The recent Novavax vaccine trial reported lower efficacy in its SA arm. This is plausibly caused, at least in part, by the high frequency of the B 1.351 lineage in SA, which is expected to partly bypass prior immunisation from #SARSCoV2 infection. https://www.sciencemag.org/news/2021/01/novavax-vaccine-delivers-89-efficacy-against-covid-19-uk-less-potent-south-africa

There are many other differences between the two arms of the trial that could have contributed to the difference in efficacy. Of note, there is a trend for HIV+ patients having been less effectively protected, efficacy in SA being ~60% (rather than ~50% with HIV+ included).
2/

We may wish to remember that the Oxford-AZ vaccine trial reported similarly high differences in efficacy between their Brazilian and UK arm, which were largely 'explained away' by a moderately convincing 'dosage error story'.
3/

An important point is that the numbers of reinfections in those vaccine trials are small and thus come with considerable uncertainty. Such values should NOT be interpreted in terms of hard 'point estimates' (i.e. 90.3% vs. 60.1%).
4/

The Novavax trial results actually represent a success in terms of vaccine effectiveness, with ~50% when including HIV+ people in SA (~60% when HIV+ excluded). It may be good to reflect on these number in the light of some pessimistic predictions from a few months ago. 5/

Even if we assumed viral mutations explain all the pattern, the results are not a disaster. This is a vaccine that targets a viral lineage which is >1 year old, and vaccine part-escape - and in the long term full-escape mutants - were to be expected.
6/ https://twitter.com/BallouxFrancois/status/1331042684291739648

It may also be helpful to consider the results of #COVID19 vaccine trials in the wider context of vaccine efficacy against other respiratory viruses. They look pretty amazing when compared to seasonal flu vaccine efficacy ... 7/
Source: https://www.cdc.gov/flu/vaccines-work/effectiveness-studies.htm

There is evidence, interpretation thereof and wild speculations. The three types of discourse should not be conflated, in particular when the topic is emotive and the audience does not have the scientific background to easily distinguish between the three.
8/

Note: This thread replaces an earlier one, as the presentation of the effect of the inclusion/exclusion of HIV+ people in the SA arm of the trial was confusing. I don't like deleting tweets, but chose to do so as there is already enough confusing material in circulation ...
9/