Coronavirus Is Spreading Quickly Through Our Most Vulnerable Communities
Case data indicates that the growth in COVID-19 cases is disproportionately fast in already vulnerable counties that will be hardest hit by a major outbreak.
Last week we published the COVID-19 Community Vulnerability Index (CCVI) for each state, county, and census tract in the United States, and we analyzed it in-depth in yesterday’s post. The index captures the expected social, health, and economic impact of the virus once it arrives in a community.
Today, we’re examining how this vulnerability relates to the numbers of confirmed cases of COVID-19 in a community. Our analysis uses case data at the county level made available by Johns Hopkins University on April 13th.
We find that, early in the outbreak, low vulnerability counties were more likely to be hit by the virus, but this trend was short-lived. This month, COVID-19 cases in high vulnerability counties have been rapidly catching up to — and even overtaking — those in lower vulnerability populations. Although the virus reached highly vulnerable communities later, its spread within those communities is much faster.
This is troubling news: it’s precisely these more vulnerable counties that are least prepared to manage a rapid surge in cases.
Here’s what we’re learning about the spread of COVID-19 in vulnerable communities:
1. There are now COVID-19 hotspots in several highly vulnerable areas.
As of April 13th, there were 547,929 cases of COVID-19 recorded across the US — a number that’s rising by about 30,000 per day. But cases are distributed very unequally: the worst-affected 20% of counties account for 96% of all cases. New York City alone has 19% of all cases in the country. By contrast, 15% of counties in the US still have 0 confirmed cases.
The map below overlays cases of COVID-19 per 100,000 people on the vulnerability of individual counties. Darker colors indicate counties with greater vulnerability, and the red circles show the number of cases per 100,000 people — the bigger the circle, the larger the number.
The relatively small circles in the Northwest show that, while Washington State was the first to be hit by COVID-19, in terms of cases per 100,000 people it does not stand out. Rather, current hotspots are in New York, Louisiana, and Georgia. These areas in the South are highly vulnerable, as are parts of New York (such as Brooklyn and the Bronx).
2. In April, cases in highly vulnerable counties soared.
For our remaining analyses, we divided counties into three categories of vulnerability: low, medium, and high, each containing one-third of counties in the US.
The graph below shows how many counties in each category reported their first-ever case of COVID-19 for each day between March 1st and April 11th (a few counties affected before March 1st are not shown). We see that between March 1st and March 18th, low vulnerability counties started to get hit by COVID-19.
After a lag of several days, COVID-19 reached similar numbers of medium and high vulnerability counties, and these have since been catching up.
Low vulnerability counties were, on average, hit several days before medium and high vulnerability counties (the black line marks the date when half of counties with cases were hit).
The next graph captures a startling finding. If we ignore counties and simply consider the entire population of each vulnerability category as a whole, we see that until April 1st, populations of low-vulnerability areas consistently had more cases of COVID-19 per 100,000 people.
However, in early April cases in the high vulnerability population started to accelerate, catching up and overtaking low vulnerability areas by April 5th — the space of less than a week!
This graph is particularly worrying because it hints that the spread of cases within a county may be faster for more-vulnerable counties.
Let’s examine this in more detail.
3. COVID-19 is spreading faster in vulnerable counties.
To assess the speed at which the virus spreads, we use doubling time, an indicator also used by The New York Times. We calculate it by looking at the date when the county had only half as many cases as it had on April 13th, and counting the number of days between the two dates. We have this value for 33% of counties in the US, as we only calculate doubling time for counties with at least 20 cases. Nonetheless, doubling time is a good metric to gauge the current speed of spread, rather than the speed at some past point in the outbreak (though we calculated that as well by taking days between case 1 and case 10 per 100,000 population, and the conclusions are the same).
We find that, in the average county, cases are currently doubling every 8 days. But the figure varies by vulnerability: low vulnerability counties take 8.6 days to double their number of cases, medium vulnerability counties take 7.8 days, and high vulnerability counties just 7.1 days.
The difference between those numbers may not seem great, but that’s because we tend to underestimate the implications of exponential growth.
As the graph below shows, if the trend were to hold, after one month low vulnerability counties would have 12 times as many cases as they’d had at the beginning of the month, while high vulnerability counties would have 21 times as many.
This confirms what we feared: more-vulnerable counties see faster disease spread once the virus hits, and therefore have a weakened ability to deal with cases by an already less-resilient health system.
What’s next?
“Flattening the curve” is most critical in highly vulnerable communities, which have the least resilience in the face of a major outbreak of COVID-19, but these communities are struggling to slow the spread of the virus. Vulnerable communities need support now, with strategies to bring down transmission, increase healthcare capacity, and bring in extensive testing to effectively combat further spread.
But vulnerability is about more than just the immediate impact of the disease: financial resilience, job security, poor outcomes for children, unattended chronic disease, and many more hardships will afflict millions living in communities ill-equipped to deal with the shock this pandemic will bring.
Understanding why these communities are seeing faster spread of COVID-19 will be crucial in order to effectively intervene, not only to stop the disease in the short term, but to protect the most vulnerable communities for years to come.
This work was made possible by everyone at the Surgo Foundation, including but not limited to (in alphabetical order) Yael Caplan, Vincent Huang, Hannah Kemp, Tich Mangono, Sema Sgaier, Peter Smittenaar, Nick Stewart, and Staci Sutermaster.
Technical notes
- Case data was retrieved on April 13th from JHU Github repository.
- Cases are confirmed cases (including those that died, recovered, or are active; but excluding those never confirmed in a test).
- JHU combines New York City’s five boroughs — Bronx, Queens, Kings (Brooklyn), Richmond (Staten Island), and New York (Manhattan) — into a single “New York County’’ when listing the number of confirmed cases. Manhattan has the lowest vulnerability among the city’s five boroughs, while some of the other boroughs have much higher vulnerability. We have included New York County throughout this article, except in one graph, where this is noted in the legend. We are working on integrating borough-specific case numbers from USAfacts.org.
- Want to analyze the numbers for yourself? The CCVI data is publicly available at census, county, and state level (link). The JHU data is available from Github.
- Questions, concerns, comments? Get in touch at covid19@surgofoundation.org
