This is true, but based on very small numbers, so caution required.
In the Oxford/AZ trial on the placebo :
- ten were hospitalised
- two of those suffered "severe" disease
- none died
On the active, there were zero of any of these.
So in terms of comparing death, there's no direct evidence whatever that deaths are reduced - there were zero deaths on placebo and zero on active. We can infer that deaths are likely to reduce significantly, but it's quite impossible to quantify.
And to say that there is 100% reduction in hospitalisation based on just ten cases is rather optimistic. The margin of error on such a small sample is very high.
None of which is to say there's any question mark over the efficacy being excellent.