GIS – The power of deduction just not representative observation

GIS is a powerful tool to represent straight out facts visually. But also a tool that can be used to rise new questions. The application is particularly useful in mixed method approach of a research study.

For example, CDC produced the following data and map on the rates of HIV diagnosis among adults and adolescents.


The maps show the distribution of new HIV cases being higher in the west, the northeast and the south. It shows states of Florida, Georgia, Nevada, Louisiana, and DC are among the top five. These facts among few other obvious ones show the observatory power of maps.

The map also leads to so many research questions :

  • Is there a cultural factor that influences high rates in Nevada, given it has a very indigenous cultural makeup?
  • What do Alabama and Mississippi do better to control the rates, when they are the geographical adjacents to LA and GA.
  • Are there specific policies in DC that help people with HIV lead a better life – policies that help them deal with the stigma.

The are few questions among many that the map brings out.

Hence GIS can be used to represent data but also can be used as a tool to form research questions in mixed method analysis.



GIS in developing countries

GIS a relatively new and is finding its way to improve health. Developing countries deal with more technological challenges to make use of GIS. But nonetheless, a very effective project with implication and impact on the local community of Ahmedabad, India is shown below. The project was impactful – was able to communicate the uneducated & also the educated mass on the issue. It went on to be heard by local policymakers too.


The above map developed by UMC in help with AMC staff shows the spatial distribution of slums and community health centers across the city.

UMC staff visited AMC’s existing UHCs and CHCs to understand their functioning and understand the requirements for upgrading facilities. UMC developed a methodology which involved meetings and interviews with health staff including medical officers, pharmacists, lab technicians, multi-purpose workers (MPW) etc. Separate SWOT analysis was conducted with medical health workers and with pharmacists, lab technicians and MPWs. UMC with assistance of AMC staff, identified slum pockets and communities on the base map of Ahmedabad and also marked existing health facilities provided by the AMC. These maps were later transferred to a GIS environment to analyze the accessibility of health facilities by slum dwellers. This assisted in locating newer health facilities as proposed under the NUHM. UMC also has prepared model layouts for the new proposed health centres for the AMC. A detailed phase-wise budget and a proposal was prepared for the AMC for submission to Government of India under the NUHM.

Source: Office of Urban Development Authority of Ahmedabad, India.

Importance of Granulation (large scale map) in Maps


I created the above map using the tutorials in ESRI website for ArcGIS & the content from GIS Tutorial for Health. The map explores the Mammography clinics in relations to counties in Pennsylvania. The pattern of high concentration around cities of Pittsburg and Philadelphia is evident from the maps.

Few other observations from the map are:

Potter and Sullivan counties have fewer women aged 40-74, but still, there are no clinics. They are obvious areas of the state where clinics are needed.
Monroe, Clearfield, Jefferson counties have higher women aged 40-74, but a relatively
small number of clinics.
Philadelphia & Pittsburg surrounding areas have enough clinics, but remote northwestern and northeastern counties need more clinics.

The power of GIS can be further explored to look into the cities that sound to have more mammography clinics, in the map below :


This map shows that though Allegheny county hosts Pittsburg, there is a pattern of concentration of clinics in the south relatively more urban part of the city. The pattern correlates with other healthcare facilities in the county that counts towards health equities in this county.

Granulation to the smallest unit possible brings in more refined data on what seems to be different in small scale.

Applicaiton of GIS in Health: Proximity Analysis

GIS is a powerful tool with various applications. One very useful of such is the ability to produce proximity analysis. It essentially gives an idea of how close one variable is to another, cartographically. For example, look into the below map.

Proximity Analysis of Injury with playgrounds, Pittsburg, PA


I created the above map using the tutorials in ESRI website for ArcGIS & the content from GIS Tutorial for Health. The above map shows the location of injury to residents and looks for how it is related to the playgrounds in the city. A simple analysis using ArcGIS software yielded the following results:


This helps us have an idea that most of the injuries happen away (1200 feet away, if not atleast 600 feet away) from the playground. Another interpretation of the results would be: 16% more injuries occur more than 600 feet away from the playground as compared to within 600 feet.  we can conclude by simple spatial reading that playgrounds have a probable protective effect against injuries. Let promote playground on the week of the French Opens!

#active lifestyle – happy lifestyle#

Do Socioeconomic Factors Influence Texans’ Decision to Get Vaccinated? – A cartographic Approach

Texas has one of the highest vaccination rates for childhood diseases overall, 97.4%, according to CDC. But the number of children not vaccinated because of their parents’ “personal beliefs”—as opposed to medical reasons—has risen since 2003, when such exemptions were introduced, to more than 44,000 so far in 2017 according to CDC. The 4:3:1:3:3:1:4 series is an overall measure that encompasses many vaccines that are recommended for children. Various demographic factors (sex, gender, race, availability of commercial health insurance) influence the decision to get vaccinated, were looked at.

Gis steps

The county-level data on the socioeconomic factors were obtained from US Census Bureau (American Factfinder). The health insurance data was obtained from Small Area Health Insurance Estimates (SAHIE). The vaccination rates were obtained from Texas Immunization registry through DSHS. The data was cleaned and geocoded to be analyzed in ArcGIS to produce maps as shown in Figure 1. Pearson’s correlation coefficient was used to analyze the relationship between vaccination rates and independent variable.


The non-vaccination rates are higher around the major cities of Dallas, Austin-San Antonio, Houston and some northwest Texas counties. Population density has a positive correlation with the non-vaccination rate. Other demographic factors have a positive correlation in certain counties as opposed to others.


Source: American FactFinder, Texas Immunisation Registry

The limitation on the immunization data is it being an optional registry so it would not be accurate to run statistics off this information to estimate an immunization rate. In future, it is productive to expand this concept to use regression analysis to try to find the odds of the relationship expressed in the maps and to find if there is a significant association.