an autodidact meets a dilettante…

Canto: So we’re looking at medcram coronavirus update 98 now, and it’s a fascinating one entitled ‘Rapid COVID 19 Antigen Testing at Home – A Possible Breakthrough’, though it comes with the clear proviso – this would require co-ordinated political action, and that won’t happen in the USA, not just under Trump, but at any time.

Jacinta: Well, but especially under Trump. But the issue is one of trying to get much more testing done, with far less emphasis on the sensitivity of the test, because rapid-fire, fast turnaround testing is far more useful than expensive, hard-to-evaluate slow-turnaround testing, which puts a premium on sensitivity. But before we get to all that, Dr Seheult looks at a paper on viral loads vis-à-vis covid19 patients. They looked at nasal and throat swabs, and then checked the Ct values over time. 

Canto: The Ct values are a measure of viral load and it works inversely – a 3.32 reduction in Ct value means a ten-fold-increase in viral load. 

Jacinta: Yes, so a low Ct value means a high viral load, and of course viral replication works exponentially, at least during the early infection period, so your viral load can be massively different from one day to the next – think about that for testing, and delayed results. 

Canto: A Ct value of 40 is close to undetectable, depending of course on the sensitivity of the test. And the value can go down as low as 5, all approximately of course. The course of the virus is generally, exponential growth, then a tapering off of the growth rate, reaching a peak, then a descent to finally a remnant population of largely disabled viral scraps, with of course mortality intervening along the way in the worst cases. 

Jacinta: Far from the majority of cases, thankfully. So the ‘gold standard’ test is the reverse transcriptase polymerase chain reaction (rt-PCR) test – also called real-time PCR, I’ve just found out. It’s relatively expensive at around $US100, with turnaround times – and this might depend on demand and other factors – of between 3 and 9 days. There aren’t enough of these tests to go round, but they are very sensitive, detecting the virus reliably from a Ct value of about 35, or maybe even 40 (for argument’s sake, Seheult says). But there are other, cheaper, less sensitive tests, called paper tests, that can be rolled out more easily to the general public. The paper is coated with monoclonal antibodies that can detect antigens – substances that evoke an immune response. These paper tests cost at most a couple of dollars each, and would be sensitive to a viral load measured at a Ct value of around 32. These figures aren’t exact but this would make the test around 50-55% sensitive. 

Canto: But there’s this issue called the ‘threshold of transmissibility’, which is important in all this, and a virologist, Dr Michael Mina, shown speaking on this update, explains:

So people who are transmitting probably have Ct values that are below 30 and the vast majority probably have them below 25 or so.

As Seheult explains, people may be testing positive at that range above 30 (i.e. low viral load) but not transmitting the virus. This is especially so if they’re on the downward trajectory, as described above, and what the rt-PCR test (or assay) is detecting are those remaining viral fragments. And as Dr Mina points out, it’s the downward trajectory that’s being picked up for the most part, because the initial upward trajectory is exponential. Here’s what he says:

A lot of people are saying, ‘we need the really sensitive tests to be able to detect people early on in their infection’, but almost all the time that people spend with this virus near the limit of detection of PCR is on the tail end of their infection. This is a virus that, once it hits PCR positivity levels, it’s growing well in its exponential phase and it’s probably a matter of hours, not days, before it passes the threshold to be detected on some of these slightly lower sensitivity assays. And then it may persist for weeks or possibly months even in some cases at very low RNA levels. So it’s after people are well beyond their transmissible period that we’re actually seeing the loss in sensitivity of these assays. It’s very rare that you actually detect somebody with a Ct value 0f 39 in that window on their way up, because they’re only sitting there for a few hours before they get down to a 33, so if you’re missing Ct values of 39… it’s really not that important..

Jacinta: Not that important, but the point Dr Mina is making is really important – if the threshold of transmissibility is at 33 or below vis-à-vis Ct values, then a high-sensitivity test may even be a barrier to focussing on getting at the most transmissible subjects. 

Canto: Yes, especially when you have an alternative test that can be applied much more regularly with a quick turnaround – results on these paper tests take ten minutes! And being cheap, you can test as often as you feel you need to. If you’re positive, you quarantine yourself for a while, keep testing, find yourself negative, wait for a few more days, considering the low sensitivity of the test, keep testing in case there’s a recurrence, and when it’s still ok after a few days you can resume your life, go back to work or school, whatever, being pretty sure you’re past the infectious phase. 

Jacinta: Yes, as Dr Mina says, 9 out of 10 people go undiagnosed with the virus in the USA, according to the CDC – indicating the inadequacy of testing. And he goes on to say, of the other one out of ten, those that are caught, are mostly post-infectious, at the ‘tail end’. The point is that, because of the woeful lack of testing and the long turnarounds, they’re catching far fewer of the transmissible cases, the ones they want and need to catch, than the pitiful few that they actually find testing positive. 

Canto: The bottom line being that if they tested with a far less sensitive, but cheap and readily available quick-result assay, they would capture far more of the transmissible cases, and save lives. 

Jacinta: Dr Mina and many colleagues have written a paper on this, entitled ‘test sensitivity is secondary to frequency and turnaround time for covid19 surveillance’, and he points out that with this approach they would drive down the ‘r effective’ – the reproduction number – which is the number of people who can be infected by a carrier at any specific time – to well below 1. So if you were to give a significantly high proportion of people in in the worst affected areas these types of tests, you could bring the numbers down very rapidly, and this would eliminate the need for contact tracing. It would have an effect on schools, workplaces and so forth – because if you’re given one of these long-turnaround tests and your results eight days later turn out negative, that may be because you had just contracted the virus when the test was taken, but it didn’t show on the test – so you go back to school and infect people. With regular testing this problem would be eliminated. Hate to belabour the point, but – people are dying. 

Canto: It seems the CDC put a high priority on sensitivity, and so rejected these cheap paper tests, neglecting the obvious problem of turnaround more or less completely. The low sensitivity tests usually miss the subjects that are beyond infectivity. If they were on the upward trajectory they would likely be caught by the next test. It’s this upward trajectory that is the infective period. You would think regulatory organisations like the FDA or the CDC would twig to this, but not so much in the Trump era, when non-scientists are put in charge. Yet another failing of the individualist, anti-collaborative, egotistical destruction of all government agencies…

Jacinta: Or just the unwieldiness, the lack of finesse, of lumbering bureaucracies. Or a mix of both. 

Canto: Anyway, as things deepen and darken in the USA, we might have to skip a number of updates to keep up with the chaos, the failures, the resistance and everything else. It isn’t a great time to be living in the USA, but as outsiders we’re kind of ghoulishly fascinated by the mess they’re making of this pandemic, and much else besides.

Jacinta: But also genuinely sympathetic to those who are trying to make thing work in the teeth of all the craziness. 

Canto: Today, September 16, marks the day that 200,000+ deaths from covid19 have been recorded next to the name of the USA, according to Worldometer’s stats. Taiwan, a country that is separated from the so-called ‘China virus’ only by the narrow Formosa strait, has suffered only seven deaths in the nine months that this virus has raged. It has a population of about 24 million, slightly less than that of Australia, into which we find ourselves thrown. Australia has had 824 deaths thus far, and we’re regarded as something of a model!

Jacinta: Yes, several cheers for Tsai Ing-wen and for female leadership, sans egotism. And a special thanks to Audrey Tang, Taiwan’s digital minister – but she’s actually real, and a life-saver. We need more of her. 

Audrey Tang