Number of people with coronavirus infections may be dozens of times higher than the number of confirmed cases

Authored by thl.fi and submitted by matakos18
image for Number of people with coronavirus infections may be dozens of times higher than the number of confirmed cases

An antibody study carried out by the Finnish Institute for Health and Welfare (THL) has found that the number of people infected with coronavirus in the Hospital District of Helsinki and Uusimaa (HUS) may be many times higher than the confirmed number of cases.

Antibody studies can be used to collect information on the share of the population that have been infected with coronavirus. THL studied the prevalence of antibodies in blood samples of persons who had undergone laboratory tests for various reasons in the Hospital District of Helsinki and Uusimaa.

"The study is the first preliminary assessment of the extent to which the coronavirus epidemic has spread in the HUS area, and the share of people who may have had the mild or asymptomatic version of the disease," says Merit Melin, Senior Researcher at the Finnish Institute for Health and Welfare.

The sampling includes just over 400 samples

The samples were collected from HUSLAB’s clinical chemistry laboratories over a period of three consecutive weeks 23 March – 12 April 2020. The sampling included a total of 442 samples from men and women between the ages of 15 and 90, who had given a blood sample for reasons other than anyinfection. These surplus samples from laboratory tests were examined without personal identifiers.

Two different tests were performed to determine the presence of antibodies. A commercial rapid test was used for sample screening, the performance of which was tested at THL before the study. In addition, all positive results were verified with THL’s own test. The THL confirmation test measures the ability of antibodies to neutralise the virus and provides a very reliable result.

Samples take later contain more antibodies

The sampling taken during week 13, included 145 samples of which one sample (0.7 percent) was observed to contain antibodies. No antibodies were found from the 150 samples in week 14 sampling. The sampling taken during week 15, included 147 samples of which 5 samples (3.4 percent) was observed to contain antibodies.

Antibodies form on average over a period of two weeks meaning that the results reflected the presence of infections two weeks prior to sampling. Due to the small number of samples and findings examined, the results must be interpreted with some caution.

"The material used in the study does not represent the population as well as a random sampling of the population, so the result is still very preliminary at this stage. However, it is in line with the results of a previous antibody study conducted on blood donors in Denmark,” says Merit Melin.

More information on the presence of antibodies in the population and more detailed estimates on the share of people infected with coronavirus will be provided by the random sampling-based population study that THL initiated last week. The study examines the presence of antibodies in different age groups and in Finland’s different regions.

Finnish Institute for Health and Welfare

KawarthaDairyLover on April 16th, 2020 at 11:58 UTC »

Very curious to see the random sampling result here. Any word on when we might see the Stanford serological study?

raddaya on April 16th, 2020 at 11:19 UTC »

OP, you may want to flair this as Press Release, just btw.

So this is from the Finnish National Institute for Health and Welfare's antibody tests collected from random blood samples. The relevant part:

The sampling taken during week 13, included 145 samples of which one sample (0.7 percent) was observed to contain antibodies. No antibodies were found from the 150 samples in week 14 sampling. The sampling taken during week 15, included 147 samples of which 5 samples (3.4 percent) was observed to contain antibodies.

Antibodies form on average over a period of two weeks meaning that the results reflected the presence of infections two weeks prior to sampling. Due to the small number of samples and findings examined, the results must be interpreted with some caution.

As the article mentions, this is nearly identical with the Danish antibody test data - it seems to me that the antibody tests are all pointing towards the "iceberg" of unconfirmed cases existing and that it is at least fairly large.

matakos18 on April 16th, 2020 at 11:15 UTC »

The sampling taken during week 13, included 145 samples of which one sample (0.7 percent) was observed to contain antibodies. No antibodies were found from the 150 samples in week 14 sampling. The sampling taken during week 15, included 147 samples of which 5 samples (3.4 percent) was observed to contain antibodies. 

Antibodies form on average over a period of two weeks meaning that the results reflected the presence of infections two weeks prior to sampling. Due to the small number of samples and findings examined, the results must be interpreted with some caution.

"The material used in the study does not represent the population as well as a random sampling of the population, so the result is still very preliminary at this stage. However, it is in line with the results of a previous antibody study conducted on blood donors in Denmark,” says Merit Melin. 

More information on the presence of antibodies in the population and more detailed estimates on the share of people infected with coronavirus will be provided by the random sampling-based population study that THL initiated last week. The study examines the presence of antibodies in different age groups and in Finland’s different regions.