Of course. Every scientist wants to publish a study that definitely shows an effect, or definitely rules out an effect. The data here--the relatively small number of positives for the youngest groups, plus the lack of a normal distribution--make that very difficult. But I think Drosten should have emphasized this, rather than implying that his study clearly indicated that children have viral loads as high as older people. You can say, the study provides no evidence that children have lower loads, rather than, we should assume that children are just as infectious as adults.
As I noted earlier, the analysis using the second classification, actually found the highest mean viral load for the grade school students. Drosten never points this out, though for anyone looking at the means, it sticks out like a sore thumb. Why is that one group so high? There were only 16 subjects, 15 - more than 100 times fewer than for the older age groups. This is the kind of outlier you can get when you work with small sample sizes.
You could lump the first two groups together. That would provide a larger sample size and a smaller standard error. Just doing a rough calculation, I find a mean of 4.74. a SE of 0.16, and 95% confidence limits of 4.42 - 5.06. The upper bound of that is lower than the lower bound of all the other groups except 5, 8 and 10, with 10 being ruled out because of small sample size. And if you lumped all eight of the older groups together (just do 7, the oldest group is way too small, and the numbers it does provide indicate it's about the same as the other seven, anyway), you get a mean of 5.26, a SE of around 0.03, and 95% limits of 5.20-5.32.
Statisticians would probably say that's an arbitrary comparison, but 20 is about the age when humans reach physical and mental maturity. i think it's entirely reasonable to compare people below that age with people above. The ten year groups Drosten uses are arbitrary, too. Suppose we use five 20 year groups instead?
1-20: 4.42 - 5.06
21-40: 5.12 - 5.34
41-60: 5.02 - 5.24
61-80: 5.10 - 5.34
81-100: 5.12 - 5.52
Complete separation between 1-20 and all the other 20 year groups except 41-60, which is really close,.
Yes, that's an important point. But probably there are multiple factors involved. Just a high viral load could be due to several. factors, including time in relation to symptoms, and infectivity could be due to other factors than high viral load. We really don't know if certain individuals are super-spreaders because of intrinsic characteristics, or if it was just a chance event.
Except that every major country that has estimated prevalence of infection has come up with a far higher estimate (than the 0.26% cited by CDC):
Spain: 1.1%
France: 0.70%
Netherlands: 0.67%
Belgium: 0.65%
Sweden: 0.47%
Denmark: 0.38%
Wrt the flu, in 2017-18, one of the worst recent seasons, the U.S. mortality rates were:
18-49: 0.02%
50-65: 0.05%
65+: 0.86%
The under 50 rates of the flu and C19 do seem to be roughly comparable, though different studies do vary widely. And of course, C19 is more contagious, so mortality rates per overall population are higher. Over 50, though, there is a huge gap. The studies in the Netherlands, France and Spain indicate rates of 0.3-0.5% for 50-70. And keep in mind that we still don't know what the long-term consequences may be for many survivors of C19.
Facebook
Twitter
Instagram