Research published recently by the University of Bern estimates around 20% of Covid-19 cases are asymptomatic.
When the researchers analysed 79 studies done from March to June 2020 and adjusted them for various factors, chief among them removing pre-symptomatic cases – those with no symptoms that went on to develop them – from those classified as asymptomatic, they found the rate of asymptomatic patients was 20%.
Similar analysis of seven studies run on defined populations yielded a rate of asymptomatic patients of 31%.
Further analysis by age suggests children are more likely to be asymptomatic. 27% of hospitalised children and 11% of hospitalised adults were asymptomatic.
The results of the study also suggest asymptomatic and pre-symptomatic cases are less likely to spread the virus. The spread risk for asymptomatic and pre-symptomatic cases appeared to be 35% and 63% of that of symptomatic cases.
The authors only included studies that provided information about follow-up through the course of infection, which allowed reliable assessment about the proportion of asymptomatic people in different settings. However, they warned that bias in the selection of test participants in the underlying studies could result in under- or over-estimation of asymptomatic cases, something likely to be so in studies focused on healthcare workers and blood donors. In addition, the impact of false negative RT-PCR tests was not included – RT-PCR tests are highly specific and almost never give a false positive.
A separate study of data from the cruise ship Diamond Princess, which adjusted out pre-symptomatic cases, yielded an asymptomatic rate of nearly 18% among those testing positive. 3,063 tests covering most of those on board found 634 people to be positive. Before leaving the ship 48% displayed no symptoms. However, based on what is known about symptom onset, modellers statistically adjusted for pre-symptomatic cases, which yielded an asymptomatic rate of close to 18%.
The four databases that the researchers at the University of Bern searched are not comprehensive, but they cover the majority of publications and they do not believe that they have missed any studies that would change their conclusions.
Study published on PLOS (in English)