Although the UK has one of the highest rates of COVID-19 testing in the world, we are struggling with demand - with Baroness Dido Harding outlining to the Health and Care Committee on 17 September that available testing capacity was oversubscribed by a factor of 3-4 times. Many are "seeking tests without symptoms" against current guidelines, but then the majority of infections are believed to be asymptomatic. What could we have done differently to buy us some more time before we hit our limits?
The bottleneck is not really the number of tests that can be produced but rather the number that can be processed. Interestingly, last week research on newer saliva-based tests highlighted that such tests lend themselves to pooled sampling. Much like one of those "guess the weigh" logic problems, the concept is simple. These saliva tests use the polymerase chain reaction technique to massively replicate any genetic material and then detect whether particular target regions are present. Pooling involves taking a small sample from each of a number of saliva tests and then testing the aggregated material in one go. If the aggregated test proves negative, all the underlying individual samples can be assumed to be negative. If the aggregated test proves positive, then you test the individual samples one by one.
However, and here is the catch - pooled sampling also works on the current swab tests. Indeed, a number of countries, such as China, USA, India and Germany are already using pooled sampling with swab tests. There are indeed a number of different sampling techniques, such as the use of overlapping samples, as set out in the following illustrations from Nature.
The mathematical strategy that could transform coronavirus testing - 10 July 2020, Nature (https://www.nature.com/articles/d41586-020-02053-6)
As long as the positive infection rate is relatively low, these techniques mean that somewhere between a tenth and a sixth of the number of tests need to be processed to achieve the same level of identification.
A reasonable question might be why the UK is not using similar techniques to significantly increase predictive value of each test that is being processed. Of course there may be other factors that mean that processing capability is not the only the bottleneck. But before you go - let me share one other piece of information.
The concept of pooled sampling is not an idea that developed during this year in response to the COVID-19 crisis. Back in 2012, researchers in the USA were successfully experimenting with the use of pooled nasopharyngeal swabs for influenza virus in response to the challenges that were posed by the Influenza A(H1N1) pandemic of 2009.
We should be rightly impressed by the speed of innovation that we are seeing in so many fields in the detection, prevention and treatment of COVID-19 this year. However, it is very regrettable that successful trials of pooled nasopharyngeal swabs from 2012 that were precisely intended to address the need for mass testing in the face of a pandemic have been overlooked in the UK.