Covid-19 may be far more prevalent than previously thought

ONE OF THE trickier tasks for epidemiologists studying the coronavirus pandemic is determining just how prevalent it is. According to Johns Hopkins University, there are now 886,000 confirmed cases of covid-19 around the world, or roughly one case for every 9,000 people. But these numbers reflect not only how many people are known to have contracted the illness, but how many are being tested for it. This varies widely from country to country. Germany is testing 500,000 of its citizens each week. Britain, which is limiting tests to those with severe symptoms, is managing just one-tenth as many.

A new paper by a group of 62 scientists attempts to shed light on the matter using some clever statistics. The study, which was overseen by Neil Ferguson and Samir Bhatt of Imperial College London, looked at data for 11 countries in western Europe. To arrive at accurate estimates for the likely prevalence of the virus in each country, the researchers’ epidemiological model calculates backwards using data on observed covid-19 deaths. By observing the timing of deaths alongside known information about transmission rates and time lags of infections, the researchers arrive at better estimates of covid-19 prevalence.

The results suggest the virus is far more prevalent than official case counts would suggest. Across the 11 European countries studied, official government statistics indicate that there are 319,500 confirmed cases, affecting less than 0.1% of the population. The Imperial researchers peg the more likely figure at 18.5m cases (with a confidence interval of between 7m and 43m), which suggests that, as of March 28th, 5% of the population in these countries have contracted the virus.

This high infection rate is partly because of the highly contagious nature of the virus. The epidemiologists estimated the “basic reproduction number” of covid-19 to be 3.9, meaning that in a population where no one is immune, and no precautions are taken to control the outbreak, each infected individual passes on the virus to nearly four other people. By implementing “non-pharmaceutical interventions”—such as school closures, the banning of public events and nationwide lockdowns—governments have successfully reduced this reproduction number. Once it drops below one, the number of new infections starts to fall.

The researchers estimate that timely interventions over the past month have reduced the reproduction number of the virus across the 11 countries studied by two-thirds to 1.4, thus saving between 21,000 and 120,000 lives. In Italy, alone, interventions have prevented some 38,000 deaths, according to the researchers. The past few weeks have been tough for those hit by the crisis. They can take some solace in knowing that it could have been much worse.