People should be wary of COVID test results | Letter to the editor

Posted

The Leader recently reported that the North Start Clinic offers several COVID-19 tests, including the BD Veritor System for Rapid Detection. 

Being able to test for the infection is a critical part of our public health strategy but please don’t assume that a negative test means you do not have the infection. The article quoted an employee of the clinic saying “…negative test results, right around 100 percent accurate.” 

Unfortunately, this statement could lead to a false sense of security, since the test can give a false-negative reading in 16 percent of cases testing positive with the more sensitive PCR test. Typically tests are rated for sensitivity (does it generate false-negative results) and specificity (does it generate false-positive results). The term “accurate” is too general and can be misleading. www.healthnewsreview.org/ gives some great examples. The FDA data sheet on this testing system, www.fda.gov/media/139755/download, gives useful information:

1. The Veritor test was compared to the PCR test in only 226 samples, a very small number.

2. PCR detected 31 as positive but Veritor missed five of those, or 16 percent. Thus, 84 percent sensitivity.

3. Veritor did report the
195 negative cases as negative, for 100 percent specificity.

4. The test did get Emergency Use Authorization, but has “not been FDA cleared or approved.”

5. FDA states “Negative results should be treated as presumptive, do not rule out SARS-CoV-2 infection and should not be used as the sole basis for treatment or patient management decisions.”

In the future, The Leader might check its COVID-19 reporting with our Public Health Officer Dr. Locke when possible. He said I could share his comments regarding my concerns:

“The ability of tests to rule out infection (in real life situations) is way more complicated due to all the variables in testing — viral load at time of testing, adequacy of the swab specimen, handling of the specimen, and more. PCR testing, the gold standard, is generally regarded as being about 70 percent sensitive, i.e. up to
30 percent false negative. Antigen testing is even worse.  Diagnostic tests can accurately “rule in” infection but cannot fully rule it out. And it gets even more complicated — a big determinate of test sensitivity AND specificity is “pre-test probability” — the likelihood that infection is present. In low prevalence populations (like in Jefferson County) sensitivity and specificity are decreased. More false negatives and more false positives.”

Our best tools for ending this pandemic are still masking, physical distancing, hand washing and avoiding crowds.

Though I suspect they would agree, this letter is not an official statement of the Jefferson Healthcare or Public Health Boards on which I serve.

Dr. Kees Kolff
PORT TOWNSEND