Note: Happy Thanksgiving!  Q&A will return on Tuesday, November 30th.  In the meantime, if you have questions, please send them along.  

Question: How sensitive and specific are the COVID tests today (I’m more curious about the PCR tests) compared to the beginning of the pandemic? If I recall correctly, it was estimated in March last year that around 1/3 of negative tests were possibly false negatives and there were all kinds of issues with them, including CDC having problems with their own tests. 

Answer: You’re right to recall that the original PCR test that CDC developed at the beginning of the pandemic was deeply flawed, so much so that it had to be recalled in mid-February 2020.  This huge mistake stalled the U.S. pandemic response, leaving us blind to COVID spread in the early months.  Later in 2020, we learned that the original CDC test could fail up to 33% of the time.  Since that January/February 2020 debacle, however, COVID PCR tests have been known for their accuracy.  Systems may have improved to increase throughput and to reduce testing time, but I’m not aware of improvements in PCR test sensitivity or specificity, which have been very high.  The main thing that continues to change is COVID-19 incidence– as incidence increases, false positivity rates decrease and false negativity rates increase.  I’ve included herein a table I made last year to show how false negative/positive rates depend on incidence (Table 1).  And for more on sensitivity/specificity, see Q&A of 11/6/20.

Table 1. False negative and false positive results depend on the true level of incidence.