Staying on the Map Without Losing Any Ground
By updating its GIS, the Arkansas Department of Health has eliminated the need for a full-time programmer and simplified day to day operations.
By Karen Howard

The mission of the Arkansas Department of Health calls for the protection and improvement of the health of the people of Arkansas, as well as the enforcement of regulations for environmental health protection and assurance of quality care. Among the many department-sponsored public health programs designed to serve the people of Arkansas are the Immunizations, Maternity Planning, Women, Infant, and Children (WIC) Services, and Environmental Health Protection Programs.
      To provide such services at the local level, the department has over 90 local health units across Arkansas' 75 counties. One of the core challenges faced by the department is to stay abreast of the whereabouts of its citizens and the geographical areas needing more focus, such as more services or health education. But, with a state population in excess of 2 million people and many in rural areas, how do you do this accurately and efficiently, based on the resources we have?
      The Bureau of Environmental Health Services (BEHS) has been using GIS since 1989 to graphically depict and geographically analyze data not only for environmental programs, but for other programs in the agency like Health Statistics, Epidemiology, WIC, and Blood Lead Monitoring. BEHS was at somewhat of an advantage because we had seen GIS technology successfully used in parcel-based mapping, illustrating its promise as a spin-off for broader use. We discovered that with GIS, we could input into the system things such as monitoring sites, hazardous waste sites, Local Health Units, and various establishments that BEHS licenses or inspects.
      Until we started using GIS, one of our biggest problems was not having meaningful information at the local, community, or neighborhood levels. We were only able to show county-based statistics. Using GIS now allows us to answer questions more clearly, such as "How many?", "How far are they?", "What else is there?", "How many within 1,350 yards or five miles?", "What ground water is affected by chemical run-off?" and "Where is the nearest health unit from my work?" Obtaining the answers to these and other important questions, not in a black and white data table, but in full color graphics, illustrates the power and potential of this exciting technology.
      The most basic human need, water, served as the impetus for upgrading our paper-based geographic mapping to GIS. Our Division of Engineering needed an advanced system for monitoring and protecting the ground water, essentially, the Public Drinking Water Systems. The EPA (Environmental Protection Agency) would frequently update chemicals and safety levels that needed to be monitored for the Safe Drinking Water Act regulations. This added to our already overwhelming amount of information.
      There were many needs and hopes the Division of Engineering had for GIS. In addition to the monitoring points, they needed to be able to see the boundaries of the Public Water System Service Areas (i.e., where the last service connection for the Water System Utility is), what features were upstream or downstream, where hazardous industries are located and their relative proximity to drinking water sources. The Well-Head Protection Program needed to be able to "draw" a protection zone around a well-head and see all features on the land in and around the protection zone. In addition, they needed the help a GIS could provide in looking at the characteristics of the land in and around the water systems. This is crucial because the division reviews plans for land applications of chemicals or wastes to ensure that these applications will not produce hazardous run-off into drinking water supplies. Maintaining accurate records of things like these are absolutely vital in preventing chemicals/pollutants from entering the public's drinking water.
      Given that we were just starting to use GIS for the first time, we were running on an in-house mini-computer with a GIS developed for that platform, and the fact that we did not invest on the training or programming side of our endeavor, this initial project took us a very long time to complete, and was a struggle. The transition from paper maps, and old latitude and longitude coordinates that had never been checked for accuracy, was initially a trial and error process. Our first challenge was to separate the geographic data from the program data, but keep a relationship intact between the two pieces of information. Secondly, we had to learn some basic mapping terms related to topographic maps, coordinate systems, and electronic digitizing. We hired a cartographer to perform geographic data entry services of transforming our paper maps and map-related data to an electronic form. After we started, we realized that much of the geographic information could have been purchased already in an electronic form and we did invest in a couple of map layers, but some of the information was not quite as precise as the other data we had already digitized into the system.
      We learned some very interesting things during this process. From the beginning of our monitoring of these sites, our field engineers would have to note their latitude and longitude coordinates by hand and then later input them into character based computer databases. When we imported these coordinates into the GIS, we discovered we had sites all over the map, literally. We had coordinates show up in Oklahoma, Texas, and Louisiana. Needless to say, the GIS quickly informed us of just how inaccurate some of our information was. Assessing this by hand by taking our hardcopy maps and looking up all the coordinates would have taken an inordinate amount of time. GIS was already on its way to paying for itself.
      Looking back on these early efforts, even though we are only talking about a few years ago, I now call that our "old days" of GIS. As technology changed rapidly over to client/server PC-based systems, and more robust GIS solutions for the desktop computing environments were developed, our old technology began holding us back. Our needs were simple. The Wang-based GIS we started out with wasn't effective because users couldn't access the data they needed on an individual level. It also required a full-time programmer, something we wanted to avoid.
      The solution was a client/server, Windows-based system, that could be placed on users' desktops, with as many standard functions as possible. Many requirements of the users are the same-query the data, overlay the map layers, zoom in, zoom out, produce a report, produce a map. We chose a "Windows-based product called WinGIS, from the PROGIS Corp. out of Bellingham, Wash. One of the major advantages this product had over others that we reviewed at the time was the fact that it was developed under Windows-not transitioned from DOS to Windows. It continues to be flexible in an open systems environment and is simple and straight-forward for our users.
      Now you do not have to have or be a cartographer or computer programmer to implement a GIS solution. The only thing the user needs to "bring to the table" is a thorough understanding of their data. One you determine what you want to see on the map, the software does the rest. The geographic features data can now be obtained from a multitude of sources and importing that data into the system is accomplished with software utilities.
      One of the agency's larger programs, the Women, Infant, and Children (WIC) Program, came up with a challenge for our GIS - find where there are areas of potentially eligible clients and geographic areas that may be in need of services. This was a three-step process: (1) find where current participants are located, (2) examine Census data to determine where geographically there are potentially eligible populations; and (3) overlay these pieces of information to show to the local staff.
      By combining a technique called address matching or geocoding with our GIS, we are now able to take our client addresses and assign a latitude and longitude coordinate to that address of a point on a street segment. For example, the 4800 to 5200 block of Main Street is a segment of Main Street. The address of 4815 Main Street would be at a point on that street segment and each street segment is made up of beginning and ending latitude and longitude coordinates for the end points of the line segment. Using address matching, we were able to locate 96,000 WIC participants and import those points into a map layer in WinGIS. We also purchased and imported the Census Block Group geographic boundaries into a map layer. The WIC Program is now able to "see" their data at levels they never could before and use the data with greater efficiency and impact in planning for the future. Simple queries like "Show me all of the Census Block Groups where there are greater than 50 persons below the poverty level and color these areas red," when combined with the map layer of the WIC participants, shows the program immediately where there are areas that may need to be targeted for additional service.
      GIS has made a remarkable difference in terms of time, especially when dealing with data (millions of records) on a statewide level. Old tabular reports showing total numbers of records by zip code or county just do not have the same impact as those we can now generate at the desktop level using GIS and a color plotter or printer. Having the ability to focus on data at the neighborhood, street, or block levels aids us tremendously in determining specific areas needing our services. GIS is the only way to see these vast amounts of data at these levels. As a state agency, this helps our bottom line because we can target our efforts more effectively and efficiently.
      We can look at potential sites where a mobile health unit might serve the public's needs better than a permanent, stationary clinic. And, we can focus on areas most in need of immediate health education. GIS makes this happen. We have been very pleased with the WinGIS product we chose for our GIS solutions. Our projects to date have met with great success and the PROGIS Corp. has been very responsive to our needs. They asked for feedback about what features we, as users, needed in future software versions and then actually incorporated most if not all of these needs. We have put the software through some tough paces with the varied nature and size of the projects to date and have only had one problem with a feature type called donut polygons. PROGIS quickly responded to the problem and provided us with a custom solution. That kind of responsiveness makes the difference in achieving the results we have enjoyed.
      Successful GIS projects can now be accomplished without a staff of geographers or computer analysts. The software is very intuitive which reduces the intimidation factor for many users. Once you determine what you want to do with your data, you simply learn the functions that each icon performs. Our users who had already used and were familiar with Windows were up and running in less than a day.
      The experience we have had here at ADH over the last several years has clearly demonstrated that the most important factor is making the information available on the user's individual desktop. When data is available at the individual level, the user can process the information they need, when they need it, ask their own questions of the data and generate their own reports. This is simply invaluable.

About the Author:
Karen Howard is a senior program analyst with the Bureau of Environmental Health Services, Arkansas Department of Health. She may be reached at 501-661-2586.

Back