In real numbers, the trial looked at 7,651 people in 90 clusters of people who were exposed to someone with Ebola. It divided them into two randomly assigned groups, by cluster. The experimental group was offered the vaccine immediately, and the control group was offered the vaccine after a 3 week delay. The delayed group was the experimental control – it wasn't medically ethical to give no vaccine or give a placebo.
In the 48 immediate vaccination clusters, 3035 (74%) of 4123 contacts and contacts of contacts were eligible for vaccination and 2014 (49%) were immediately vaccinated. In the 42 delayed vaccination clusters 2380 (67%) of 3528 contacts and contacts of contacts were eligible for vaccination and 1498 (42%) were vaccinated after a 3-week delay.
A total of 75 laboratory-confirmed cases of Ebola virus disease were identified in the 90 randomized clusters during the study period, of whom 33 patients died. The case-fatality rate was 52% (15/29) in the immediate vaccination rings and 39% (18/46) in the delayed vaccination rings. This included people who were eligible but refused the vaccine.
In each group, after a 10-day wait for the vaccine to take effect, they counted subsequent Ebola cases: zero in the immediate group, and 16 in the delayed group.
With this kind of trial design, it isn't possible to know for sure just how many vaccinated people were really exposed to being infected. Even if it was possible to know that, these numbers (in the thousands) are large enough only to predict (not directly measure) what might happen for large populations.
So, bring on the statisticians… Although a direct reading of the test results showed a vaccine efficacy of 100%, in statistical terms, at a 95% Confidence Interval (CI), this becomes 74.7% to 100.0%; with a
p=0.0036). A
p value lower than 0.05 is good, and here
p=0.0036 means these numbers are highly reliable at predicting what might happen to a larger population.
Other good news from these very early results suggests the vaccine is safe, at least for the short run, up to 84 days after vaccination. 43 serious adverse events were reported; one serious adverse event was judged to be caused by vaccination, a febrile (fever) episode in a vaccinated participant, which resolved without any further complication.
That article you linked from The Guardian was very good. It did reflect the statistical interpretation where it said:
"In terms of vaccines, which are usually trialled in hundreds of thousands of people, Kierny said the numbers were small but highly promising. It is likely when larger numbers are collected that efficacy will be between 75% and 100%."
