Monday, March 2, 2020

Mapping An Outbreak





Hello Everyone:

It is a lovely and windy Monday.  Before we get started on today's subject, mapping epidemics, a quick reminder that Super Tuesday, Game Day, is tomorrow, March 3, 2020.  If you are in one of the fifteen states, including California, that goes to the polls tomorrow and information, please go to usa.govsos.ca.gov for California voters, or your state's secretary of state website.  Finally, as of writing we says thank you and farewell to three nominee candidates: Tom Steyer, Peter Mayor Buttigieg, and Minnesota Senator Amy Klobuchar (D).  For now, let us take about mapping epidemics.

Image result for mapping epidemics
Mapping a London epidemic
nationalgeographic.org

The coronavirus continues to spread like wildfire, around the world.  Having a map to chart the path of the deadly virus is useful to understand how it spreads, where it goes, and most important where to target public health interventions.

Marie Patino of CityLab reports, "One of the best of examples is the dashboard built by Johns Hopkins University's Center for Systems Science and Engineering..." (citylab.com; Feb. 11, 2020: date access Feb. 25, 2020).  Using mapping software from Esri, JHU aggregated data from global health officials in real time.  Take a look at
Image result for mapping epidemics: jhu online dashboard
sciencealert.com
the map on the lower left-hand side and what see that the bigger the red bubble gets, the higher number of reported cases in a given geographical areas.  As of fourteen minutes ago, the worldwide death toll from coronavirus is over 2,700, with the vast majority in mainland China (cnn.com; Feb. 25, 2020).  To put it in proper context, the coronavirus is now more deadly than 2003 SARS epidemic.

The pandemic threat management company, Metabiota, listed the coronavirus to its list over 130 pathogens is it currently tracking around the the world.  Like Johns Hopkins' epidemic tracking map, Metabiota also created a map, albeit with a more pleasing color palette except for the orange dots indicating the presence of the virus, below left.  Ms. Patino asks, "But is this really new?"
Image result for mapping epidemics: metabiota: twitter
Metabiota (@metabiota)
twitter.com
 The disease is new but mapping epidemics goes back centuries.  One of the earliest attempts was in the 17th century.  Recently, medical cartography has taken off and become more democratic thanks to technological advances made possible in the digital age.  In addition to more powerful devices, the internet makes data gathering and sharing easily and quickly accessible.

Image result for mapping epidemics: historic
philadelphiaencyclopedia.org


Ms. Patino writes, "For centuries, geographers and health-care officials have used so-called geographic information systems (GIS) mapping to craft theories on why a particular outbreak occurred based on the clustering of casualties.  Now they can also project how an ongoing epidemic might evolve--in real time--and set policy based on those models" (citylab.com; Feb. 11, 2020)  Este Geraghty, the chief medical officer and health solutions director at Esri, told CityLab,

The underlying goal of using a GIS in the past was to develop a greater understanding of what happened, and of some of the root causes,... But today, GIS has evolved, so we can do so much more (Ibid)


Image result for cholera map: Arrieta
Plague map 1690-92
Attributed to Fillippo Arrieta
Bari, Italy
researchgate.net
 "The first occurrence of disease mapping can be traced back to 1692, according to Tom Koch, professor of geography at the University of British Columbia and author of the book Cartographies of Diseases" (Ibid).  This was during a period when the plague was once again rampaging through Europe, and Fillippo Arrieta, an Italian royal auditor, spatially visualize the strateg, for containing the spread of this flea-borne disease in the region of Bari, Italy (Ibid).  On his map, Bari was separated from the rest of the country by a dashed line signifying a "cordon santaire" (Ibid).  Within the cordoned off area are two smaller areas, further separated by thick lines.  If you look close enough, you can see a large "D" in the top right, indicting an infected province.

Image result for yellow fever: valentine seaman
Yellow Fever map 1797 (1804)
Attributed to Valentine Seaman
New York, New York
brianaltonenmph.com

According to Prof. Koch's research, the first truly detailed spatial imagining of an epidemic did not appear until about 1797 and the publication of Valentine Seaman's yellow fever outbreak maps in the Medical Repository (Ibid).  Valentine Seaman overlaid the locations of yellow fever outbreak, signified by dots, with the locations of dump and sewage sites in lower Manhattan.  Those locations were indicated by a thick letter "S."  After analyzing the data, Valentine Seaman concluded that outbreaks of deadly disease with linked to those sites and the putrid odors (Ibid). "Even if Seaman's theory was not quite right--yellow fever is carried by the mosquitoes that were breeding in these wastes--disease mapping was born" (Ibid).  Prof. Koch added "Over time, technology improved and disease related data become more available,..."(Ibid).  Be that as it may, it was the great cholera epidemic that struck Europe and, in particular, the United Kingdom, that saw a boom in disease maps.  In an email, Prof. Koch observed, And then it was...watershed (Ibid).


Image result for John Snow: cholera
Dr. John Snow and a detail from his cholera outbreak map
London, England
hauntedwalk.com
  Dr. John Snow's (not Game of Thrones) famous maps of the London cholera epidemic of 1854 remains the best example of the explosion in spatial visualization of this disease, as Britons struggled to understand its causes. Dr, Snow was not the only one mapping the outbreak.






Image result for Richard Grainger: cholera
London cholera outbreak 1850 (detail)
Attributed to Dr. Richard Grainger
researchgate.net
Dr, Richard Grainger hypothesized a connection between the disease and altitude.  He very precisely mapped the city, drawing all of London's districts and sub-districts, including the location of sewers and wells.  On to this, he overlaid elevation information and shaded in the areas according to the intensity level of the outbreak.  The darker the blue area, the more intense the outbreak.  Notice the deep blue areas near The Thames south bank.  Prof. Koch wrote, "...the masses near the banks of the river Thames in South London struggled with less clean air than Londoners who lived higher up" (Ibid).


Image result for Esri disease outbreak maps
Flu Index in the United States
esri.com

In the digital age, the computer has provided the technology to generate maps at light speed and the internet enables data-sharing and transmission faster than ever.  Computers have become extremely efficient tools for their data processing capabilities, resulting in the creation of geospatial models that allow health care officials to follow where an ongoing epidemic, like the coronavirus, and identify high-risk populations.  The benefit of geospatial models is that it makes public health interventions possible.  Este Geraghty told CityLab, "...when she joined Esri, the public health community was very familiar with GIS."  Ms. Geraghty,

They had an understanding and they had been mapping, but they had been doing it with desktop tools, not web-based GIS" (citylab.com; Feb. 11, 2020).

The result is larger datasets more readily available, and GIS users can create their own prediction models based on the datasets.  Among the wealth of information available readily accessible is census data, shipping or air routes, and social media content.  On that note, Twitter has been extremely useful for people in the hot zones to share their experiences and public health information.

Image result for dashboard for Zika virus in the U.S.: Esri 2016 CityLab
Esri dashboard map of Zika
Originally seen at healthtechmagazing.net
No longer available
citylab.com
Marie Patino reports "In 2016, the U.S. Centers for Disease Control and Prevention used Esri's products and expertise to monitor the diffusion the Zika virus" (Ibid).  According to Ms. Geraghty, the mosquito-borne virus, the survival and reproduction rate of the insect is tightly linked to five variables: temperature, rain fall, land use, population, and elevation (Ibid).  After analyzing the data, researchers were able to cross-reference the results with census data.  Since pregnant women are particularly vulnerable to the Zika virus, the census data overlay permitted researchers to identify the hot zones with the largest at-risk population.  This led to more effective policy-making and testing, while locals were encouraged to use insecticide and larvicide as part of an arsenal to limit the spread of the disease.

The dashboard map (above) followed the number of Zika cases in the United States was made available to the public.  The states filled in darker shades of red indicate the number of cases.

According to pandemic threat management company Metabiota chief executive officer Nita Madhav, the company "has accumulated and cleaned data 2,400 outbreaks since its creation 2008" (citylab.com; Feb. 11, 2020).  Metabiota's epidemic tracker is publicly available but most of its modeling capacities are only available to its clients, including the CDC and the United States Agency for International Development.  Ms. Madhav told CityLab,

This [data on previous epidemics] can actually help inform future decision-making,... and can show use that these epidemics should not come as a surprise.  This is something that happens frequently over time (citylab.com; Feb. 11, 2020).

Image result for tracking the coronavirus
nytimes.com
Metabiota evaluates when and where outbreaks are more likely to happen; Ms. Madhav added, "is currently 'nearcasting'--projecting the short-term evolution--of the coronavirus" (citylab.com; Feb. 11, 2020).  The company also created a way to measure the fear (Ibid) metric, labeled pathogen sentiment score (Ibid).  Ms. Madhav explains, "To calculate that score,... researchers combine existing data regarding an outbreak--like mortality rate, for example--an put it into a 'scoring algorithm" (Ibid).  The publicly available scores are ranked on a broad scale: the coronavirus ' sentiment score is defined as high (Ibid).  More granular results are available but not posted on the company's website.

The Johns Hopkins dashboard it built upon multiple data feeds.  It uses CDC, WHO, ECDC and DXY report--Chinese, American, global data sources.  The resulting map is the only reliable data sources, but tracking an epidemic can be tricky: some cases go unreported.

Este Geraghty, at Esri, reflects on where the field of geospatial mapping might go.

GIS is moving in a direction that is maybe more democratized...There is always a need for these GIS professionals, but a lot of people whose job is totally different and need maps to make decisions don't have to have the expertise.  They just need to know enough about mapping (citylab.com; Feb. 11, 2020) 
   

No comments:

Post a Comment