GIS issues pursued in these applications involve downloading and importing spatial data into ArcGIS and projecting map layers based on the application and geographic scale at hand.
Chapter 5: Downloading and preparing spatial and tabular data
Where are the concentrations of a city’s older houses that are likely to have lead-based paint? Do children who have elevated levels of lead in their blood live in those areas? The purpose of the study in chapter 5 is to identify clusters of children who have elevated blood lead levels for the targeting of lead-screening programs. You work with elevated blood lead level samples that have been aggregated to census tracts and census tract data on housing built before 1970, when lead-based paints were still used.
In chapter 5, you download and prepare US Census data. You must clean up the data by renaming variables and deleting rows that do not conform to data table formats and by modifying census tract identifiers in the table so that they match comparable identifiers in the census tract map layer. Then you join them to a downloaded census tract boundary map. Ultimately, you will produce a very nice bivariate map using choropleth and dot-density displays. In that map, you will place randomly located points within polygons in proportion to an attribute of interest. The tutorials in chapter 5 seek additional explanation of observed clusters by using additional census variables to spatially explore a public health problem.
Chapter 6: Geocoding tabular data
General spatial information is available from basemaps, but how do you format the data your organization has so it can be converted to points on a map? Data of interest often includes point locations of patients’ or clients’ residences and health care or other service delivery locations, such as the scenes of traffic accidents. If data includes street addresses, ZIP Codes, or other spatial identifiers, GIS has the tools to plot points of interest for use in analysis.
For the example in chapter 6, you need to spatially enable data that is in tabular form. You’ll geocode existing facilities within a county so they can be placed as points on a map. You also place patients on a ZIP Code map over a wide area. Having this data mapped, you can readily see potential service gaps. Then you map suitability measures for a health clinic location to aid in identifying locations in gap areas that would be an attractive site for a new facility.
Chapter 7: Processing and analyzing spatial data
What are some neighborhood factors that lead to child pedestrian injuries in a city? Is poverty a factor? What about the lack of safe public areas for play? In chapter 7, you explore the determinants of serious juvenile-pedestrian injuries for the purpose of designing prevention programs. The basis of your study is a sample of serious-injury data that has been geocoded and can be compared to census data on poverty and to map layers for streets, neighborhoods, and parks that have playgrounds and playing fields.
The GIS work in chapter 7 includes preparatory steps for extracting study region maps from county maps, and then focuses on detailed proximity analyses using park buffers, like those seen in figure 1.2.
Chapter 8: Transforming data using approximate methods
How can health-care analysts combine data from different, incompatible polygon boundary sets? Chapter 8 explores how to transform this data so it can be used for comparative analysis. Often the spatial unit of analysis for a health study is a custom set of polygon boundaries designed for the phenomenon at hand. An example is the hospital service areas and hospital referral regions used in the Dartmouth Atlas of Health Care Project (http://www.dartmouthatlas.org) at the Center for the Evaluative Clinical Sciences at Dartmouth Medical School in Hanover, New Hampshire. Although appropriate for studying patterns in the quality of health care across the country, these custom areas have the limitation of not sharing boundaries with census statistical areas (that is, they are noncoterminous sets of boundaries). Thus, census data cannot be used directly for supportive analysis of these custom areas and must be spatially apportioned to the noncoterminous boundaries.
Another common case of noncoterminous data involves regional analysis of spatial areas such as emergency management service zones for a city where such zones become the de facto unit of spatial analysis. Detailed census variables on income, poverty, educational attainment, and so on are not easily attainable at these levels. Advanced GIS functionality such as using spatial joins, however, can produce some very accurate approximations (or apportionments) for transforming data from one set of polygons to another incompatible set.
Chapter 9: Using ArcGIS Spatial Analyst for demand estimation
Chapter 9 is an introduction to the ArcGIS Spatial Analyst extension. Spatial Analyst uses or creates raster datasets composed of grid cells to display data that is distributed continuously over space as one continuous surface. In this chapter, you prepare and analyze a demand surface map for the location of heart defibrillators in Pittsburgh, where demand is based on the number of out-of-hospital cardiac arrests in which potential bystander help is available. You also learn how to use ArcGIS Spatial Analyst to create a poverty index surface combined with several census data measures from block and block group polygon layers.
Chapter 10: Studying food-borne-disease outbreaks
Chapters 10 and 11 provide a change of pace — opportunities for you to apply and extend the GIS skills and health applications you have learned in the previous nine chapters to new case studies that you develop. We provide the source data and guidelines for analysis, as well as a broad outline of steps; however, it is up to you to carry out the GIS work on your own in an independent case study. In chapter 10, you prepare map layers, including geocoding incidence addresses as the basis for analyzing outbreaks of food-borne illness. Then you use data to simulate the impact of an outbreak. You also do a proximity analysis based on patterns in reported disease cases.
Chapter 11: Forming local chapters of ACHE
Chapter 11 concludes this workbook with a second independent case study, following a setup similar to that in chapter 10. Staff members of the American College of Healthcare Executives (ACHE) want you to use GIS to help them set up ACHE chapters across the country that provide educational and other services to health-care professionals.
In chapter 11, you perform a buffer analysis of existing affiliates that propose becoming ACHE chapters. The buffers will help determine the territories that are served as well as the gaps that suggest where new chapters should be established. You do some work interactively using ArcGIS, but for steps that must be done repeatedly over time, you build an ArcGIS model that generates a macro to automate these steps.
Introduction to ArcGIS and map documents
This book is designed for use with ArcGIS 10.2 for Desktop software. ArcGIS is a full-featured GIS software application for visualizing, managing, creating, and analyzing geographic data. The more advanced levels of ArcGIS offer advanced data conversion and geoprocessing capabilities. ArcGIS has numerous extensions that include ArcGIS 3D Analyst for three-dimensional rendering of surfaces, ArcGIS Network Analyst for routing and other street network applications, and ArcGIS Spatial Analyst