It is often said that a picture is worth a thousand words. I would argue that a map is worth much more.
Can maps provide valuable information about the coronavirus pandemic? Absolutely.
I’ve had a fascination with maps since I was a young boy and used to buy topographic maps with my allowance. I enjoyed geography in school and when I studied political science and political economy in university, I took courses in human geography and statistics. Later, in one of my jobs, I used geomatics to analyze public opinion data.
Geomatics is defined as a “discipline concerned with the collection, distribution, storage, analysis, processing, presentation of geographic data or geographic information.” It’s related to GIS (geographic information system), which pertains more to the software used for this analysis.
Geomatics can be used to look at spatial information about where the SARS-CoV-2 virus is spreading. Combined with demographic information about race, poverty levels and access to healthcare, it can provide insights on the types of places and populations most at risk for Covid-19.
British Columbia so far can be considered a success story in Canada for its handling of the pandemic. But one area where the province has failed badly is its lack of transparency about geographic spread of the virus.
Numbers of cases are reported by health region only. And in the case of those of us under the Interior Health Authority, this represents a vast area of more than 215,000 square kilometres covering much of the southeast quarter of the province. It stretches from Manning Park to beyond Fernie, Williams Lake and Golden.
Are cases under IHA clustered in urban centres like Kelowna or Kamloops? Are they found along transportation corridors like Hwy 97? Are they in northern areas where poverty plays a role? We don’t know, because in B.C. that information is secret. Knowing there have been 175 cases (as of the morning of May 4) in IHA tells us very little.
Fortunately, other jurisdictions are much more transparent. In the U.S., information is reported at the county level. Johns Hopkins University provides excellent geomatic information with an interactive map showing the spread of the disease in every American county.
First a caution about Covid-19 and statistics. They must be taken with many grains of salt. Estimates of case numbers may be more reflective of the amount of testing than actual cases. They probably severely underestimate the numbers of cases because many people who are positive show no symptoms or only mild ones and may recover at home without ever being tested or reported.
Death statistics may be more reliable because you don’t need a test to show that someone is dead. But they are a lagging indicator because people die weeks after being exposed to the virus. Probably too, many people dying of Covid-19 are recorded instead as dying of such causes as stroke or pneumonia, which may have been brought on by the virus.
At best, the numbers are ballpark estimates that can highlight trends.
On the Johns Hopkins map, one of the useful ways to identify epidemic hotspots is to look at confirmed cases by population. This shows the number of cases per 100,000 population. The advantage of this approach is it shows areas that are hardest hit in terms of their population numbers (and medical resources). Simply looking at case totals would bias the analysis toward the most heavily populated counties.
By searching for dark purple counties on the map and then looking at confirmed cases by population, it is easy to find the hardest-hit areas – whether or not total case numbers are high. By clicking on a county, you can view demographic data such as race, poverty levels, hospital beds, population 65 and older, confirmed cases, fatality rates and more.
Some trends emerge suggesting the characteristics of areas that are most vulnerable. These appear to be major factors:
- Population density
- Level of poverty and, likely related, racial composition
- Presence of meat packing plants, or prisons
- Presence of long-term care facilities
- Tourism or transportation centres
Below, I’ve identified several hard-hit counties and areas in the U.S. that serve as examples. It’s by no means a complete list.
McKinley, New Mexico
The map of cases by population shows a big purple blob at the intersections of Arizona, New Mexico, Utah and Colorado. McKinley County, which surrounds Gallup, New Mexico, now has a rate of nearly 1,600 cases per 100,000 population. That’s higher than in hard-hit neighbouring counties in Arizona.
This area is home to the Navajo people and in relation to their population size, they’ve been devastated. Poverty, lack of water, poor communications and other factors have probably all played a role in the disease’s spread.
More than three quarters of the people in McKinley are American Indian and almost a third live in poverty. High numbers of those under age 65 have no health insurance.
These articles discuss the outbreak in the Navajo Nation generally as well as the situation in McKinley specifically:
Queens, New York
New York City is the epicentre of Covid-19 in the U.S. and currently (May 4) the county and borough of Queens has the highest number of confirmed cases (52,305), and deaths (4,035) in the entire country. Its number of confirmed cases per 100,000 population is now almost 2,300, which is among the highest in the region, although areas like Bronx and Richmond (Staten Island) are slightly higher.
This major wide outbreak extends from Philadelphia to Boston and beyond, encompassing New Jersey, which is growing as a hotspot.
Queens, with almost 2.3 million people, is a mixed-race borough, consisting of whites (36%), Asians (26%), blacks (19%) and “other races” (13%), which would include Latinos. With an 11.6% poverty rate, it’s not the poorest of New York’s boroughs, and comparatively fewer people lack medical insurance. Nonetheless, some of the hardest hit parts of Queens are those with high numbers of blacks and Latinos.
Most significant though are its population density and the reliance of its citizens on public transit in a city where relatively few drive. The area of Queens is just 280 square kilometres, meaning its density is 8,139 people per square kilometre – clearly a risk factor.
These articles discuss the situation in Queens and neighbouring boroughs:
- These Queens neighborhoods have been hit hardest by Covid-19 – QNS
- Brooklyn may now be deadliest county for Covid-19 in U.S., overtaking Queens – NBC
- ‘A tragedy is unfolding’: Inside New York’s virus epicenter – New York Times
Nobles County in Minnesota is a good example of how some of the most serious outbreaks can be overlooked if we only look at total cases or total deaths. Currently there are 940 confirmed cases and just one death.
But taking into account its small population – just under 22,000 – it’s being slammed. The number of confirmed cases per 100,000 people is a staggering 4,287.54 one of the highest rates in the U.S.
Nobles is not especially poor, with a poverty rate of less than 11% and with relatively few people lacking health insurance. But it’s home to the JBS pork plant in Worthington, MN. And it’s not alone among Midwestern U.S. rural communities with outbreaks at meatpacking plants where employees work in close quarters, often without adequate personal protective equipment (PPE).
These stories discuss the situation in Nobles and other communities with outbreaks at meat packing plants:
- Hundreds of workers test positive for Covid-19 at Worthington meatpacking plant, Minnesota officials say – Argus Leader
- JBS pork plant in Worthington set to partially re-open this week – Bring Me the News
- ‘We’re modern slaves’: How meat plant workers became the new front line in Covid-19 war – The Guardian
- CDC: Nearly 5,000 meat plant workers infected by Covid-19 – U.S. News & World Report
- South Dakota meat plant is now country’s biggest coronavirus hot spot – New York Times
Blaine County has fewer than 23,000 people living in a rural area of 6,892 square kilometres. But it’s a tourist centre with skiing, notably the Sun Valley ski resort near Ketchum.
In March, a large wedding was held there, immediately following a major skiing event. Wherever the virus came from, Blaine County became a major hotspot in terms of confirmed cases per 100,000 population – now 2,199.02.
Currently there are 497 confirmed cases and five deaths – certainly small compared to major metropolitan areas, but large for the size of the community.
The county has 3,804 people aged 65 and over and only 25 hospital beds, four of them ICU. This shows how vulnerable small communities can be when the virus is brought in by tourists from elsewhere.
These articles discuss the outbreak in Blaine:
- Why an Idaho ski destination has one of the highest Covid-19 infection rates in the nation – The New Yorker
- The people in Idaho’s coronavirus epicenter have a message for the rest of us – Buzzfeed
Georgia was one of the last states to introduce social distancing measures and it’s been one of the first and most aggressive in its recent reopening. Yet if you look north of Tallahassee, the Florida capital, you’ll see a large blob of dark purple on the Johns Hopkins map. This blob surrounds the small Georgia city of Albany.
Dougherty, the county where Albany is located, is more than 70% black. It has a poverty rate of nearly 30%. More than 13,000 of its 91,243 population are 65 or older.
Currently it has 1,536 confirmed cases and 124 deaths. Its confirmed cases per 100,000 population is 1,683.42 – high, but lower than some of the neighbouring counties like Terrell, with 2,160.03.
The outbreak is believed to have taken off as the result of a super-spreading situation at two funerals in early to mid-March. Soon hospital beds were overwhelmed.
These articles describe how Dougherty became one of the country’s hotspots on a per capita basis:
- Days after a funeral in a Georgia town, coronavirus ‘hit like a bomb’ – New York Times
- How a small Georgia city far from New York became one of the worst coronavirus hotspots in the country – Business Insider
The above five cases are certainly not the only hotspots in the U.S., but they illustrate the diverse ways this virus can strike communities, large and small.
One feature the Johns Hopkins data lacks is charts showing the growth curves over time in specific counties, although this information is available at national levels. Counties with the highest numbers of confirmed cases may have seen earlier outbreaks, giving the virus more time to spread.
This is only a brief survey of the data. By digging deeper, we could learn a lot more.
Hopefully sometime soon we’ll see this kind of information available for British Columbia and the rest of Canada.