When testing accuracy comes into question

There’s been a lot of excitement about antibody tests that might be able to tell people when it’s safe to go back to work. But there have been reports of unproven and fraudulent tests flooding the market.

That happened in part because during the FDA loosened restrictions in mid-March and allowed companies to sell antibody tests, also known as serological tests, without proving that the tests worked. There was pressure to get antibody tests on the market. But the decision was a trade-off, FDA officials acknowledge, and accuracy suffered.
 

Now the FDA will require test makers to submit emergency authorization requests and to send in validation data for their antibody tests. The FDA is also setting specific performance recommendations for all test developers and will work with the National Cancer Institute to independently validate antibody tests on the market.

There’s also another area we need to look and it's the more common test, called a PCR test. It’s the one that detects if you currently are infected with Covid-19 based on a saliva or nose swab.

Much of the conversation has focused on increasing the number of tests, but there’s another number that may matter even more: the accuracy of the test. The gold standard to diagnose Covid-19 has been the RT-PCR test, a test that finds and measures genetic material known as RNA from the virus. But some surveys are finding that these tests aren’t always correctly identifying who is sick.

A recent survey from the Cleveland Clinic found that five of the most common tests had a 15% false negative rate – meaning the test result was negative even though the person was, in fact, infected. It’s crucial to get more accurate tests so we can measure how many people really are sick, and start thinking about when it will be safe to reopen America.