A new paper published by researchers at the University of Göttingen suggests that the official numbers of COVID-19 cases is underestimating the reality in the field — dramatically so.

According to the team’s model, only 6% of all infections with the SARS-CoV-2 virus have been detected worldwide, placing the real number of infections, potentially, in the tens of millions. The authors say that their findings should serve as a warning against relying too heavily on the reported number of cases for policymakers.

While definitely worrying, the results are based largely on reports from cases in the Wuhan province, China. As there are growing concerns that the country has misreported data pertaining to the COVID-19 epidemic to the wider world to appease its own political machinations, having “concealed the extent of the coronavirus outbreak [by] under-reporting both total cases and deaths it’s suffered from the disease,” the findings are best taken with a grain of salt. While the exact figures reported on in this paper may suffer due to the unreliability of data, the larger general trends identified in this paper may still be sound.

A drop in a bucket

“These results mean that governments and policy-makers need to exercise extreme caution when interpreting case numbers for planning purposes,” Sebastian Vollmer, Professor of Development Economics at the University of Göttingen and co-author of the report, explained in an article for the University.

“Such extreme differences in the amount and quality of testing carried out in different countries mean that official case records are largely uninformative and do not provide helpful information,” adds Dr. Christian Bommer, the report’s second author. “Major improvements in the ability of countries to detect new infections and contain the virus are urgently needed.”

The duo drew data from a recent study published in The Lancet Infectious Diseases journal which estimated the mortality rate of COVID-19 and the time until death — i.e. the time between contracting the virus and a patient’s death. Based on these figures, they developed a mathematical model to help them estimate the quality of official case records, giving them an idea of how many cases are likely detected out of the total number spreading through society.

All in all, official numbers “dramatically understate” the true number of infections, the two report.

The team believes that some European countries such as Spain and Italy are seeing much higher casualty rates from the virus than others, for example Germany, because they have only detected a smaller number of their overall infections — and this lack of data artificially makes the virus seem more deadly here. According to their estimates, Germany has detected around 15.6% of infections compared to only 3.5% in Italy and 1.7% in Spain. the United States and the United Kingdom — two countries that have received widespread criticism from public health experts for their delayed response to the pandemic — are looking at ever lower detection rates, of 1.6% and 1.2% respectively. On the other end of the spectrum are countries such as South Korea, which appears to have detected about half of all its SARS-CoV-2 infections.

If these results are true, it would mean that there are currently in excess of ten million infected in the United States, over five million in Spain, around three million in Italy, and around two million in the UK. In Germany, the team estimated that the number of infections is close to 460,000. On the same day this report was published (31st of March), there was a total of about only 900,000 confirmed cases worldwide.

How accurate the team’s estimates are hinges on how accurate the official data made available by Chinese authorities is, in turn. Regardless of this, the report is a good reminder that the official figures currently at our disposal aren’t the reality on the ground, they’re just the best attempt we currently have at gauging it. In the absence of mass testing, it’s simply impossible to know how many are truly infected. Until such measures become possible, the report aims to caution policymakers and healthcare experts on the limits of the data they work with.

The report “Average detection rate of SARS-CoV-2 infections is estimated around six percent” is available on the University of Göttingen’s page.