So, as predicted, the headline vaccine efficacy figures reported are somewhat misleading and the confidence intervals overlap massively.
At 4 weeks after one dose, efficacy reported.
Pfizer 76-91%, spot value 85%
AZ 73-99%, spot value 94%
Note that the AZ figure is based on just two(!) hospitalisations, the Pfizer on 18. The AZ confidence interval seems remarkably tight given it's based on only two patients, but OTOH I'm very sure they know their stats better than I do. The three week AZ figure is 72-90%, spot value 84%, based on 18 hospitalisations, which seems more soundly based.
Bottom line here is that both vaccines look very good at four weeks with a single dose, and can't be differentiated by this dataset.
Poor reporting to give the spot values, based on a poor press release which did the same, in my opinion.
They torture the data with myriad adjustments including age, sex, prexisting conditions, deprivation index and something to do with prior test results, none of which I really understand. If you just look at the base rate of hospitalisations with and without vaccine, they are far less impressive sounding, presumably because the most vulnerable are most likely to have been vaccinated.
The Pfizer one seems to drop off after 4 weeks, I've seen in claimed that may be an artefact of the low sample size.