One more quick word on the PCR testing kerfuffle. Our local paper ran a story on the testing issue this morning. They very kindly quoted some of what I have said about the issue, for which I am appreciative. I do think it is important to put this in context.
The use of over-sensitive cycle numbers is only a problem if it affects a significant number of tests. If less than 1% of tests were a positive because the lab used a cycle number of 35 or higher, it really wouldn’t matter much. What was alarming about the New York Times article was that the experts they talked to said that as many as 90% of all “positive” results were non-infectiousness because it took so many cycles to reach the threshhold. And when more cycles are used, there is more opportunity for actual false positives, as opposed to non-infectious ones, to occur. Contrary to what some have suggested, there is a good body of research correlating cycle number to culture positivity. It is pretty much common sense that if there are more total virus fragments detected in a sample, it is more likely that some of it is viable virus that would replicate in culture. If you are using a cycle number limit of 38, as Minnesota is, almost all samples near that limit are not infectious, according to this body of research.
It is clinically and epidemiologically relevant to know what the cycle number was when the sample reached the threshold of positivity, and to know the distribution of positives versus cycle number at threshold, and what the trends are in that data. Again, when this kind of basic information is not released, people naturally wonder why. Just release the damn information. If it only affects a very few positive results, then we can rest easy that it really isn’t much of a problem. If it relates to a large number of tests, however, it would be appropriate to ask what steps have been taken to validate that the marginal results do reflect people who are infectious, and what public policy should be if in fact the number of true “cases” is dramatically lower than has been reported by the state to date.
This is particularly important when people literally are talking about closing schools because of one “positive” test result. In those cases, no action should be taken until it is verified that the positive result likely reflected infectiousness. If the appropriate information is provided with the test result, that takes no extra time. And it is also very easy to do double testing, to avoid other potential false positive issues.