False reassurance and Mass rapid testing for asymptomatic Coronavirus infection and infectiousness

This requires a little bit of background. The Johnson PM of the UK government announced a ‘moonshot’ programme, many months ago in September 2019, in which there would be mass daily screening for C19, for the asymptomatic, at a cost of over a billion UK pounds.

… literally millions of tests processed every single day … theatres and sports venues could test all audience members on the day and let in those with a negative result, all those who are not infectious … workplaces could be opened up to all those who test negative that morning and allow them to behave in a way that was normal before COVID … Our plan – this moonshot that I am describing – will require a giant, collaborative effort …

https://www.gov.uk/government/speeches/pm-press-conference-statement-9-september-2020

Today, January 2020, the government has begun and now ‘paused’ such a programme due to start in schools because of worries it may do more harm than good.

The key point is that despite doubts about the accuracy of the test and the risks of false negatives, the UK government has been persistent in wanting to roll this out, in order, it claims, to keep pupils in schools and ‘to keep the economy running’. Because of the inaccuracy of this rapid lateral flow ‘moonshot’ test and the collateral harms it will lead to, and I suggest that the UK government’s primary goal is to keep the economy running, even if, as is probable this will cause increased viral community transmission of C19.

Briefly, the test being used is a rapid ‘lateral flow’ test (Innova currently). It was shown in a Liverpool study to only pick up 50-60% of cases. The test is said to be 50-60% sensitive, and other studies suggest this may be an over-estimate. This means it misses 40% of cases, or in other 40 % of negative results will be false negatives.

A false positive result occurred in two of 2981 PCR negative people—a specificity of 99.93% (99.76% to 99.99%). But lateral flow tests missed 23 of the 45 PCR positive participants, giving a sensitivity of 48.89% (33.70% to 64.23%).

Covid-19: Lateral flow tests miss over half of cases, Liverpool pilot data show. BMJ 2020;371:m4848

This means that as many as 50% of people, (students returning to university, school children , or potential visitors to care homes) will be wrongly reassured that they are Covid negative. There have been calls from scientist, including Jon Deeks the co-chair of the Royal Statistical Society’s diagnostic tests group, as well as Gill and Muir-Gray for the government to urgently re-think its strategy.

Note, as well, that although the tests is highly specific – good at identifying positives correctly, this means that because the prevalence of true positivity is low in the general asymptomatic population, then even this high specificity still produces many false positives:

The false positive rate of 0.6% means that at the current prevalence in Liverpool, for every person found truly positive, at least one other may be wrongly required to self-isolate. As prevalence drops, this will become much worse.

https://www.bmj.com/content/371/bmj.m4436

In terms of the poor sensitivity and high false negative rates : the danger of having large numbers of people under-diagnosed occur when the negative result is used as a green light for social mixing (e.g. in schools) to continue, because this means that large numbers that are actually infectious will continue to mix, and, crucially, may mix in increasingly risky ways than if they hadn’t had a test at all. At the same time, although many true positives can be told to self-isolate reducing social transmission, all the same this benefit may be outweighed if those testing negative behave as if they are really disease free when they are not, in ways that increase risks of transmission.

In other words, false reassurance may well lead to increasing risky behaviours, rather than just having no effect at all on the risk of transmission compared to the situation where no testing had been done.  In other words, the key question is: What is the net effect on transmission rates of testing compared to no testing?

Understandably some scientists are asking for real world empirical studies to find an answer, but there are no signs of such studies forthcoming, and no doubt they would be difficult to do and interpret, given potential confounders and rapidly changing social circumstances and regulations.

Some have argued that the worries about false reassurance increasing risky behaviours isn’t borne out because, for example, the introduction of seat-belts, and cycle helmets etc. didn’t lead to increasingly risky behaviours. I think that this comparison is invalid for four reasons:

a) The test signifies ‘freedom from danger to others’, a freedom from a need to be careful, from guilt or vulnerability, whereas the hat/belt signifies reduced, but not zero, personal risk but, rather, is a continued concrete physically present, and sensed, empirical reminder of vulnerability;

b) Unpredictably, people may or may not want, or value, a negative result and will tend to believe or not believe it depending upon either, for example, they experience : (i) levels of fear of the disease or fear of loss of income if told to self-isolate; or conversely (ii) a lack of fear of the disease and possible value from secondary gains from a negative result due to, for example, avoiding stresses by being permitted or told to take time away from work or school;

c) Even though under-diagnosis may be, and can be, identified as a false-reassurance and potential risk and, although public health messaging may emphasise this risk to the public, all the same under-diagnosis is an abstract concept, not an empirically experienced event, and so, I suggest, can never be fully grasped or ‘felt’ by individuals as having any personally relevant meaning that should influence individual behaviour. This means, in effect, a ‘negative’ result will always erroneously signify, and be believed by the individual to mean, total personal freedom from infection and infectiousness.

A comparison can be made with the risk of over-diagnosis with mass cancer screening, where once diagnosed the possibility of over-diagnosed becomes disavowed, that is it may be ‘known’ but has no affective value, it becomes a ‘unknown known’, and the diagnosed and treated person focuses on cure as a cancer survivor, not as a possible over-diagnosis victim.

d) individual protection versus protection of others: a negative test result, used in order to avoid or prevent self-isolation, permits continued and, if behaviors change, may even increase social mixing, potentially increasing risk to others; on the other hand, for example, wearing a cycle helmet only reduces the dangers of an existing behaviour to the individual, it prevents continuation of the pre-helmet risk to the individual and doesn’t change existing cyclist behaviour in ways that makes it riskier, or safer, for others.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s