A study carried out by researchers at the University of Southern California estimated that approximately 4.1% of the people in Los Angeles County may have been infected with coronavirus, despite many not showing any symptoms.

This would be both bad news and good news: bad news because it means the disease is far more widespread than we thought, and good news because it means that the fatality rate is much lower, since the total number of cases is higher.

However, experts not involved with the study have criticized the methodology and have warned to be wary of drawing any clear conclusions.

The study relied on antibody tests. Unlike diagnosis tests — which assess whether someone has the disease at a given moment — these antibody tests determine whether someone has had the disease and developed an immune response to it. According to this study, 4.1% of the county’s adult population has antibody to the virus. Adjusted for the margin of error, that’s 2.8% – 5.6%, or 220,000 to 440,000 adults.

The immediate takeaway is clear: silent carriers are a big problem and we should all behave as though we have the virus and take measures to prevent spreading it. Simply put, whether or not we have the virus, we should behave as though we have it.

“Given the high rate of people that may have been infected at some point with COVID-19, we need to assume that at any point in time that we could be infected and that all the other people that we come in contact with could also be infected,” Public Health Director Barbara Ferrer said.

The first study was carried out on 3,300 Santa Clara residents in early April, featuring the same type of test (and some of the same researchers). The second study recruited volunteers for testing on social media, which was criticized as an improper methodology by several experts.

Unlike the Santa Clara study, the USC study used subjects representative of the county population who were part of a market research firm database. Neither study was peer-reviewed, a critical step to publication.

But the real problem with these studies is the test itself.

We’ve all put a lot of hope in antibody tests which could show how widespread the immunity is in the overall population, but increasingly, it is becoming apparent that most (or all) antibody tests aren’t performing accurately.

Essentially, antibody tests have already been shown to not work in all situations, and the WHO has even warned that only a minority of people seem to develop these antibodies, which casts substantial doubts on the validity of antibody tests (and raises concerns about how long a potential immunity to the disease would last, but this is another discussion). For now, no antibody test has been shown to be sufficiently accurate in peer-reviewed science, and there are still many unanswered questions about this type of test.

Testing has been problematic around the globe, as most countries and states have been plagued by a shortage of supplies — not only when it comes to testing, but also for things such as protective clothing and face masks. California, like the rest of the world, needs to have a more robust testing capacity, and it needs a bit more time to validate its tests: only when this is done, can we start talking about reopening the economy. If we act earlier, we risk opening the door for a second, more devastating wave of infections.