What do China, the Diamond Princess, and Iceland have in common? They have the media in a frenzy over COVID-19.
At the time of writing, the novel coronavirus, SARS-CoV-2, confirmed cases approach 1.5 million people worldwide, with over 87,000 deaths. With cases and deaths on the rise in places like the United Kingdom and the United States, those numbers are sure to keep rising for the near future.
Anxiety is running high throughout the world, around a third of which is under some sort of lockdown. With more and more people becoming infected and critically ill, including the U.K. Prime Minister Boris Johnson, criticism over a lack of testing is widespread across the globe.
At the moment, conventional testing only determines if you are currently infected with SARS-CoV-2, it does not test to see if you’ve had it previously. So, the question everybody is asking right now is how many people are asymptomatic? That is, how many people have had the virus without showing any symptoms at all of the disease, COVID-19.
This is an extremely important question, not just to ease the minds of many left in limbo after self-isolating but without being tested, left to wonder if they really had the virus or just the seasonal flu which shares many of the same symptoms. It is also important because the prevailing theory is that asymptomatic cases are still infectious. Many scientists believe that the rapid transmission of the virus has been in part due to asymptomatic cases unwittingly infecting others.
There have been multiple studies already released which suggest high levels of cases which don’t develop symptoms, and the media has been touting figures between fifty and eighty per cent of cases being asymptomatic. If this is true then it could suggest that many, many more people have had the virus and therefore are now suspected to be immune to the disease it causes, COVID-19, which could potentially see life as we once knew it return much sooner than previously expected.
But how true are these extraordinary statistics?
Call me a cynic, but these felt a little too good to be true, so I decided to once again dig deep into the research of the three most prevalent stories. Here’s what I found.
The Diamond Princess
At the beginning of February news broke that a Princess Cruises ship, the Diamond, was being quarantined whilst docked at Yokohama, Japan, after a passenger had tested positive for SARS-CoV-2.
Recently, a study has once again been trotted out across the U.K. media, purportedly stating that fifty per cent of novel coronavirus cases could be asymptomatic.
The study by Mizumoto et al., was carried out on the some 3,063 passengers and crew between February 5 and February 20, 2020, of which 634 tested positive for SARS-CoV-2. Of those 634 confirmed cases, it is true that 320 (50.5%) of those were registered as asymptomatic within the parameters of the study. On the face of it, it would seem the media were telling the truth.
However, it was no dramatic flair on my part to add “within the parameters of the study” as this makes all the difference when getting into the gritty details.
The study states that the asymptomatic cases were characterised as, “both true asymptomatic infections and cases who had not yet developed symptoms at the time of data collection but became symptomatic later”. This is something called right censored data.
Further into the study it explains that, by using established complex models, the estimated rate of true asymptomatic cases (those which never developed symptoms even after the the time of data collection) was 0.35. Meaning that only 35% of those 50.5% deemed asymptomatic were estimated to be true asymptomatic. This brings the actual estimate of true asymptomatic cases down from the media-touted 50.5%, to a more modest 17.9%. The other estimated 65% of asymptomatic cases merely developed symptoms unusually late.
This study was also not a random sample of the population and therefore extrapolating conclusions for the entire world population, such as the media tend to do, is highly discouraged. The study itself also states this. Most of the passengers were sixty years of age or over and therefore are not a representative sample of the wider world population. Even if these statistics could be applied to the wider world population, the study still only estimates that around 18% of all cases could be asymptomatic, a far cry from the 50% the media headlines would have us believe.
The Icelandic Study
Iceland is pretty small. It has a population of just over 360,000 and has hit the ground running on testing all of its citizens for SARS-CoV-2, not just those showing symptoms. These are all facts. It is also a fact that DeCODE, a subsidiary of Amgen, have carried out much of the testing and that Dr. Kári Stefánsson, the company’s founder, told CNN in April that so far out of around 9,000 self-selected people tested, 50% of all confirmed cases have been asymptomatic.
Who can argue with those numbers?
Well, while the sample size seems more than large enough to produce reliable data, the reality is that less than 1% of those tests tested positive. Meaning that the real sample size is around 90 confirmed cases, in which around 45 cases were asymptomatic. Suddenly, those figures don’t seem that big or reliable.
When trying to determine the rate of asymptomatic cases, the two figures which truly matter are the confirmed cases and those confirmed cases which are asymptomatic. In this case, 99% of the data, which have tested negative, is disregarded.
Sadly, although an exciting headline, a sample size of around 90 confirmed cases is just far too small to draw any conclusions from, and certainly cannot be scaled up to the whole population.
The China 78%
As the outbreak of the novel coronavirus originated in Wuhan City of the Hubei Province in China, much of the initial data and research is naturally coming out of Chinese labs and scientists.
In the past week, you may have already heard something along the lines of ‘four out of five cases are asymptomatic’ in regards to the SARS-CoV-2 pandemic. This would be because an article was published in the British Medical Journal on April 2, touting the fact that a whopping 78% of people who tested positive for SARS-CoV-2 in China on April 1 were asymptomatic for the disease, COVID-19.
However, a small rummage through the sparse facts revealed that only 166 people tested positive within this 24 hour period, 130 of whom were asymptomatic. No other details were present.
The article does not describe how these people were chosen for testing, when they were likely infected, or how they determine someone is asymptomatic. Furthermore, this is data collected over 24 hours about a disease which takes an average of five days to show symptoms. Meaning, those 130 asymptomatic cases may not remain asymptomatic, they merely weren’t showing symptoms at the time of testing. We come back to right censoring again, except on a much shorter time scale than the Diamond Princess study.
Ultimately, this very small sample size and extremely limited time scale makes this data almost meaningless. Certainly not enough to base any kind of assumptions off, nor reliable enough to generalise as is suggested in the article itself. Any conclusions derived from this data are mere speculation.
So What Does This Mean?
This means, for now at least, we just don’t know the rate of asymptomatic cases across a population, nor do we have reliable enough data to extrapolate even an educated estimate.
As this pandemic continues to spread, countries across the world are struggling to keep up with testing targets. It could be that, before we can make any reliable suggestions on asymptomatic case rates, we will need to wait for the long anticipated antibody tests to be rolled out en masse.
Ultimately, much of this comes down to media sensationalism as they vie for those coveted clicks, as 50% or even 80% sounds much more exciting than just 18% or ‘we just don’t know’ in headlines.
As for the experts who have weighed in on these studies in the media, unfortunately even scientists aren’t immune to being mistaken, misled, or overly optimistic about such data. We are, after all, human.
It is natural that in these dark times people would search for the smallest of light. The real silver lining here is that the media and, more importantly, the public are listening to experts and scientists. And as I wrote in a previous article, it’s not about believing everything you hear or read, it’s about listening to and at least considering the science which is important, and this seems to be something people are doing more and more lately.
Further Reading
Day, M., 2020. Covid-19: four fifths of cases are asymptomatic, China figures indicate. BMJ, p.m1375.
Mizumoto, K., Kagaya, K., Zarebski, A. and Chowell, G., 2020. Estimating the asymptomatic proportion of coronavirus disease 2019 (COVID-19) cases on board the Diamond Princess cruise ship, Yokohama, Japan, 2020. Eurosurveillance, 25(10).
Sankey, S. (2020). What is an Antibody Test and How Does it Test for Coronavirus?. [online] Available at: https://medium.com/@stevensankey/what-is-an-antibody-test-and-how-does-it-test-for-coronavirus-32413cc1c5dc.
Tara John, C. (2020). Iceland lab’s testing suggests 50% of coronavirus cases have no symptoms. [online] CNN. Available at: https://edition.cnn.com/2020/04/01/europe/iceland-testing-coronavirus-intl/index.html.
SANKEY REPORTS 