Photo Credit: Global Diaspora News (www.GlobalDiasporaNews.com).
Isaac H. Michaels, MPH1; Sylvia J. Pirani, MS, MPH2; Alvaro Carrascal, MD, MPH1 (View author affiliations)
Suggested citation for this article: Michaels IH, Pirani SJ, Carrascal A. Visualizing 50 Years of Cancer Mortality Rates Across the US at Multiple Geographic Levels Using a Synchronized Map and Graph Animation. Prev Chronic Dis 2020;17:190286. DOI: http://dx.doi.org/10.5888/pcd17.190286external icon.
Static display of the change in US cancer mortality rates from 1968 to 2017. [A text version of this figure is also available.]
We developed the synchronized map and graph animation to visualize changes over time in yearly, age-adjusted, cancer mortality rates at the county, state, and national geographic levels for the United States from 1968 through 2017. The goal was to enable viewers to select trends of interest for a particular state, region, or time interval.
Cancer is the second leading cause of death in the United States (1). An estimated 42% of all cancer cases and nearly one-half of all cancer deaths in the United States are attributable to modifiable risk factors (2). Health officials and stakeholders need visualizations of data on cancer deaths to target prevention and treatment efforts optimally.
One way of showing changes over time in spatial data is to present side-by-side maps, each map representing data for a different time during the period (3). Animations have been used to present changes over time more granularly than static maps (4). To improve this method and accommodate data for multiple geographic levels, our project proposed a novel technique for visualizing temporal trends in spatial data — presenting an animated choropleth (thematic) map alongside a synchronized animated horizontal bar chart. We demonstrated the method by using data on cancer deaths from 1968 through 2017 in the United States at the county, state, and national geographic levels.
Age-adjusted mortality rates, stratified by year, for counties, individual states, and the United States were obtained from the CDC WONDER (Wide-ranging Online Data for Epidemiologic Research) website, including its Compressed Mortality File 1968 through 1978, its Compressed Mortality File 1979 through 1998, and its Multiple Cause of Death file for 1999 through 2017 (5). Cancer deaths were defined as deaths with any malignant cancer listed as the underlying cause. Malignant cancer was indicated by the International Classification of Diseases (ICD-8) codes 140–207 during 1968 through 1978, by ICD-9 codes 140–208 and 238.6 during 1979 through 1998, and by ICD-10 codes C00–C97 during 1999 through 2017 (6). Three ICD case definitions were necessary to compare data for all years during 1968 through 2017 (Table). The possibility of sensitivity or specificity differing among the case definitions is, therefore, a limitation of this project.
We developed and animated a horizontal bar graph and a choropleth map in R version 3.6.0 (R Foundation for Statistical Computing) (7), by using the albersusa (8), ggplot2 (9), viridis (10), ggthemes (11), gganimate (12), and magick (13) packages. Graph and map animations were rendered separately as graphics interchange format (GIF) images, and then combined. The open-source FFmpeg software suite was used to convert the animated GIF image into an MP4 formatted video (14).
The animation has a short duration, which facilitates consecutive viewings. The animation also leverages interactive features of MP4 video players, such as play, pause, vary playback speed, advance frame-by-frame, rewind, fast forward, and jump to specific places. These constitute a partial menu of possible interactive features that a data visualization might incorporate. We acknowledge that, in this sense, our animated data visualization has limitations. The combined animation’s interactivity, open layout, and high placement of the title; however, are consistent with common design practices for animated maps online that generally conform to cartographic standards (15).
Our animated data visualization presents cancer mortality rates spatially and temporally and illustrates that despite the overall decrease nationally in the age-adjusted rate from 1968 through 2017, disparities persisted among states and counties. This visualization can be used to improve public health resource targeting and evidence-based intervention efforts for states and counties with emerging or persistently high cancer mortality rates. Health officials, policy makers, and stakeholders can use data animation to inform policies and practices that influence cancer outcomes. For example, animation can focus attention on counties throughout the Mississippi Delta and Appalachia, where declines in cancer mortality have lagged compared with national declines — a known pattern that would be difficult to discern from a static data visualization.
Animated choropleth mapping is a novel visualization method for health data. Our project is the first, of which we are aware, to combine an animated graph and animated choropleth map. Although specific data, such as maximum and minimum values, might be difficult to convey by animated visualizations, animation can be effective for adding another dimension of information, particularly time, to static data visualizations. In doing so, animation can enable some data visualizations to convey patterns and relationships that are not apparent from static visualizations, especially across geographic levels. Therefore, we encourage data analysts to consider synchronized graph and choropleth map animations as an option for communicating data to public health researchers, practitioners, and policy makers.
No financial support was received for this work. No copyrighted surveys, instruments, or tools were used. The authors have no conflicts of interest.
Corresponding Author: Isaac H. Michaels, MPH, 3100 Rosendale Road, Niskayuna, NY 12309. Telephone: 347-687-9719. Email: [email protected].
Author Affiliations: 1Department of Epidemiology and Biostatistics, University at Albany School of Public Health, Rensselaer, New York. 2Health Resources and Services Administration Region 2 Public Health Training Center, Columbia University Mailman School of Public Health, Department of Sociomedical Sciences, New York, New York.
- Murphy SL, Xu JQ, Kochanek KD, Arias E. Mortality in the United States, 2017. https://www.cdc.gov/nchs/products/databriefs/db328.htm. Accessed June 16, 2019
- Islami F, Goding Sauer A, Miller KD, Siegel RL, Fedewa SA, Jacobs EJ, et al. . Proportion and number of cancer cases and deaths attributable to potentially modifiable risk factors in the United States. CA Cancer J Clin . 2018;68(1):31–54.
- Centers for Disease Control and Prevention. Vital signs: state-specific obesity prevalence among adults — United States, 2009. MMWR Morb Mortal Wkly Rep 2010;59(30):951–5.
- FiveThirtyEight. 35 Years of American death: mortality rates for leading causes of death in every US county from 1980 to 2014. https://projects.fivethirtyeight.com/mortality-rates-united-states/cancer2/#2014. Accessed July 15, 2019.
- Centers for Disease Control and Prevention. CDC WONDER, 2019. http://wonder.cdc.gov. Accessed May 25, 2019.
- National Cancer Institute. SEER cause of death recode 1969+ (04/16/2012). https://seer.cancer.gov/codrecode/1969_d04162012/index.html. Accessed July 8, 2019.
- R Core Team 2019. The R Project for Statistical Computing. https://www.R-project.org/. Accessed July 8, 2019.
- Rudis B. Tools, shapefiles and data to work with an “Albersusa” composite projection in R. R package version 0.3.1. https://github.com/hrbrmstr/albersusa. Accessed January 21, 2020.
- Wickham H. Elegant graphics for data analysis. New York (NY): Springer-Verlag Inc; 2016.
- Garnier S. Default color maps from ‘matplotlib.’ R package version 0.5.1. https://CRAN.R-project.org/package=viridis. Accessed January 21, 2020.
- Arnold JB. Extra themes, scales and Geoms for ‘ggplot2.’ R package version 4.2.0. https://CRAN.R-project.org/package=ggthemes. Accessed January 21, 2020.
- Pedersen TL, Robinson D. A grammar of animated graphics. R package version 1.0.5. https://CRAN.R-project.org/package=gganimate. Accessed January 21, 2020.
- Ooms J. Advanced graphics and image-processing in R. R package version 2.0. https://CRAN.R-project.org/package=magick. Accessed January 21, 2020.
- FFmpeg. ffmpeg tool version 4.1.4. http://ffmpeg.org/. Accessed January 21, 2020.
- Cybulski P. Design rules and practices for animated maps online. J Spat Sci 2016;61(2):461–71. CrossRefexternal icon
|Year||Deaths||Age-Adjusted Rate per 100,000 Population|
Source of original article: Centers for Disease Control and Prevention (CDC) / Preventing Chronic Disease (tools.cdc.gov).
The content of this article does not necessarily reflect the views or opinion of Global Diaspora News (www.GlobalDiasporaNews.com).
Sign up to Global Diaspora News newsletter (https://www.GlobalDiasporaNews.com/newsletter/) to start receiving updates and opportunities directly in your email inbox for free.