The parts of the United States hit hardest by COVID-19 shifted over time. On May 7, 2020, CDC and other federal agencies began flagging hotspots — counties where cases were rising — to understand how the virus was spreading and to offer targeted help to local health departments. CDC then looked back at the data from January 22 (the first U.S. case) and forward to July 15.
Visual summary. CDC
What made a county a hotspot
A county had to meet all four of these, on the day assessed:
- more than 100 new cases in the last 7 days;
- a higher 7-day incidence than in the 7 days before;
- a 3-day incidence that had risen, or fallen by less than 60%, from the 3 days before;
- a 7-day incidence more than 0.31 times the 30-day incidence;
and at least one of these: a change of more than 60% in the latest 3-day incidence, or in the latest 7-day incidence.
Case counts came from USAFacts, which gathered them from state and local health department websites, cleaned by CDC and the Johns Hopkins University Applied Physics Laboratory. Each day a county met the criteria counted as one alert.
The numbers
No county qualified from January 22 to March 7. From March 8 to July 15:
| Counties that were hotspots on at least one day | 818 of 3,142 (26%) |
| Share of the U.S. population living in them | 80% |
| Alerts | 9,898 |
| Days a hotspot county met the criteria (median; not necessarily in a row) | 10 (interquartile range 5–18) |
The daily count of hotspot counties:
- peaked at 175 in early April;
- fell and held below 75 a day from mid-April to early June;
- climbed again to 179 in early July.
Hotspot alerts by county and number of days meeting the criteria, March 8–July 15, 2020. CDC
Where: by region
| Census region | March–April | May | June–July |
|---|---|---|---|
| Northeast | 40% of counties (84% of the region's people) | 11% | 9% (16%) |
| South | 10% | 9% | 28% (76%) |
| Midwest | 8% | 8% | 10% (44%) |
| West | 13% | 9% | 22% (86%) |
Share of the region's counties that were hotspots on at least one day in each period (and the share of its population they held).
Where: cities to small towns
| County type | March–April | May | June–July |
|---|---|---|---|
| Large central metro | 97% | 46% | 78% |
| Large fringe metro | 31% | 16% | 39% |
| Medium metro | 26% | 19% | 46% |
| Small metro | 13% | 13% | 32% |
| Micropolitan | 5% | 6% | 16% |
| Noncore (outside metro and micropolitan areas) | 12 of 1,335 counties | 26 | 37 |
By summer, hotspots had spread from the biggest cities into medium and small metro areas and micropolitan counties.
Daily hotspot alerts by county type, and the 7-day average of new reported cases, March 8–July 15, 2020. CDC
What drove the rises
For counties newly flagged from May 11 to July 13, CDC and the Applied Physics Laboratory looked for causes, using HHS Protect data, health department websites, news reports, CDC deployment records and calls to state health leaders. A county could have more than one.
| Share of identified causes | May 11–31 | June 1–July 13 |
|---|---|---|
| Focal outbreaks — long-term care, food processing, prisons and jails, other workplaces | 56% | 24% |
| Community transmission | 18% | 41% |
| No discernible cause | 8% | 24% |
| More testing or reporting delays | 17% | 11% |
Causes were found for 116 of 124 counties newly flagged in late May (214 causes in all), and for 389 of 539 in June and early July (481 causes).
The response
From May 7 to July 15, CDC sent 92 teams of 375 people to 37 states and the District of Columbia, mostly for hotspots. In North Carolina, for example, CDC (from June 13) and the U.S. Public Health Service (from June 22) sent teams to help with case investigation, contact tracing and data management, and the CDC Foundation supplied extra contact tracers for local health departments.
Once a county is flagged, data on its epidemic, its people, prevention, testing and health care use guide outreach to local officials — to check the findings, hear local concerns and find the right help. That can include federal technical support, on a state's request, for epidemiology, contact tracing, testing, community mitigation, worker safety, infection control, communications and health care.
What it means
- Community spread in June and July showed how fast the virus can move even without outbreaks in high-risk group settings.
- Its spread across metro and micropolitan counties made community mitigation — masks, physical distancing, hand hygiene — more urgent.
- Responses should fit local populations, cultures, languages and politics, ideally with local community leaders involved.
- Work was under way to spot the most worrying hotspots, and communities at risk of becoming hotspots, sooner.
Limits
- Differences in testing, reporting delays and changes in reporting may have affected how well case counts tracked real infections.
- The 100-case threshold makes small counties less likely to be flagged; their trends, and their neighbors', are still watched.
- Causes came from existing sources and may be incomplete, and more so in some communities than others.
Sources
Based on Oster AM, Kang GJ, Cha AE, et al., "Trends in Number and Distribution of COVID-19 Hotspot Counties — United States, March 8–July 15, 2020," MMWR volume 69, number 33, Centers for Disease Control and Prevention; a work of the United States government in the public domain. For noncore counties the report's table gives a count of 12 as 0% while its text gives a range of 1% to 3%, so counts are shown for them here.
Licence: CC0 1.0 (public domain) · Adapted from www.cdc.gov
1
0
0
0
Comments
