A much lower CoVID death rate is warranted, based on antibody testing to SARS-CoV-2 in California.
Positive antibody tests indicate that those persons’ bodies had met the virus, and their immune systems appropriately mounted a response by generating antibodies to neutralize the virus.
As previously reported here and elsewhere, the bottom number (denominator) of the death rate calculation depends on knowing how many have been infected in total, which depends on adequate antibody testing.
The new information comes from two recent studies showing neutralizing antibody rates much higher in the California population than accounted for by the count of known infections.
On April 30, researchers from Stanford reported that antibody testing of 3,300 persons in early April in Santa Clara county detected 50 to 85 times more than the official cases known by that date.
The range estimate comes from the potential error factor of the test itself.
The numbers are somewhat different in Los Angeles County, where another study tested 863 adults for antibodies in the second week of April. The testing found 4.6% of the sample had antibodies, as reported in the Journal of the American Medical Association (JAMA).
When projected across the entire LA County population, it is likely that some 367,000 County residents had already met the virus and had developed immunity by then.
In contrast, the official count of confirmed CoVID cases in LA County by April 10 was 8,430. At that time, it was mostly only the obviously ill that sought medical attention.
This new evidence revises estimates of the number infected (and immune) by a factor of 43 – 85 times.
The word ‘infected’ here has to be used with full understanding: despite having met the virus, the vast majority with positive antibody tests never had symptoms, so actual disease would not have taken hold.
This blog previously reported on the uncertainties surrounding antibody testing, namely, that the tests have not been sufficiently independently validated.
No such antibody test is 100% accurate, but the reliability of this test is not known. It is uncertain how many are false negatives (a person actually has antibodies but the test reads negative) or false positive (a person does not have antibodies but the test reads positive).
The researchers in both studies attempted to take this into account. In the LA county study, they adjusted the statistics by considering the test might have missed a many as 17% of those that really did have antibodies.
In the opposite direction, they estimated that the test might have incorrectly given a positive result in about a half percent that really did not have immunity.
While the authors admit these estimates might not be correct, they have given it a good scientific guess by considering how similar viral antibody test kits have performed in the past.
Even given the uncertainties, these finding are substantial enough to warrant a recalculation of CoVID numbers in at least three significant directions:
- Potentially 43-85X more persons have already met the virus;
- Potentially 43-85X more persons are immune;
- Potentially 43-85X larger denominator, making the death rates significantly lower.
We can expect these results to be substantiated in larger populations when reliable, independently validated antibody tests become widely available.
This good news was predicted here and by many others, not because anyone had crystal balls, but because this is what routinely occurs with infectious diseases, amounting to no more than first-year medical student wisdom.
Various other unvalidated antibody tests have received emergency use authorization from the FDA in March. On May 4 the FDA mandated that labs share their data with the agency to validate accuracy.
By May 21 the FDA issued a list of 27 antibody tests that had previous approval withdrawn.
Tip for the smart health care consumer: As antibody testing increases, expect reports of more total past cases (meaning more immunity), which profoundly drives down the calculation of death rates.
Take heart, things are gonna get brighter (with regards to Stan Vincent, O-o-h Child)!