Two recurring news items raised substantial noise in the media concerning COVID-19.
The first item was the publication of the reported case of reinfection from Nevada. This case report was published in the Journal “Lancet Infectious Disease” and purports to show the first individual in the United States who suffered a re-infection with coronavirus.
The evidence for the reinfection is a finding that changes in the genetic sequence of the virus from the two episodes of infection were too different to have happened by chance mutations. While the evidence of the genetic sequences was convincing, and the authors had also shown that the swabs came from the same individual, there was one problem in agreeing with the authors’ hypothesis.
There was no evidence that this patient was actually infected the first time he was labeled positive by the PCR assay. In particular, there was no identification of an immune response after the first episode indicating a productive COVID-19 infection. This means that the first episode PCR positive could have been from contamination of the patient sample or debris from coronavirus material that was not infectious.
The concern about a real first infection with COVID-19 was reinforced since the patient was more ill with the second infection, something not expected even with partial immunity from the first infection. While several scientists made tortured attempts to explain how this might happen, all were highly speculative and not based in data.
And, now that almost 8,000,000 Americans have been shown to have COVID-19, even if this report is true, it is an incredibly rare phenomenon.
The second issue involved COVID-19 testing problems. The test in question this time was the Becton Dickinson BD Max PCR test, which received Emergency Use Approval on April 8, 2020. This is a PCR molecular test for SARS-Cov-2 virus and was given by the federal government to nursing homes to screen their patients. Because the results were used to place infected patients together, any errors could be potentially deadly.
The federal government is now also investigating the BD rapid COVID-19 assay that was sent to nursing homes by the federal government. No data on that is available, but again it involves false positives.
Both BD assays were reportedly giving false positive results. This could mean that uninfected patients who had a false positive test were being placed with infected patients. The FDA has sent out a notice informing users of this possibility but suggested that the false positive PCR rate was approximately 3%. Unfortunately, if the true positive rate was only 3% as we’ve seen in many situations, then half of the positives identified would be false positives.
It is interesting that the assay most people consider as a “gold standard” would have this problem. It is in some ways worse than a false negative test, which is usually confirmed by a second assay if the patients is sick. Bret Giroir, testing czar for the government, when asked about this suggested that “false positive tests are to be expected.” I don’t think that is usually true with this technique.
The most recent PCR false positive problem was the one that kept the CDC COVID-19 test from being used for 6 weeks early in the epidemic. It was due to accidental contamination of the kits. BD needs to examine every possible cause of this problem.