Just ran the numbers for 35. I've also added a column with p Value calculated using Z-test for comparison (I was not able to add the Chi test yet so happy if someone want's to post that). It seems that if we are around 50% reduction in deaths then we meet statistical significance even with a lower number of deaths.