A recent Stanford University study of COVID-19 infection in Santa Clara County suggests that the disease is vastly more widespread in the Bay Area than official data shows.
Because the test was performed on volunteers rather than a randomized population, it is likely to have over-reported the rate of infection among its subjects. Nonetheless, it suggests that our current understanding of the virus’ prevalence is wildly inaccurate.
The study, led by infectious disease specialist and Associate Professor of Medicine Eran Bendavid, aimed to estimate how many people in Santa Clara County have been infected with the novel coronavirus known as SARS-CoV-2. By measuring who among 3,300 test subjects presented SARS-CoV-2 antibodies, researchers determined that the number of infected people fell somewhere between 2.5 and 4.2 percent of the county’s 1.9 million residents as of April 1.
At the higher end of the study’s estimate, this would mean that nearly 80,000 county residents had already been infected with COVID-19 by the first day of April, at which time the county had only reported 956 confirmed cases. The results suggest that by April 1, the county had only confirmed 1.1 percent of actual COVID-19 cases. This discrepancy can largely be attributed to the fact that most testing has been performed on people who were already very ill and presenting with symptoms.
The 3,300 study participants were recruited through targeted Facebook ads and intended to represent cross-sections of the total population of the county based on ZIP code, age, sex, and race and ethnicity. White women were overrepresented in the study, so researchers adjusted the analysis to give more “weight” to underrepresented populations of Asian and Hispanic people who represented smaller portions of the testing pool.
While the study contained certain biases that cannot be adjusted for (for example, it only tested people with cars and internet access and gathered no information on the local homeless population), its data can help generate a more complete picture of the spread, severity, and fatality of COVID-19.Â
For instance, because the population of total infected people in Santa Clara County could be up to 80 times greater than current testing reflects, that would mean that the rate of death could be up to as many times smaller that widely assumed.
The study’s findings could have ramifications concerning shelter-in-place practices as well as treatment of current COVID-19 patients. Antibodies from the plasma of recovered COVID-19 patients could help patients fend off the virus.
Another implication: If a significant portion of the population has already been infected, and their antibodies can actually protect them against reinfection, shelter-in-place measures may be able to be lifted sooner
However, at a press conference Friday, the World Health Organization warned that no evidence exists that antibodies can in fact prevent reinfection or cause immunity. In South Korea, 163 recovered patients have tested positive again for COVID-19, tempering hopeful analysis of the Stanford study.
