Rumination on Vaccination Part II - statisticians view on the Pfizer COVID19 final analysis results

COVID-19 vaccine announcements are now coming at us thick and fast! Today Pfizer and BioNTech have announced the results of their final analysis on 170 COVID-19 cases, and this time they have provided us with more details. 

We know that of the 170 COVID-19 cases in total, 162 were observed in the placebo group versus 8 in the vaccine group, giving a vaccine effectiveness (VE) of 95.1%. It is also reported that the analysis was carried out on at least 43,000 participants, meaning that the 95% confidence interval (CI) for this VE figured is estimated to be around 90% to 98% and a 99% CI would be 88% to 98%. Similar to the Moderna study (results discussed here), success at the final analysis was defined as a posterior probability that the VE was greater than 30% being greater than 98.6% (P(VE>30%|data) > 0.986), and the associated efficacy boundary for success was 52.3%. We can see that today’s reported VE of 95.1% is very far away from 52.3%, meaning that the trial has overwhelming evidence of efficacy. 

Furthermore, we are told that the primary efficacy analysis is based on COVID-19 cases beginning 28 days after the first dose, meaning that COVID-19 cases occurring after just one dose of the vaccine would be included in analyses. This means that if there any participants who didn’t take their second dose of vaccine, but contracted COVID-19 28 days after their first vaccine dose, we should be able to ascertain if just one dose of vaccine provides any protection against COVID-19. This is in contrast to the Moderna study, which only counted COVID-19 cases that occurred 14 days after the second vaccine dose. 

Today’s announcement also tells us that there 10 severe cases of COVID-19 observed in the trial, with nine occurring in the placebo group and one in the BNT162b2 vaccine group. The Moderna study reported no severe COVID-19 cases in the vaccine group, but if severe COVID-19 cases had been the primary outcome, 9 versus 1 would still have given a VE of 89%. 

Today’s press release also provides more insight into subgroup analyses and safety results. We are told that vaccine efficacy was found to be consistent across age, gender, race, and ethnicity demographics, and that the observed efficacy in adults over the age of 65 was over 94%. It is also reported that the vaccine is generally well tolerated, with most adverse events resolving shortly after vaccination. We are also told that older adults tended to report fewer and milder solicited adverse events and that the only grade 3 (severe) solicited adverse events at least 2% in frequency were fatigue (3.8%) and headache (2.0%). 

It’s great to see that the interim results presented last week have followed through until the final analysis and Pfizer and BioNTech plan to submit a request within days to the FDA for an emergency use authorization.