCHS (mortality datasmoothing work): This study was conducted as initialresearch to determine an appropriate method forsmoothing mortality data maps of the contiguous U.S.and the results of this method were effectively used inthe recently published Atlas of United States Mortality.Past work has shown that the median-based smoothercalled head-banging has been a good method toeliminate spikes and retain edges in two-dimensionaldata (Hansen, IEEE Trans. Geoscience & RemoteSensing, 1991). We were interested in adding datareliability to this algorithm and extended head-banging

12

to include weights which were based on the varianceof the data to be smoothed. When applied to bothactual and simulated datasets, as expected, spikes weregenerally removed while ridges and clusters of highrates near the U.S. borders were maintained by boththe unweighted and weighted algorithms. Addingweights, however, had a substantial effect on resultingdata patterns in that they strongly influenced whetheran observed spike was retained or smoothed away. Inan effort to consider the appropriate degree ofsmoothing, we also manipulated the values of theheadbanging parameters and showed how theyinfluenced the resulting smoothed maps. In summary,this algorithm retained edges along with spikes forreliable data while unreliable isolated data points(spikes) were smoothed toward adjacent, more reliablevalues. For more detailed information, see"Application of a Weighted Headbanging Algorithmto Mortality Data Maps" (by M. Mungiole, L.W.Pickle, K.H. Simonson, & A.A. White) pp. 45-49, inthe ASA 1996 Proceedings of the Section onStatistical Graphics.

Gerald Rushton and Marc Armstrong, Departmentof Geography, The University of Iowa: ImprovingPublic Health Through Geographical InformationSystems: An Instructional Guide to Major Conceptsand Their Implementation, VERSION 2.0 (CD-ROM),November, 1997. The CD-ROM contains learner-centered instructional activities. It also containssoftware for analyzing the geographic pattern ofdisease events. The materials on this CD-ROM showthe key steps in performing a detailed analysis ofhealth data using GIS:- acquiring digital road maps of local areas frompublic domain sources or from private vendors;- acquiring software to match addresses stored inhealth files to the digital map;- making tests of statistical significance forgeographical patterns of diseases;- relating the geographic patterns of incidence tosocio-economic data from sources such as the U.S.Census;- evaluating the geographic pattern of health facilitiesin relation to need; and

- evaluating alternative locations for appropriate healthservices development.

Gerry and Marc explain that their interpretationof GIS differs from the popular �desktop mapping�concept of GIS which is common in many state andfederal programs. �We deal with health, environmentand socio-economic data at many geographic scales ofanalysis, starting with the individual entity. The CD-ROM offers instruction in the matching of healthinformation by addresses to digital maps. After this hasbeen successfully accomplished, software available onthe CD-ROM as freeware (DMAP) can be used toprepare input for GIS software to make isoline mapsshowing areas of high and low disease incidence rates.Alternatively, individually geo-coded health data canbe aggregated to any defined geography, includingcensus entities or other areas of interest--for example,the pollution footprint of a point source pollutant. Weteach the analysis of disease distributions ascontinuous spatial distributions that are tested forstatistical significance. Programs in DMAP areavailable for this purpose.� [Editor: Copies of the CD-ROM �Improving Public Health Through GIS� may beordered by mail from Department of Geography,316J.H., The University of Iowa, Iowa City, IA 52242or by FAX at (319) 335-2725. The cost, includingpostage and handling is $20 per copy ($10 for orders of10 or more)]

VI. Related Census, DHHS and Other FederalDevelopments

A. Excerpts from the June 3-4, 1997 meetings, SanFrancisco, California, "Perspectives on Privacy,Confidentiality, Data Standards and Medical/ClinicalCoding and Classification Issues in Implementation ofAdministrative Simplification Provisions of P.L. 104-191," of the Public Health Service, NationalCommittee on Vital and Health Statistics-EXECUTIVE SUMMARY: The National Committeeon Vital and Health Statistics convened a two-dayhearing on June 3-4, 1997 on perspectives on privacy,confidentiality, data standards and medical/clinicalcoding and classification issues in implementation ofthe administrative simplification provisions of P.L.104-191. The hearing, held in San Francisco,

13

California, was the seventh NCVHS hearing onKennedy/Kassebaum (K2). It was the first to deal withthe full range of issues related to the legislation. Thehearing had panels on insurers, health plans andproviders; public health and research; public hospitals,community health centers and academic medicalcenters; advocacy; integrated health systems;employers; and state health data. NCVHS Chair Dr.Don Detmer said the testimony and discussion wouldcontribute to the recommendations the Committee ismandated to make to the Secretary.

PANEL ON PUBLIC HEALTH AND RESEARCHMs. McCall [Laguna Research Associates]

focused her remarks on the uses of standardized datato improve the health delivery system. Sheencouraged the Committee to use its position toexplore and promote integrated standards that apply toall forms and settings of health care. Standardizationmust be broad-based and include standards for long-term care insurance policies. Dr. Phibbs [Center forHealth Care Evaluation] focused on the potential gainsof linking data from different sources. Herecommended that the standards include a set ofidentifying information rather than just an ID number.

Dr. Flores [Sonoma City Department of HealthServices] stressed the fundamental importance ofmaintaining the trust of all those for whom publichealth is responsible, something that depends on theirconfidence in the privacy and security of their medicalinformation. He described the dire consequences ofProposition 187, with both documented andundocumented immigrants avoiding essential healthcare for fear of jeopardizing their status in the U.S. Herecommended local oversight to control access byresearchers and others, so that local authorities knowwho is using the information.

Much of the discussion with this panel focusedon ways to protect confidentiality and still permitresearch, and on the need to make a stronger case forresearch to the public and Congress. There was a longand inconclusive discussion as to what meaningfuldistinctions can be made between research and lawenforcement in terms of their use of health records.

Panelists agreed that greater scrutiny is required fordata uses that focus on individuals. Mr. Gellmanpointed out that researchers need to clarify the blackand white areas related to confidentiality, and then dealcarefully and creatively with the many gray areas.

Much of the discussion period focused onconfidentiality issues and their relation to research andpublic health priorities. The panel and committeeexplored Dr. Luft's [UCSF] ideas about "approximatevalue" and other ways of protecting confidentialitywhile allowing research to go forward. In a discussionof the difficulties of getting data from health plans, itwas pointed out that plans are improving their datacollection and this trend will probably continue.Asked about ICIDH, Dr. LaPlante [UCSF DisabilityStatistics Center] repeated his point about theimportance of data on functional performance, but hesaid the ICIDH may not be the best standard for thispurpose. He and Dr. Iezzoni discussed the tradeoffsbetween small area research on disabilities and rareconditions to document underservice, on the one hand,and the threats to privacy this research represents, onthe other. Mr. Scanlon suggested that a certificate ofconfidentiality from the DHHS Secretary might addfurther protections for researchers from subpoena.ADVOCACY PANEL

Because of discrimination against gay andlesbian patients and people with HIV and AIDS bysociety in general and by the health care system, theconsequences of information's getting into the wronghands are very serious. Ms. Plumb [Gay and LesbianMedical Association] stressed the need to give toppriority to privacy and confidentiality. Herorganization recommends that the Committeerecognize that only universal access to health care andstrong anti-discrimination laws will protect againstdiscrimination, that it request strong measures toprotect patient information, and that it call for strongpenalties and sanctions for inappropriately divulginginformation.

Ms. Hansen [AIDS Legal Referral Panel of theSan Francisco Bay Area] urged that electronic datacollection and transfer of information not be doneunless/until the strongest possible protections are inplace, in view of past abuses and the evidence of

14

discrimination against people with HIV/AIDS. TheAIDS Legal Referral Panel is concerned that HIPAAseriously threatens patients' confidentiality in that theadministrative simplification provision emphasizes thegoal of reducing the cost of health care and givesprivacy concerns secondary consideration, at best.She urged that constraints be imposed on theimplementation of HIPAA until a comprehensivefederal privacy statute is enacted.

PANEL ON PROVIDERSMs. Forbis [American Association for Medical

Transcription] drew attention to the burgeoning of ahuge, unregulated industry to transcribe dictatedmedical records, with serious implications for theconfidentiality of those records. She called theindustry attention to the fact that some transcriptionservices are becoming data repositories and looking athow to market the information they control. AAMTrecommends that information release forms includeinformation about where documentation is stored, whocontrols it, and the patient's right to review it. Qualityassurance standards for dictation and transcriptionprocess should be developed, and ASTM's standard onsecurity and confidentiality of transcribed healthrecords should be adopted.

Dr. Simons [U.S. Public Policy Committee ofthe Association for Computing] encouraged theCommittee to call on ACM for disinterested technicalassistance. She discussed the current threats toconfidentiality, and called for aggressive Committeeleadership in this area. She strongly criticized the useof the Social Security Number as a unique identifier.Much of the discussion with this panel focused on Ms.Forbis' alert about medical transcription and thegeneral need for new solutions that take into accountthe global nature of information exchange. Dr.Detmer observed that the Committee and thegovernment are being challenged to be unusuallyproactive in their approach to policy planning.

PANEL ON INSURERS, HEALTH PLANS AND PROVIDERS

Turning to confidentiality, Mr. Gellman askedfor comments on proposed legislation. Mr. Matejka

[California Medical Billing Association] asserted thatthe Bennett Bill is very restrictive, and he appealed forclarity in the guidelines about who is entitled access todata on patients. Ms. Franks noted that Californiaalready has strict patient confidentiality laws.Panelists commented on the law's provision thatinformation can be disclosed for billing purposeswithout patient authorization.

Asked to comment on the National ProviderIdentifier, which Mr. Scanlon noted they all seem tosupport, the panelists noted that the longer the numberis, the more room for error and that it will be good toreplace the current "sloppy" systems. Concerns wereexpressed about possible misuse or misinterpretationin connecting providers to mortality data.

PANEL ON PUBLIC HEALTH AND RESEARCHDr. Phibbs (Palo Alto V.A.; Stanford

University; speaking for himself only) focused on thepotential gains of linking data from different sources,something that will be facilitated by unique identifiersand data standardization. He recommended that thestandards include a set of identifying informationrather than just an ID number. To demonstrate thebenefits of data linkage, he described his recentanalysis of neonatal mortality and hospital patientvolume, published in JAMA. Using several datasources increased the predictive accuracy of the modeland yielded substantive changes in the results. Henoted that it will be important to link the emergingclinical data sets to existing data, to remove the effectsof selection bias, among other things. As anotherexample of data linkage issues, he discussed thebenefits of linking data in the V.A. system with thosefrom care outside it, as veterans receive care in bothplaces. Finally, he noted that increased data linkagealso increases the risk to patient confidentiality, and hecalled for reasonable and appropriate steps to protectconfidentiality without denying data access to qualifiedresearchers.

Dr. Flores (Sonoma County Public HealthDirector and President, Health Officers Association ofCalifornia) appeared on behalf of the Health OfficersAssociation. He stressed the fundamental importanceof maintaining trust with all for whom public health is

15

responsible, which depends on their confidence in theprivacy and security of their medical information. Inparticular, the concerns of California's immigrantpopulation about confidentiality and security havebeen severely exacerbated by Proposition 187, leadingto precipitous drops in clinic attendance and the failureto get care for serious conditions. This puts bothindividual and societal health at risk. Dr. Flores alsodiscussed the concerns around disclosure of HIVstatus. He noted the potential advantages to publichealth from linking information systems, benefittingimmigrants as well, but also the considerable risksthrough intrusion by the Immigration andNaturalization Service, insurance companies,apartment owners, and others who would use medicalrecords against people. He recommended localoversight to control access by researchers and others,so that local authorities know who is using theinformation. Localities also need access to informationderived from shared information systems. The goal isto use health information for "systematizedimprovements in health care" and to avoid their usefor "systematized entrapment." In conclusion, hestressed the need to assure immigrants that their healthcare information will not be used against them, and heurged the Committee to "first, do no harm" with itsdata sets.Discussion

Much of this discussion focused on ways toprotect confidentiality and still permit research, and onthe need to make a stronger case for research to thepublic and Congress. The panelists said that some, butnot all, research can be done with nonidentifiable data.Dr. Flores said public health's use of encryptedidentifiers for AIDS information has no disadvantagesfor epidemiology but prevents contact tracing. Specialaccess may be necessary for public health. Mr.Gellman noted that some legislative proposals requirepatient consent before their records can be used forresearch. Given that every user group asserts theirspecial entitlement to access, it is incumbent onresearchers to make the case for research moreeffectively than has been done in the past. Failing that,research is at risk. Several panelists pointed out thatthere are already limits on access to data for research

uses, but they agreed that more public education isneeded. Dr. Detmer noted the paucity of research onthe nature and extent of abuses and problems in thisarea.

Mr. Gellman stimulated a long andinconclusive discussion with a query as to whatmeaningful distinctions can be made between researchand law enforcement in terms of their use of healthrecords. It was posited that the INS and lawenforcement are interested in individuals, whileresearchers generally are not, but this distinction wasdisputed. Nevertheless, it was agreed that greaterscrutiny is required for data uses that focus onindividuals. Ms. Coltin pointed out that a single uniqueidentifier that could be encrypted would obviate theneed for additional data, but Dr. Phipps stressed that adataset is necessary to definitively identify anindividual. Responding to a question, he said it istheoretically possible, but not currently realistic, tohave a reliable unique identifier. Mr. Gellman pointedout that any access to a person's medical record is aviolation of that person's privacy, whether or not itleads to a decision regarding that individual. He alsoobserved that researchers need to clarify the black andwhite areas related to confidentiality and then dealcarefully and creatively with the many gray areas,including this one.

The panelists were supportive of the IRBmodel for health services researchers around theiraccess to individual data. Asked about new provisionsthat require the patient's authorization for research useof medical records, Dr. Phibbs and other panelists saidit is "a disaster."

PANEL ON PUBLIC HEALTH AND RESEARCHDr. Newcomer stressed the need to track the

service utilization of people in group housing,including both elderly and nonelderly disabled people.Currently, although as many people live in grouphousing as in nursing homes, the industry is virtuallyinvisible in national data systems. It also is not clearlydefined or well regulated in state and local ordinances,threatening an erosion of the quality of care. Untiladequate information systems are in place, hospitaldischarge abstracts can be used to track marker

16

conditions. Another possible source is Medicareclaims, to track diagnoses. Neither of these datasystems is adequate, however. One alternative is torely on survey systems, but these are national and donot yield useful community-level estimates.

Definitions are a key problem because the term"group housing" encompasses a wide range of settingsand purposes. To simplify matters, states use a simplerdefinition based on having five or more unrelatedpersons in the same home. Many group homes are notincluded in the National Health Interview Survey andother surveys. Dr. Newcomer provided the Committeewith a critique of 75 surveys, all of which havefundamental problems. He recommended putting andmaintaining better information on hospital dischargeabstracts and claims data as the first step toward betterinformation in this area, and also that NCVHS oranother body look at how to improve the sampleframes used in the NHIS and other national surveyswith respect to group housing. In addition, thedisability information in the American HousingSurvey and the Census must be improved.

Dr. Abbott (California Department of HealthServices, appearing as an individual) noted thatCalifornia has been developing approaches to many ofthe issues in the National Committee's mandate. Hestressed the importance of standards to public healthand research, with identifiers playing a key role infollowing individuals longitudinally and acrossprograms. He then described his state's project todevelop an approach to a unique identifier, and theevolution of a core data set for this purpose with fivecore and seven confirmatory data elements. Thesystem is premised on the patient's voluntarycooperation, and uses stable data items the client canremember. The Department of Health Services has agoal of full implementation by June 1998, and it hopesother state agencies and departments will also adoptthe system. They did not choose the SSN because it isnot reliable and can too easily be linked with otherinformation that can be damaging to the individual.Other stakeholders such as schools were involved inthe process.

Dr. Abbott echoed Dr. Luft's recommendationthat federal standards be viewed as minimums on

which states and localities can build. Regardingconfidentiality, he noted the need to determine whetherdata are confidential in their own right or whencombined with other data. California uses a kind ofIRB to review requests for data by researchers andothers. He described its process and criteria, noting theimportance of public tolerance of the use of healthinformation and thus the need to establish and maintaintrust. He urged the Committee to maintain the balancebetween confidentiality and data access rather thanbeing too restrictive.

PANEL ON HEALTH PLANS AND EMPLOYERSThe PBGH [Dr. David Hopkins, Pacific

Business Group on Health] is one of the most activeemployer coalitions in the country. Because two-thirdsof all the covered lives in the PBGH network areenrolled in HMOs, PBGH formed a negotiatingalliance in 1994. It has done a variety of qualitystudies, leading to efforts to improve health careinformation systems. Its data initiative developed athree-part vision involving computer-based patientrecords, open-architecture systems, and built-in real-time feedback mechanisms. The Business Group alsohas involved a core group of managed carestakeholders in a commitment to build a datainfrastructure, with a short-term focus on adoptingunique patient and provider identifiers and uniformdata standards. PBGH is exercising its leadership ingetting employers to use ANSI enrollment standards.

Dr. Hopkins stressed that for all theseinitiatives, the private sector is dependent on thefederal government to make key decisions. He offeredthe following recommendations: 1) Adopt the SSN asthe core element of the unique individual healthidentifier. Confidentiality concerns can be met throughencryption and other security measures. 2) Do notestablish confidentiality provisions that will inhibit theconstruction of integrated databases for legitimateresearch and evaluation. Patient privacy can beprotected by restricting the use of identifiableinformation. 3) Ensure that provider group identifiersfit the realities of today's marketplace, in whichCalifornia may have the most complex set ofarrangements. Provider group identifiers need to

17

indicate each provider's contractual relationship to thepatient at the time of the encounter, thus connectinghim/her to a particular medical group or IPA. 4) Makeevery effort to meet the mandated deadlines, becauseother stakeholders cannot act until the HHS Secretaryhas determined the standards to be used, and 5) Afterthe standards are developed, publicize them widely.

PANEL ON PROVIDERS [Discussion]Mr. Gellman questioned Mr. Schinderle about

the assertion in his written testimony that the healthcare industry has "an excellent track record" inprotecting privacy. He noted that the only publishedinvestigations in North America (Denver, 1976 andCanada, 1980) showed widespread abuses and lack ofcontrol. Mr. Schinderle acknowledged there are noindependent studies to confirm his feeling that theindustry takes adequate care of patient records. Mr.Kassis repeated his recommendation for accreditationto improve practices and bolster public confidence.Both Mr. Gellman and Ms. Frawley expressed supportfor the idea. Ms. Frawley noted that the NationalResearch Council found that organizations have nodata to validate their claim to be safeguardingconfidentiality. The industry needs to developstandards in this area and hold itself accountable.Dr. Mor suggested that JCAHO, or something like it,might be an appropriate mechanism.

Asked about incentives, Mr. Schinderle said hehad tax incentives in mind, and/or other ways ofcompensating for the up-front cost of conversion. Headded that the costs are offset by savings in a ratio ofabout 4 to 1. Dr. Oliva said the vendors she has talkedto welcome standards and expect they will cut costs.Mr. Kassis cautioned that incentives could motivatevendors to cut corners in order to meet deadlines, andDr. Mor observed that they penalize those who do there-engineering on their own. Mr. Schinderle said thevendor community has been unwilling to respond toCalifornian managed care needs because the other 49states don't need the software yet; now the hope is thatthe legislation creates the critical mass to move thevendors to change their systems to make electroniccommerce possible.

Dr. Detmer pointed out the need for a national

strategy for the health component of the nationalinformation infrastructure, including the computer-based record, help facilitating its dispersion, andfinancial incentives as well as privacy, confidentialityand security provisions and data dictionaries. Thesefactors coalesce as part of a public/private nationalstrategy. He thanked the panelists and organizers, andinvited those wishing to make public comment.

B. Excerpts from the June 24-25, 1997 meetings,Washington D.C. Public Health Service, NationalCommittee on Vital and Health Statistics: Dr. Detmerreviewed the Committee's work on standardizationmandated by the Kennedy-Kassebaum ("K2")legislation, as it has prepared for forthcomingrecommendations to the Secretary. The Subcommitteeon Health Data Needs, Standards and Security heardfrom a total of 134 witnesses in eight days of publichearings, including two in San Francisco, and at fullNCVHS meetings. Dr. Detmer added that theExecutive Subcommittee developed an executive workplan at the San Francisco meeting. Two additional daysof hearings on security standards are scheduled forAugust, with attention to the NRC report. TheSubcommittee on Privacy and Confidentiality heardfrom 42 witnesses at six days of public hearings plusan additional 40 at the San Francisco hearings oncross-cutting issues. Dr. Detmer commented that theeffort to get out of Washington, D.C. and communicatewith people in the West was very worthwhile.

UPDATE FROM THE DEPARTMENT AND HHS DATA COUNCIL

Mr. Scanlon reminded the Committee of thesix major themes in the Council's work plan: healthdata standards, health information privacy, serving asthe focal point for work with NCVHS, coordinating theDepartment's positions on ad hoc issues, developing adata collection strategy for the Department, and healthinformation applications of the national informationinfrastructure (NII). He focused his remarks on the lasttwo of these.

Developing and overseeing an HHS datacollection strategy is a challenge because of theDepartment's size and diversity. The Council is

18

working on a multi-year plan for FY 1997 through2001, covering survey content, periodicity and budgetplanning as well as doing a data "cross-cut" to identifygaps and redundancies. The survey integration plan,on which the Committee has been briefed severaltimes, is part of this broader effort. Its major feature isusing the National Health Interview Survey as the hubfor other population surveys. NHANES will go intothe field in 1998; the Medical Expenditure PanelSurvey will be in the field most of the time. Othercomponents of the data collection strategy are theHHS research planning initiative, strategies for statelevel data, and work on special populations andrace/ethnicity data. An inventory of surveys andsurvey plans will be available on the HHS website.HHS is also reconceptualizing the area of data onproviders, resources and capacities, with reference tothe changing nature and structure of the health caresystem.

The research planning initiative began in HHSlast year, associated closely with welfare reform, thechief question being whether the nation has thecapacity to monitor the changes occurring throughdevolution and reform, particularly at the state level.Developing strategies for state level data is a key effortin the Department -- for example, for a state and localarea integrated telephone survey system. The DataCouncil also has a workgroup on race and ethnicitydata that has been reviewing all the previousrecommendations. It is heading the Department'sreview of OMB's recommendations on Directive 15.

Regarding the NII, Mr. Scanlon noted that theCouncil is charged by the Vice President with leadinginteragency efforts on data standards, privacy,telemedicine, and consumer information. It formed aninteragency working group on telemedicine which islooking at regulatory barriers, co-sponsoring ademonstration project, and developing an inventory oftelemedicine efforts. In the area of consumer healthinformation, HHS has developed a project calledHealthFinder to help consumers get reliable healthinformation over the Internet.

Asked about the evaluation of telemedicine,Mr. Scanlon said the Data Council's broad focus is todevelop a methodology that can identify the boundary

and effectiveness of this technology and to compare itsoutcomes to other methods. Dr. Detmer observed thatthe country lacks, and could benefit from, a robust andcomprehensive strategy for the health applications ofthe NII, which he suggested the Committee can helpdevelop. He noted that the context is global, and theU.S. must see its role in terms of partnerships.

HHS IMPLEMENTATION OF P.L. 104-191,PRIVACY COMPONENT

The Secretary's recommendations for theprotection of health privacy are being drafted and willbe presented to Congress by August 21. The NationalCommittee's recommendations will be fed into thatprocess. In addition, the Department has commissioneda report on privacy and health research, which isavailable on the Data Council's website.

DATA STANDARDS COMPONENTOn June 11, the Data Council heard from the

implementation teams that are developing regulationson the K2 standards. A public meeting will be held onJuly 9 at which the teams' co-chairs will present theinformation and elicit feedback about the Department'spositions. The draft regulations will be ready byOctober, and the final rules in February, 1998. Dr.Braithwaite expressed appreciation for theCommittee's hearings, which have provided importantinput into the Department's recommendations.

The crux of those recommendations is that theANSI X12-N version 3070 standards be used for allbut a few transactions in the claims area. Pharmacyclaims will use NCPDP standards and dental claims,the 837. HCFA's National Provider Identifier andNational Payer ID will be recommended as theproposed standards. In the coding area, the Departmentwill recommend the continued use of ICD-9-CM,adding that systems should be ready to move to theICD-10-CM by 2001. For procedures, the Departmentrecommends the continued use of the ICD-9 Vol. 3 andCPT-4 and HCPCS, but with the goal of moving to anas yet undetermined single procedure coding system by2002 or 2003. Recommended security standards willbe policy (not technical) standards. Implementationguides will be available free on the Internet for all the

19

standards.Discussion: Turning to the unique provider

identifier, Dr. Detmer noted that the Departmentrecommends that an 8-digit alpha-numeric identifierbe assigned to all providers, with identifyinginformation. It is proposed that NCVHS endorse thisproposal and recommend that HHS publish it forpublic comment.

STATE-BASED STANDARDS AND PRIVACYINITIATIVESElliot Stone (Massachusetts Health Data Consortium):All the relationships among stakeholders have beencrafted by the MA Health Data Consortium, and theyall have a vested interest in seeing that the nationalstandards are implemented expeditiously. TheAffiliated Networks do not advocate buildingcentralized regional data repositories, but instead favorstandardizing components of private data bases heldby individual providers. A comprehensive health datasystem will be achievable once national standards arerecommended by the Secretary.Walter Suarez(Minnesota Health Data Institute): Regarding privacyand confidentiality, he noted that Minnesota hasstringent data privacy legislation and argued for anational legislative framework that states can buildupon.

Other Discussion: Mr. Gellman asked thepanelists to elaborate on their comments on federalprivacy legislation and the issue of preemption versusallowing more stringent state laws to prevail. Dr.Suarez spoke in support of a more stringent federalstandard rather than a lower common denominator, sothat states are not inclined to push it higher. Mr. Rubinsaid that despite the obvious tradeoffs, he supportspreemptive federal law that does not allow state-levelcustomization because so much health care isdelivered on a regional basis. Dr. Suarez noted thatstates can act independently in ways such as movingfaster, or implementing more transactions, than federallaw requires. Mr. Stone urged the Committee toconsult with Massachusetts' Attorney General and hisstaff, a good resource, and noted that the mostimportant issue has to do with employer access toinformation.

Returning to the subject of a single nationalprivacy law versus multiple ones, Dr. McDonaldasserted that it is contradictory to support standardsand argue for individual state privacy laws. Mr. Rubinagreed, noting that many Washington health plans arebecoming regional or national. The other panelists saidthey saw this trend in their states as well, along withthe facts that most physicians belong to more than oneplan and patients are expected to move from plan toplan. Mr. Rubin added that people are also starting tosee the merits of standardization in the security area.

DISCUSSION OF PRIVACY RECOMMENDATIONS:

Mr. Gellman presented the draftrecommendations prepared by the Subcommittee onPrivacy and Confidentiality, explaining that thedocument emphasizes the urgent need for federallegislation. It also takes positions on subjects aboutwhich the Subcommittee is in agreement (e.g., liberalaccess for research and public health) and identifies theareas where it is not (e.g., preemption and disclosureswith written authorizations).

STATUS REPORT ON OMB DIRECTIVE 15 REVISION

Dr. Detmer welcomed Katherine Wallman,Chief Statistician at OMB. In view of the fact thatOMB had not yet released the recommendations of itsinteragency committee, she provided the Committeesome background and a general idea of what theycould expect. The administration has had standards forcollecting race/ethnicity data since 1977. Particularlysince the 1990 census, the standards have beencriticized as not in tune with changes in thedemographic composition of the population. In 1993,OMB undertook a comprehensive review of thestandards in cooperation with other agencies that useand produce data, to culminate in recommendations forany modifications to be incorporated beginning withthe next decennial census. Decisions must be made byfall 1997.

The agency has made an effort to go out todiverse constituencies and elicit their concerns and

20

suggestions, and also to ensure that extensive researchwas conducted. There were four public hearings andan extensive Federal Register notice with commentperiod in summer 1994, resulting in testimony from100 people and some 800 letters. OMB publishedanother major notice a year later that provided asynopsis of the issues raised and laid out researchplans. The research involved three major studies: theCPS supplement, the National Content Survey, and theRace and Ethnicity Targeted Test. There also havebeen focused studies in specific areas such as health,education and civil rights.

The research results and public commentshave been studied by the interagency committee,which has just issued a report and recommendations.This will be the subject of the Federal Registerannouncement to be released in early July, followedby a 60 day comment period. On the basis of thesecomments and further analysis, OMB plans to releaseits recommendations by mid-October. Ms. Wallmanstressed that the standard is intended as a minimum setof categories for cross-government use; more detailedcategories can be used as needed in specific contexts.

RECOMMENDATIONS ON PRIVACY AND CONFIDENTIALITY

Mr. Gellman presented the revised draft of therecommendations, prepared by the PrivacySubcommittee in response to the previous day'sdiscussion. Its major points remain the same: that thenation is in a health privacy crisis, that privacyprotection has eroded over the last 20 years, and thatthe issue urgently needs attention. It asks forlegislation by the end of this Congress. He and Ms.Frawley described the changes in detail; in some, theystrengthen some recommendations (e.g., ondiscrimination) and clarify various positions in theareas of disagreement. Dr. Amaro spoke in support ofDr. Harding's concerns about special protections,noting that members of less advantaged groups are atthe greatest risk of loss of privacy as well as at high

risk for such sensitive conditions as substance abuse,STDs and HIV. She urged that the Committee offerlanguage to move toward solutions. Mr. Gellmanresponded that despite the widespread awareness ofthis problem, no solution has been found. Severalmembers observed that here the Committee isstruggling with the result of social inequities, includingthe lack of universal health care. Dr. Detmer expressedagreement with these concerns but stressed the need tomove forward with the overall recommendations andnot let them founder on this issue. He noted that thefull Committee, the Privacy Subcommittee, and thoseworking on security, as well as parallel processes in theDepartment, will continue to struggle with theseconcerns. Non-discrimination policies are critical,given society's decision to limit access to basic healthcare. Dr. Detmer expressed agreement with theseconcerns but stressed the need to move forward withthe overall recommendations and not let them founderon this issue. He noted that the full Committee, thePrivacy Subcommittee, and those working on security,as well as parallel processes in the Department, willcontinue to struggle with these concerns. Non-discrimination policies are critical, given society'sdecision to limit access to basic health care.

Net Site(s) of Interest for this Edition: You maywant to visit a very well organized instructionalpresentation on infectious disease epidemiology andrelated topics at http://www.sph.umich.edu/~jkoopman/802Web/. Start with Chap1.htm. The coursewas prepared by Jim Koopman MD MPH, Dept. ofEpidemiology SPH-1, University of Michigan Schoolof Public Health, 109 Observatory St., Ann Arbor, MI48109. [Jim may be reached at e-mail jkoopman@sph.umich.edu or voice (313) 763-5629]

21

Final Thought(s)

These are exciting times for GIS and public health. There are many opportunities to collaborate and work togetheron a wide range of public health issues involving GIS science. I am pleased to spotlight one such opportunity withthe following communique received from GIS colleague Susan Perlin: �Dr. Susan Perlin is leading an effort atthe Environmental Protection Agency�s National Center for Environmental Assessment to conduct a study onchildren�s respiratory health using the National Health and Nutrition Examination Survey (NHANES-3) data. Aspart of the study, we want to use GIS technology to overlay environmental data with the NHANES-3 data. Wewould like to discuss this type of study with researchers who have experience with using GIS to analyze healthdata in conjunction with environmental exposure, or exposure surrogate, data. We would be particularly interestedin hearing from individuals who are involved in this type of study and are using NHANES data. We would beinterested in possible collaboration, too. [Contact: Susan Perlin at e-mail perlin.susan@epamail. epa.gov or voice202-260-5877]

Charles M. Croner, Ph.D., Editor, PUBLIC HEALTH GIS NEWS AND INFORMATION, Office of Researchand Methodology, National Center for Health Statistics <cmc2@cdc.gov>. Copyright Notice: This report is in thepublic domain but its contents are not to be altered or changed without prior written approval of the editor.

August 17-20, 1998 �GIS in Public Health� Conference...Plan Ahead Now