COVID-19 antibody testing problem to be solved: is the data credible? How precise is it?
With the new corona virus infection rate reaching 5%, virus detection equipment manufacturers and regulatory agencies must strictly prevent the occurrence of false positives. But what about the market?
As of now, there are dozens of COVID-19 immune antibody tests in the world, but some of them have not been approved by regulatory agencies. The US FDA, CDC and NIH recently stated that they will work together to verify the accuracy of the results of the test methods sold in the US as soon as possible.
However, these three organizations have not made clear what level of accuracy they need to achieve. At present, none of the manufacturers of these antibody blood tests have published the sensitivity and specificity data of their products. And the accuracy of many data available tests may simply not be good enough.
Earlier last week, Roche CEO Severin Schwan stated that some COVID-19 tests on the market were "worthless, or of little use.
According to data collected by EvaluatedTech, of the 29 commercially available detection methods for detecting new coronavirus antibodies in the United States, only 3 have obtained the FDA's Emergency Use Authorization (EUA), which is about 10%.
And even if there is a EUA test method, it has not been evaluated by FDA or any other regulatory agency for accuracy.
COVID-19 Antibody Test Sensitivity
However, some manufacturers claim that their products are highly accurate. On April 16, Abbott stated that the recently released IgG test method in the United States was tested at least two weeks after the first symptoms of COVID-19 patients, with a sensitivity of 100% and a specificity of 99.5%.
The antibody test sold by BD and developed by Biomedomics claims that the sensitivity is 88.7% and the specificity is 90.6%.
In addition, companies known to be conducting COVID-19 antibody or serological testing include Siemens Medical, Danaher and Diasorin.
Roche also plans to launch antibody detection methods in May.
In view of its CEO's doubts about the accuracy of existing tests, the outside world is very much looking forward to the accuracy data of Roche products.
At present, the COVID-19 detection method on the market faces two main problems.
First, is the accuracy statement of the test credible without independent verification?
Abbott and BD are both reputable companies and generally do not present data that they cannot prove, but many other antibody tests sold around the world come from small groups and laboratories, and data control may not be as strict.
Secondly, how accurate is the COVID-19 test?
Prior to the advent of the vaccine, countries that aimed to lift restrictions on isolation and hope to start accelerating economic operations will rely heavily on antibody testing to determine who can safely return to work.
The scale of verification tests conducted by these companies is very different. Among them, the antibody test conducted by Abbott included 1200 samples. While Epitope's test only included samples from 54 healthy people. COVID-19's IgM and IgG tests only had 20 and 30 PCR samples, respectively. Concerns about the comparison of data between different studies are magnified here.
In addition, the accuracy of detection is also high. According to previous estimates, the prevalence of COVID-19 in the United States is about 5%. At this level, the risk of false positives becomes a major problem.
If the serological test has a specificity of 90%, the positive predictive value will be 32.1%, which means that the false positive result is nearly 70%; while the specificity of 95% test still means a 50% false positive rate.
Only at 99% specificity, the false positive rate is close to the acceptable range, but even at 99% specificity, there will still be 16% false positive results.
The predicted value of theoretical detection when the disease infection rate is 5%
If you want to avoid a large number of people who are not immune to the virus from coming into contact with others again, global medical regulatory agencies, including the FDA, need to develop strict accuracy standards for testing methods and products.
The FDA has indicated that it will take appropriate action against companies that conduct or sell untested testing methods or publish false accuracy data, such as issuing warning letters to stop their illegal testing.
In fact, behind trying to prove the accuracy of its own detection method, there is a deeper demand of the company, that is, sales. If half of the US population is tested by COVID-19, the US market size will exceed US $ 3 billion.
The traditional large medical device companies are obviously the most likely to obtain this "cake".
Abbott said that by June this year, the company is expected to be able to conduct 20 million tests per month.
Roche said that the number of tests will be 5 times that of Abbott. This will be an amazing income.