Updated: May 6, 2020
According to Public Health England, over the last 24 hour recorded cycle, according to DoHSC, 8,240 individuals were tested in the UK for infection with SARS-CoV-2. Of these, only 3,009 tested positive. While we are all awaiting the delivery and roll-out of the SARS-CoV-2 antibody IgM/IgG test, it is worthwhile examining these results in some more detail.
According to Public Health England, the majority of COVID RT-PCT tests are being carried out on those already in hospital, where the test is intended to confirm clinical symptoms of COVID19. Unfortunately, our collective addiction to words over numbers leaves us with an imprecise "majority", but the implication is that at least 25% (and perhaps more) of those that we thought had COVID19 do not appear to have evidence of virus infection. There is no doubt that these individuals need serious medical attention - so where is the flaw - given our daily fascination with the daily declaration of number of positive tests. Let us go through the possibilities: 1) The sensitivity of the test - the test could be failing to detect the presence of virus. The sensitivity of any test is the proportion of tests that prove positive. Tests should be as sensitive as possible without leading to an unacceptable number of positive tests in the absence of virus. In the early days of the crisis, there were difficulties with faulty reagents, but these have been addressed and the problem is now the overall supply. Preliminary analysis in China which has been supported by more recent investigations have suggested that the sensitivity of the RT-PCT test may be as low as 75% but studies are limited and there is no agreed gold standard. In any case, the perceived sensitivity of the test is affected by the following two factors. 2) The application of the test - current guidance from Public Health England sets out how specimens should be taken from upper and lower respiratory tract. Prior research indicated that swabs from the mouth and throat were less likely to detect the presence of virus than those from the nose. Failure to follow the guidance leads to a greater risk of a false negative. However, clinicians in hospitals are not novices, so at this stage of the crisis we would not expect this to be a significant source of false negatives. 3) Virus may not be present in sufficient amount - the RT-PCR test is detecting fragments of viral RNA. The question is therefore what is the pattern of viral shedding. Prior studies indicate that low rates of viral shedding are seen in the first few days and likely peak at 6 or more days after the onset of clinical symptoms. So swabbing too early may lead to false negative tests. However, a cohort study in Wuhan detected that SARS-CoV-2 continued to be shed throughout the period of hospitalisation, so insufficient virus in those tested in hospitals is unlikely to be an explanatory factor for false negatives in a hospital setting.. There remains one further factor - 4) Some of those in hospital with clinical symptoms associated with COVID may not in fact have COVID. According to Public Health England, all those presenting in hospitals in the UK with either evidence of pneumonia, ARDS (acute respiratory distress syndrome) or influenza-like symptoms are being tested for COVID. Some may indeed have influenza. Whilst flu season peaks between December and February, continuing flu activity has been seen in the past as late as May. The classic COVID19 presenting symptoms may also be seen in other respiratory diseases such as bronchitis and sinusitis. Greater transparency over those currently in hospital with respiratory symptoms/disease will clarify how important this group could be. The aim here is not to challenge the urgent need for more "testing, testing, testing". Instead, it is to encourage us all to be less dependent on single measures, whether they be numbers of positive tests or deaths reported with COVID. We all struggle with assessing the likelihood of low probabilities. We look for simple indicators and clear guidance. Those providing guidance will welcome rigorous challenge to assumptions and sensitivities because it is only through this process that guidance will be improved and a better understanding achieved of how best to meet the health and economic challenges ahead.