@gernot_ruzicka @beftwi @t_kurz @CoronaZeroCH @UZH_ch @uzh_EBPI @AgneUlyte 1/ The authors of medrxiv.org/content/10.110…… should either rule out any such bias or precise it and take it in their analysis/conclusions as best as possible into account. It is therefore essential they inform precisely about:
2/ a) When (exact date) sample collection occurred for each participant.
b) How long- and based on which validation data they consider the latency from onset of symptoms until the test gets reliably positive to be. virology.uzh.ch/dam/jcr:0f9bbd… says nothing about that.
b) How long- and based on which validation data they consider the latency from onset of symptoms until the test gets reliably positive to be. virology.uzh.ch/dam/jcr:0f9bbd… says nothing about that.
3/ Furthermore the authors should in a clearly comprehensible way explain based on which exact data/calculation and/or references in what extent they think infections might have been missed when the serologic analysis was negative (= false negatives).
4/ The publication gives a value of 11 for 2330 Analyses. It is well known how such a value ist calculated technically. Nevertheless it would ease verification if the authors precised the exact formula how they have calculated it.
5/ The publicly available document virology.uzh.ch/dam/jcr:0f9bbd… precises that sensitivity of the test has been determined on 104 plasma samples, which is relatively few, and it is not specified whether or how many of those samples were samples from children.
6/ If the test has not been validated on children or only on very few children, this adds up to the uncertainty concerning the sensitivity of the test and false negatives.
https://mobile.twitter.com/petertomasi1/status/1342688277334290432
@threadreaderapp unroll
• • •
Missing some Tweet in this thread? You can try to
force a refresh
