А это его исследование. По-моему, тут все сказано про бессимптомную детскую выборку...

From January to 26 April, 2020, virology laboratories at Charité and Labor Berlin screened

59,831 patients for COVID-19 infection, 3,712 (6.2%) with a positive real-time RT-PCR result. We divided patients according to two categorizations to investigate whether there is a relationship between patient age and viral load. The first categorization is based on ten-year brackets, ages 1-10, 11-20, 21-30, 31-40, 41-50, 51-60, 61-70, 71-80, 81-90, and 91-100. The second categorization is based on broad social strata: kindergarten (ages 0-6), grade school (ages 7-11), high school (ages 12-19), university (ages 20-25), adult (26-45 years), and mature (age over 45). Patient counts in each age group, and number and percentage of PCR positive patients are shown in Table 1. A comparison of age stratification in tested cases versus the Berlin population is shown in Figure A1. Of note, whereas younger age groups have lower

detection rates (Table 1), this does not imply an age-based estimate of infection prevalence because of mostly symptoms-directed testing.

Due to the small sample sizes in the pediatric age groups, we examined diagnostic indications for 47 cases (1-11 years of age) for whom this information was available. Fifteen cases had indications pointing toward underlying disease or hospitalization. Average viral loads in these cases were lower than in children tested in outpatient departments, practices, or households (Figure A2). This corresponds to the observation that hospitalization occurs after some days of symptoms, a time when viral loads in throat swabs are beginning to decline (5).