Sorry to keep harping on the COVID-19 pandemic, but it seems our waking hours are consumed with lots of hypotheses (of origin, rate of infectivity, and projections of outcome over the next few months), conspiracy theories, and proposed treatments. This study from Stanford (just out; see below) might be important for several reasons:
[a] If the rate of infection is far greater than that reported, then in human populations many more people carrying the virus are asymptomatic than previously thought.
[b] These data suggest a greater likelihood of “herd immunity” might be able to occur.
[c] These findings also suggest the currently reported fatality rates of ~2-3% are too high, i.e. the actual COVID-19 infection fatality rate is between 0.12% and 0.20%.
[d] Therefore, statistics of “COVID-19 death rate” divided by “number of COVID-19 confirmed cases” — makes little sense, if the total number of infected cases is 50-100 times greater than the “number of confirmed cases.”
[e] Lastly, this article has been uploaded on the preprint server medRxiv, which as we all know, is one of Fred’s favorite web sites. 😉 Thus, the study need not be in final form; it may likely be revised before submission to a journal.
More than 48,000 Santa Clara County residents likely been infected by coronavirus
Survey of blood samples suggests between 2.5% and 4.2% of county residents may have coronavirus antibodies
by Gennady Sheyner / Palo Alto Weekly
Updated: Fri, Apr 17, 2020, 7:47 pm
The number of coronavirus infections in Santa Clara County could be between 50 and 80 times higher than the officially confirmed count, preliminary results from a community-based study by a team of Stanford University researchers indicates.
The prevalence study, led by Stanford Assistant Professor Eran Bendavid, has not been formally published and is still undergoing peer reviews. It has, however, been published on the preprint server medRxiv. As such, it is effectively a first draft, subject to change, based on input before formal publication.
That said, the early findings indicate that between 48,000 and 81,000 residents in Santa Clara County were infected as of April 1, back when the official count was 956. The estimate is based on 3,330 blood samples that were taken from volunteers in Mountain View, Los Gatos and San Jose on April 3 and April 4 and tested for antibodies to SARS-CoV-2 .
When adjusted for Santa Clara County’s population and demographics, the number of positive results suggests that between 2.49% and 4.16% of the county’s 1.93 million residents have had COVID-19.
The study’s results “represent the first large-scale community-based prevalence study in a major U.S. county completed during a rapidly changing pandemic, and with newly available test kits,” the authors wrote.
The most important implication, the preprint notes, is that “the number of infections is much greater than the reported number of cases.”
“The population prevalence of SARS-CoV-2 antibodies in Santa Clara County implies that the infection is much more widespread than indicated by the number of confirmed cases,” the researchers concluded. “Population prevalence estimates can now be used to calibrate epidemic and mortality projections.”
Jay Bhattacharya, a professor of medicine at Stanford University and one of the study’s authors, said the goal of the study is to understand how widespread the disease is.
“To do that, we need to understand how many people are infected,” Bhattacharya told this new organization on April 4, as the second day of tests was kicking off. “The current test people use to check whether they have the condition – the PCR (polymerase chain reaction) test – it just checks whether you currently have the virus in you. It doesn’t check whether you had it and recovered. An antibody test does both.”
Participants in the prevalence study were targeted through Facebook ads, with the goal of getting a representative sample of the county by demographic and geographic characteristics, the study states. Because the sampling strategy relied on people who have access to Facebook and a car, there was an overrepresentation of white women between 19 and 64, as well as an under-representation of Hispanic and Asian populations, relative to the community, according to the study. The study attempted to compensate for that by weighting the results for race, sex and ZIP code so that they better reflect the countywide population.
The group’s analysis indicated 50 blood samples from the study, or 1.5% of the total, tested positive for either immunoglobulin M (IgM), the antibody that the body produces when the infection occurs and that disappears after several weeks, or immunoglobulin G (IgG), the antibody that appears later, stays longer and provides the basis for immunity.
After weighting to match the county population by race, sex and ZIP code, the prevalence rate was adjusted to 2.81%, according to the study. Other factors, including uncertainties relating to the sensitivity of the tests that were used, contributed to the range of up to 4.16%.
County, state and federal health experts have consistently acknowledged that the number of COVID-19 cases is far higher than the official statistics show, a problem they attribute largely to the lack of widespread testing. Even though California is looking to greatly ramp up serological (blood) testing and to establish new community-testing sites, the state continues to experience both a shortage of tests and a backlog in processing tests.
As of April 15, more than 246,400 tests had been conducted in California. In Santa Clara County, there were 17,774 tests completed as of April 17, with 10.52% testing positive for the coronavirus.
The new study suggests that the undercounting of COVID-19 infections — the extent to which they vary from official case numbers — is far greater than has been assumed.
“The under-ascertainment of infections is central for better estimation of the fatality rate from COVID-19,” the study states. “Many estimates of fatality rate use a ratio of deaths to lagged cases (because of duration from case confirmation to death), with an infections-to-cases ratio in the 1-to-5-fold range as an estimate of under-ascertainment. Our study suggests that adjustments for under-ascertainment may need to be much higher.”
The Stanford study suggests that the undercounting of cases can also be attributed to a lack of widespread testing and reliance on PCR for case identification, which misses “convalescent” cases (those who have already recovered from the infection). The official count also doesn’t capture asymptomatic or lightly symptomatic infections that go undetected, the study states.
The range of results also reflects uncertainty in both test sensitivity (how good it is at correctly identifying COVID-19 antibodies) and test specificity (how likely it is to produce a false positive). Researchers relied on tests manufactured by the Minnesota-based company Premier Biotech, rather than the newly developed serological test by Stanford, which has been used to test health care workers.
Bendavid told this news organization earlier this week that the tests were chosen because they are very easy to use (they produce a line reading similar to a pregnancy test) and produce results within 15 minutes. They are, however, less precise than laboratory-based tests and give you an underestimate of how many people have coronavirus – a shortcoming that was factored in the study.
To determine their accuracy, the research team used the kits it received from Premier Biotech to test blood samples from Stanford Hospital patients that were shown to be positive through a DNA test, as well as samples that were known to be negative because they were taken before the pandemic. These results led researchers to conclude that the sensitivity is about 91.8%, a rate that was factored in to produce the final range.
The authors acknowledge the study’s other limitations. While they factored in sex, race and ZIP code, the survey does not account for age imbalances or a potential bias, favoring individuals who were in good health and, therefore, able to volunteer. The effect of such biases, the study notes, is hard to ascertain.
Bendavid and Bhattacharya had both argued in the past that the COVID-19 fatality rate is far lower than many experts had assumed. That’s because the number of actual infections far exceeds the official case counts.
“If the number of actual infections is much larger than the number of cases – orders of magnitude larger – then the true fatality rate is much lower as well. That’s not only plausible but likely based on what we know so far,” Bendavid and Bhattacharya wrote in a Wall Street Journal opinion piece on March 24.
As of April 10, the study notes, 50 people in Santa Clara County had died of COVID-19 in the county, with an average increase of 6% daily in the number of deaths. Given the trajectory, the study estimates that the county will see about 100 deaths by April 22.
Given the study’s estimate of 48,000 to 81,000 infections in early April – and a three-week lag from infection to death – the 100 deaths suggest that the actual infection fatality rate is between 0.12% and 0.2%.
That’s a far cryt from the county’s mortality rate based on official cases and deaths as of April 17 — 3.9%.
The study states that the new data “should allow for better modeling of this pandemic and its progression under various scenarios of non-pharmaceutical interventions.”
“While our study was limited to Santa Clara County, it demonstrates the feasibility of seroprevalence surveys of population samples now, and in the future, to inform our understanding of this pandemic’s progression, project estimates of community vulnerability, and monitor infection fatality rates in different populations over time,” the study states.