Dark numbers block infection control mode

The dark numbers indicate the extent of undetected cases – the proportion of undetected cases out of the total number of cases. It is an important factor in understanding morbidity and mortality as well as planning and evaluating infection control measures. The more dark spots, the worse the control of transportation routes طرق [1, 2].

The dark numbers depend not least on test capacity and test criteria. The lower the testing capacity and the narrower the criteria, for example to test only with certain symptoms, the more unconfirmed cases. This is particularly problematic when a virus such as sars-cov-2 is also spread by asymptomatic individuals [3].

Current estimates of dark numbers in the USA, France, Great Britain and Italy range from 70 to 90 percent [4-7] It is based on antibody testing or modeling. Due to the limited sensitivity of different tests, the actual dark number is likely to be higher [6].

To do a good check on the dark numbers, the World Health Organization recommends that the proportion of positive responses to tests for active infection be less than 5 percent. [8]. At the beginning of January 2021, Sweden was at a record high of 20 percent, while our Nordic neighbors were below 5 percent. [9].

In order to map the prevalence in the population, the Swedish Public Health Agency has conducted a few population studies of active infection in 2020 [10]. The most recent was performed at Week 49 and showed a national prevalence of 0.7 (0.4-1.2 percent). As in previous studies, the result was not a problem, that is, whether it was reasonable in light of the reporting of confirmed cases.

During weeks 48-49, regions of Sweden reported a total of 72,406 new cases [11], which corresponds to just over 0.7 percent of the population, which is slightly higher than the estimated prevalence rate of 0.7 (0.4-1.2 percent). The result is strange. Population studies should record all cases, even those with no or few symptoms. District testing prioritizes cases with obvious symptoms.

Even in international comparison, the numbers seem surprising. During the same period, the prevalence of active infection in the UK was 0.9 per cent [8]. Meanwhile, the country had about half as many confirmed cases per capita as Sweden, despite repeated testing.

The December study isn’t the only one that raises questions. In the Swedish Public Health Agency studies in August and September, no participants showed a positive test result [10]. This was statistically unreasonable given the number of confirmed cases over the same period.

I think there is a bias in population studies. They are based on self-sampling and require a certain commitment. Perhaps the interest in participation is greater among people who are at higher health risks, and therefore have lower risks and are less affected. Conversely, interest may be lower among those who have been sick or ill, and who know they have had or have contracted the coronavirus.

I only know of a Swedish covid-19 population study that holds this measure. Made in Norrbotten and referred to antibodies over 22 – 23 [12]. With an estimated seroprevalence rate of 1.9 percent (416 and 503 confirmed cases through week 22 and 23, respectively), the darker figure would be at least 85-90 percent, in line with international studies. Otherwise, the studies in Sweden could shed light on the minuscule dark numbers. This should be evident during a pandemic. Otherwise, how should we be able to assess the risks and impacts of measures?

A study on the effects of the sars-cov-2 vaccine on herd immunity was recently published in Luxembourg, Austria and Sweden. [13]. Sweden has high dark numbers but it is uncertain. This complicates the analyzes and reflects pictures of an uncertain case of the epidemic in Sweden.

Lakartedningen 4-5 / 2021

Lakartidningen.se 2021-01-27
(Updated on 2021-01-28)