Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Rapid Covid tests miss 90% of asymptomatic cases

nature.com

Post reply on HN
Print view View original

1–10 of 236 posts

Re: Rapid Covid tests miss 90% of asymptomatic cases

#5

Post by henry-j »

Referenced paper: https://doi.org/10.7326/M23-0385

Not much behind the paywall of the article OP linked:

“Rapid antigen tests are much more reliable at detecting COVID-19 in people with symptoms than in those without, finds the largest study to compare home rapid tests with gold-standard PCR tests1.

Apurv Soni at the UMass Chan Medical School in Worcester, Massachusetts, and his colleagues, followed 5,353 initially uninfected people between October 2021 and January 2022, when the Delta and Omicron variants of SARS-CoV-2 were circulating. For 15 days, the participants regularly took rapid tests at home and sent samples to a laboratory for PCR testing.

More than 150 participants tested positive for SARS-CoV-2 on PCR tests. On the day of infection onset, rapid tests detected almost 60% of infected participants who had COVID-19 symptoms, but only 10% of those who didn’t. However, repeating the test 3 times, 48 hours apart, led to detection of infections in 75% of asymptomatic participants. Two tests conducted 48-hours apart identified 92% of symptomatic participants.

The authors say that an informational campaign about the proper way to take rapid tests might help to preserve public confidence in the tests’ utility.”

Re: Rapid Covid tests miss 90% of asymptomatic cases

#7

Post by yabones »

In my anecdotal experience, rapid tests are only really good for telling you if what you're sick with is covid or not, and barely even that. They will not give you a predictive result, and will not identify an early stage illness, only when you're at or near the peak of symptoms.

Re: Rapid Covid tests miss 90% of asymptomatic cases

#8

Post by Symmetry »

Duh, they only detect levels of viral load sufficient to make a person infectious, around 1e6 RNA copies per cc of fluid. And viral load tends to peak somewhere in the 1e10 to 1e14 range, people at 1e6 really have to work at coughing directly on to growth plates to create a sample.

Back in the early days of Covid-19 people didn't typically begin displaying symptoms visible without an x-ray until they'd nearly reach their peak viral load. More recently people tend to start displaying symptoms earlier. This is probably due to the adaptive immune system learning about covid through immunization or prior infection, but could also be something related to Omicron variants becoming dominant.

But in any event, it takes a while to get the results back from a test more sensitive than a rapid test. If you take a rapid test and send off a sample for PCR every hour then you'll collect a positive PCR sample long before you collect a positive rapid antigen sample. But you might very well get the results from the rapid test first since viral loads tend to increase rapidly in the early stages and can cross the ~4 orders of magnitude in the time it takes to turn around a PCR test.

EDIT: Actually, I should say and those peak viral loads I quoted are from 2020 when I was paying the most attention to this stuff and have likely changed with mass vaccination/prior infection and Omicron.

Post reply on HN