How COVID-19 Took Hold in the United States, February–April 2020

This report was posted on May 1, 2020, early in the COVID-19 pandemic, and describes what was known then. Much has been learned since.

The first confirmed case of COVID-19 in the United States was reported on January 21, 2020. From then through February 23, 14 cases were diagnosed in six states: 12 in travellers arriving from China and two in their household contacts. Another 39 were found among people brought home from affected areas abroad. The outbreak looked contained.

Then, from late February, cases appeared in people who had not travelled to affected areas and had no link to a known case — the sign that the virus was spreading on its own. By mid-March transmission was widespread, and by April 21, 793,669 confirmed cases had been reported in the United States, most from community spread. Over a 3-week period from late February to early March, the number of U.S. cases rose more than 1,000-fold.

Confirmed U.S. cases by date of report, February 20–April 21, 2020, with the early rise in three states after large gatherings. Credit: Centers for Disease Control and Prevention.

CDC's review, by Anne Schuchat and the CDC COVID-19 Response Team, names four things that drove the March acceleration.

1. The virus kept arriving

After China halted travel from Wuhan and other cities in Hubei Province on January 23, and the United States restricted non-U.S. travellers from China from February 2, air passenger journeys from China fell 86%, from 505,560 in January to 70,072 in February. But in February 139,305 travellers arrived from Italy and 1.74 million from the Schengen countries, where the outbreak was spreading fast; those numbers fell only in March.

Genetic sequencing told the same story. Washington State's outbreak strains suggested an introduction from China around February 1, but northern California's showed many introductions — from China, from Europe and from other states — and New York's March strains pointed to Europe and other parts of the country.

Cruise ships added more. Passengers from many countries mix closely on board, and crews stay on for voyage after voyage. In one example, 101 people from nine separate Nile River cruises between February 11 and March 5 returned to 18 states and tested positive, nearly doubling the number of known U.S. cases at the time.

Cases linked to nine Nile River cruises, February 11–March 5, 2020, by state of residence. Credit: Centers for Disease Control and Prevention.

The response: entry restrictions on non-citizens coming from China, later widened to other countries with sustained spread; travel health notices, and in the end advice to avoid all nonessential international travel and all cruise travel; quarantine for travellers exposed in places such as Hubei Province and the Diamond Princess cruise ship off Yokohama, Japan; and airport screening, begun on January 17 and expanded. By April 21 CDC staff and U.S. Customs and Border Protection officers had screened approximately 268,000 returning travellers, and testing confirmed 14 cases among them. On March 14 the CDC Director issued a No Sail Order suspending cruise ship operations in U.S. waters, renewed on April 9.

2. People gathered, then went home

In February, with few cases and those apparently contained, no one recommended limits on gatherings. In the last week of the month, three events seeded outbreaks:

• Mardi Gras in Louisiana, with more than 1 million attendees;
• an international professional conference in Boston with approximately 175 attendees;
• a funeral in Albany, Georgia, with more than 100 attendees.

Dougherty County, Georgia, a small rural county that includes Albany, went on to have one of the highest cumulative incidences in the country: 1,630 per 100,000 population. From mid-March, recommendations to limit mass gatherings and travel followed.

3. The virus reached the places where it spreads best

Nursing homes, long-term care facilities and hospitals bring people at high risk of severe illness into close contact with staff, many of whom work in more than one facility. Other workplaces amplified spread too, among them several meat packing plants in rural areas. Clusters were tied to religious services, in the United States and elsewhere, and crowded, dense settings may help explain the very high case numbers in the New York metropolitan area.

Responses in these settings centred on infection control: finding and isolating sick people, cleaning and disinfection, limiting visitors, spacing workers out through shifts, and proper use of protective equipment. To protect health care capacity, authorities at every level issued stay-at-home orders and closed schools and nonessential workplaces. On April 3 CDC advised cloth face coverings in public, given growing evidence of spread from people without symptoms.

4. Much of the spread went unseen

The virus arrived during flu season. Surveillance systems tracking respiratory illness in clinics and emergency departments, including in Seattle where early clusters appeared, saw nothing unusual in the early acceleration because seasonal influenza dominated.

After Santa Clara County, California, confirmed its first community case on February 27, it tested patients at four urgent care centres from March 5 to 14. Influenza accounted for 23% of respiratory illnesses; among those who tested negative for flu, 11% tested positive for SARS-CoV-2 — about 8% of all patients with respiratory symptoms.

Blood samples told a similar story. In Seattle in March, 1 of 221 leftover clinical samples from healthy adults without respiratory illness showed antibodies — few, but across a whole population still a large number of unrecognised infections. None of 59 children with respiratory infections from January to March had antibodies. With infections still relatively rare, only very widespread testing could have found every case, and spread from people without symptoms made detection and containment harder still. Responses included more surveillance and testing, teleworking, stay-at-home orders, school closures, social distancing and cloth face coverings.

What it meant

These factors acted at the same time, so the report cannot say how much each contributed, or how much each response helped. Its examples are illustrative, not complete; limited testing shaped the epidemic curve, especially early on; and the case counts understate the true number.

Some measures that mattered most at the start, such as quarantine and airport screening, may count for less once the virus is spreading widely in communities, but much of the rest would still be needed. Early antibody surveys suggested that even where recognised cases were most numerous, most people had not been infected and remained susceptible. Sustained effort, the report concludes, would be needed to prevent further spread.

Sources

Based on "Public Health Response to the Initiation and Spread of Pandemic COVID-19 in the United States, February 24–April 21, 2020," by Anne Schuchat and the CDC COVID-19 Response Team, Morbidity and Mortality Weekly Report 69(18), posted as an MMWR Early Release on May 1, 2020, published by the Centers for Disease Control and Prevention; rewritten in hubnx's own words.

• The Seattle antibody findings were personal communications to the authors from researchers at the University of Washington and Seattle Children's Hospital.
• The report's table of mitigation measures and its box on New York are not reproduced.