COVID-19 Infections in U.S. Are About 10 Times Greater Than Reported Cases, Says Study

About 40 percent of infected persons show no symptoms but may still transmit disease.


Researchers with the Centers for Disease Control and Prevention (CDC) have just published the results of blood tests seeking antibodies for COVID-19. This seroprevalence study in the Journal of the American Medical Association (JAMA) scanned blood samples taken from more than 16,000 people from 10 areas around the U.S. during varying time periods in April and May. The samples are not random but were gathered from patients who were visiting their physicians or hospitals for treatments and checkups.

Based on these test results, the researchers conclude that at most of the sampled areas "it is likely that greater than 10 times more [coronavirus] infections occurred than the number of reported COVID-19 cases." In fact, they estimated that the ratio of undetected infections to reported cases ranged from being six times greater in Connecticut, to 24 times greater in Missouri.

The researchers caution that "extrapolating these estimates to make assumptions about population immunity should not be done until more is known" about how much long-term protection against re-infection antibodies may provide. They further note that their findings "may reflect persons who had mild or no illness or who did not seek medical care or undergo testing but who still may have contributed to ongoing virus transmission in the population." The CDC recently estimated that about 40 percent of people infected with the virus that is responsible for COVID-19 exhibit no symptoms and the chance of transmission from such people is around 75 percent.

The researchers estimate in the JAMA study that the percentage of people who had been infected by the time the samples were collected in April and May range from 6.9 in the New York City area to 1 percent in the San Francisco Bay area. These results are from early in the pandemic and estimates based on later antibody testing suggest, for example, that more than 20 percent of New York City residents had already been infected by the end of April.

What are the implications of these findings for the future course of the pandemic in the U.S.? "The study rebukes the idea that current population-wide levels of acquired immunity (so-called herd immunity) will pose any substantial impediment to the continued propagation of [COVID-19 coronavirus] in the U.S., at least for now," states an accompanying JAMA op-ed. "The size of the epidemic projected through early May 2020 in this study falls far short of the estimated herd immunity threshold of approximately 60% to 70%." As the op-ed further observes, "Seven of the 10 study locations are currently experiencing substantial, as-yet uncontrolled increases in new COVID-19 cases." On the other hand, other very preliminary evidence of more widespread T-cell mediated immunity to the COVID-19 coronavirus suggests the herd immunity threshold could be closer than antibody tests by themselves may reveal.

Parsing the data presented in the JAMA study allows for the very crude calculation of the infection fatality rate (IFR) for COVID-19, that is, the percent of infected people who will die of the illness. Dividing the number of infections estimated in the JAMA study by the total number of deaths in the various regions at the time the samples were collected, yields an overall crude IFR of around 0.47 percent. This compares favorably with the median IFR estimate of 0.5 percent derived from the well-respected machine learning model developed by independent data scientist Youyang Gu. Bearing in mind that these samples were obtained during the early part of the pandemic, the IFR for the various regions in the study range from a low of 0.1 percent for Utah and a high of 1.4 percent for Connecticut, both in early May. Earlier this month the CDC upped its overall COVID-19 IFR to 0.65 percent.

The IFR for seasonal flu is conventionally estimated at around 0.1 percent—although some researchers think that's too high—which suggests that COVID-19 is nearly 5 times more deadly than the flu. So far, more than 80 percent of U.S. COVID-19 deaths occurred in people over the age of 65 years. This death rate is similar to CDC estimates that between about 70 percent and 85 percent of seasonal flu-related deaths occur in people 65 years and older.

The JAMA authors conclude: "Because persons often do not know if they are infected with [the COVID-19 coronavirus], the public should continue to take steps to help prevent the spread of COVID-19, such as wearing cloth face coverings when outside the home, remaining 6 feet apart from other people, washing hands frequently, and staying home when sick." Until effective vaccines become available, that's good advice.