Amid the uniquely unsettling novelty the coronavirus epidemic has brought, one thing is more uncertain than any other: how many of us have, or have had, the virus?
On Tuesday, for example, Italy's civil protection chief, Dr Angelo Borrelli, briefed reporters with the latest statistics: cases up 3,612 to 54,030. The problem was that, earlier in the day, he had suggested there might be as many as 600,000. Asked to account for the difference, he said his earlier answer had taken into account asymptomatic cases. So what is it? 54,000 or 600,000?
Italy is not alone. Earlier this week, the Self-Defense Forces Central Hospital in Tokyo published a report into the Diamond Princess cruise ship, which was quarantined in February with more than 800 of the 3,711 people on board eventually diagnosed with Covid-19.
According to the report, cited by the Japan Times, four-fifths of those infected showed no symptoms or just mild symptoms. Scans showed their lungs had suffered some physical consequences from the virus, but the patients had not been led by their symptoms to believe they were infected. Quite the reverse.
Even in China, the poster child for lockdown and trace tactics which is edging back towards business as usual, there is uncertainty about asymptomatic cases. Estimates are classified. Officials insist all asymptomatic cases are unearthed by so-called "contact tracing", yet even if this is true and every such case is detected, the numbers are not trivial: the South China Morning Post obtained details which suggested some 40,000 asymptomatic cases had been discovered in this way. It's hard not to believe there are many more.
One way or another, it is clear that far more of us have, or have had, Covid-19, than the official figures tell us. But how many more? Only mass testing will give us a real picture of what is happening, and thus needs to happen at a far greater rate in the UK than is currently the case:
Without mass testing, two very different pictures of the spread of the virus are emerging.
On the one hand, there is the optimistic scenario, suggested by an Oxford paper which on Wednesday gained a huge amount of attention for a model suggesting that, given the current state of the UK outbreak, some 68 per cent of the population "would have been infected by 19/03/2020" – some 45 million of us, by last week, most presumably without even knowing it.
Such a result would be dramatic. The outbreak is overblown, it would confirm. Most of us have already had the disease and are now fine. The NHS has survived the epidemic without even knowing it.
But that was just one paper. There is another scenario. This one is that yes, there are lots of asymptomatic carriers (though far from 45 million), but rather than being beacons of hope and joy, they represent hidden problems – mines sown among the rest of us, ready to explode and cause a new outbreak just when it appears we might be on top of this one.
According to researchers at the Centre for Mathematical Modelling of Infectious Diseases at the London School of Hygiene and Tropical Medicine (in a paper awaiting peer review), children may represent such "mines" because only around 20 per cent of them develop clinical symptoms (as opposed to 70 per cent in older adults). Yet it is not yet clear how infectious such largely asymptomatic people are. This critical question "must be resolved to effectively forecast and control Covid-19 epidemics", the authors note.
Such infection is certainly possible. In Wuhan on Tuesday, the one new case was that of a doctor reportedly infected by an asymptomatic patient. And on the Diamond Princess, studies suggest some of the 20 per cent of infected passengers who suffered serious symptoms either went several days without obvious outward signs of the condition or caught it from those who never went on to develop serious symptoms.
So is the epidemic mostly over or every bit as bad as we think, with the potential to erupt when we imagine we're through the worst?
A single truth emerges: we will only discover the genuine picture through mass blood testing – a simple finger-prick tests that works like this:
The testing must be followed, once the outbreak (whatever its true size) is controlled, by picking representative samples in the population to screen regularly, just as such samples are quizzed by pollsters to get a more or less accurate snapshot of national voting intention.
In the middle of a mass global outbreak, it seems impossible to carry out whole population testing. Yet fate has twice thrown up unusual circumstances in which this has happened. The Diamond Princess was one such. The other was the 3,000-strong town of Vo, in Italy, where the entire population was tested after one of its citizens became the first to die of the disease.
There, too, researchers found "a significant proportion of the population" (three per cent) had been infected without knowing it, representing the vast bulk of the 89 cases discovered. More generally, in the researchers' estimation, "asymptomatic or quasi-symptomatic subjects represent a good 70 per cent of all virus-infected people".
There is good news in this: mass testing in Vo meant there were no further deaths. And factoring in asymptomatic cases also dramatically lowers the disease's mortality rate, which is currently based only on those who receive tests – usually those showing obvious or severe symptoms.
And finally, if you had a fever or a hacking cough earlier in the year and, looking back, now swear you must have had Covid-19 – well, you might just be right. For you, Tommy, the war might truly be over. Best stay inside anyway, though, eh?