Mapping Nationally Designated U.S. Cancer Centers

Updating National Cancer Institute and National Comprehensive Cancer Network maps to interactive web maps with live-update features.

Julia Donlon

Abstract: My primary interest for this project was to practice building a dataset of known locations which could be displayed as an interactive web map. I am interested in public health applications of GIS and believe that well-designed web maps are important resources for healthcare accessibility and informatics in the digital age. My guiding spatial and design question was: how can I improve the market quality of web mapping for U.S. comprehensive cancer centers and laboratories?

What is the current online market for mapping cancer care centers?

The National Comprehensive Cancer Network (NCCN) is a "not-for-profit alliance of 33 leading cancer centers dedicated to improving and facilitating quality, effective, equitable, and accessible cancer care."

This map is technically informative, but the names make it busy, and the static image is lacking in interactivity and ease of engagement.

Member institutions are listed alphabetically below this map, linking out to different webpages where more information about them can be found.

The National Cancer Institute (NCI) has designated this map of cancer centers which "deliver cutting-edge cancer treatments" to patients across the U.S.

This map is more visually engaging than the former, but I believe it can still be improved. The website publishes this static image, coded into three types: comprehensive care centers, cancer centers, and basic laboratories. Below the image, centers are listed by state, where their information is linked to a separate webpage.

In this project, I am interested in creating an interactive web map which allows the user to zoom to their area, select a nearby center, and view important data related to it. What if users did not have to perform various alignments, first looking up their state, then thinking about which cities are closest to them, then following the resulting links, to access cancer care information? My vision for an interactive map streamlines user engagement to a single click, occasionally preceded by a swipe to zoom.

Methods and Workflow

My  Google Sheets dataset  of cancer centers' addresses, phone numbers, websites, classifications, and coordinates.

Dataset Creation: I created a dataset which merges data from the NCI and NCCN maps by hand. This was the majority of the tedious work for this project: I took information which was largely decentralized, scattered across various websites, and centralized it into a single editable document which provides the basis for my final web map.

Step 1: Click links on the NCI and NCCN websites to find website, phone, and classification information. Input this identifying information to the spreadsheet.

*It would have been possible to do this by geocoding the way we learned in lab, through the Stanford locator extension and OpenRefine, but that system requires finding an address anyway and since Google Maps has such a seamless coordinate feature, I chose the quicker option. I verified again with the final MapBox basemap, where most cancer centers are labeled if you zoom in near my points.

Step 2: "Geocode" center location using Google Maps (and cross-referencing addresses with website information). Enter address and coordinate information into the spreadsheet.

Web Map Creation: I used my merged dataset to create an interactive version of the NCI map, which is color coded by center type.

Blue centers are NCCN members (comprehensive care centers).

Red centers are listed simply as "cancer centers" (non-comprehensive) by NCI.

Yellow centers are listed as basic laboratories by NCI.

Clicking on one of the dots results in a pop up with the name of the center, its address, phone number, classification (one of the three NCI categories), and a link to its website.

Step 1: Edit the web map code from the Week 7 Lab (Web Mapping with Map box GL JS) in a plain text editor, replacing the CA Haunted Places dataset with the new Google Sheet containing locating and identifying information for each cancer center.

Step 2: Upload the .html file to a new GitHub repository, where any future edits to the file can now easily be made by clicking the pencil icon in the top right and "commit changes..." when finished. Aspects I tweaked from the original code included:

  • Color coding — I created another variable in my sheet which I coded to determine the color of the dot, in order to match my interactive map to the NCI map, which classifies centers into three categories.
  • Pop-ups — I wanted to include different features in my pop-ups, which appear when the user clicks on a dot. I had to format the code to include a link to an external site, unique to each pop-up, accessed from my spreadsheet data, and had to manually adjust the address formatting. I learned the intricacies of html coding for formatting and aesthetics here!

The Final Product

I successfully created an interactive web map which resembles the static NCI map, with three cancer center classifications, but which is much more user-friendly and streamlined than both of the original maps.

User Efficiency: The user can now zoom to their area on the map and click around to nearby dots to learn more about the centers near them. Website links are directly provided in the pop-ups, and users can still visually search the map for the appropriate center type for their needs based on color.

Organizational Efficiency: When centers are added, addresses, phone numbers, or website hyperlinks changed, or more information is desired, the organization can easily update the spreadsheet data source and these changes will be almost instantly enacted on the web map.

My final web map: An interactive map suitable for auto-updates using cloud spreadsheet source data. Accessible, navigable, and useful.

I am very excited to transfer these skills to a bigger undertaking in public health research this summer!

Data Sources

References

National Comprehensive Cancer Network. "Member Institutions." Accessed June 10, 2024.  https://www.nccn.org/home/member-institutions/ 

National Cancer Institute. "Find a Cancer Center." Accessed June 10, 2024.  https://www.cancer.gov/research/infrastructure/cancer-centers/find 

Maples, Stace, "Geocoding with OpenRefine & locator.stanford.edu," (2022), Earthsys144,  https://github.com/mapninja/Earthsys144/blob/master/Labs/Week_07/Geocoding_with_Locator.md 

My  Google Sheets dataset  of cancer centers' addresses, phone numbers, websites, classifications, and coordinates.

*It would have been possible to do this by geocoding the way we learned in lab, through the Stanford locator extension and OpenRefine, but that system requires finding an address anyway and since Google Maps has such a seamless coordinate feature, I chose the quicker option. I verified again with the final MapBox basemap, where most cancer centers are labeled if you zoom in near my points.