If I understand the video correctly, efficacy in the Oxford trial was determined based on the number of positive antibody tests, whereas efficacy in the Pfizer trial was determined based on symptoms.
If one assumes that some people without symptoms in the Pfizer trial would have tested positive, and that none of the people with symptoms in the Pfizer trial would have tested negative, it seems as if the efficacy of the Oxford vaccine might be significantly greater (i.e. closer to the Pfizer vaccine) if the same efficacy standard were to applied to both.
Basically, it's unclear to me how the efficacy of the two vaccines would compare if the same criteria were to be applied to both.
>>>The Pfizer/BioNTech vaccine reports 90% efficacy, which means that – of the 94 confirmed cases of COVID-19 – their vaccine prevented COVID-19 symptoms for 90% of those who received the vaccine compared with placebo.<<<
In science, vaccine efficacy and effectiveness mean something different.
theconversation.com
>>>The Oxford–AstraZeneca team is the only one of the three leading vaccine developers that monitored for asymptomatic infections, by collecting weekly swabs from some participants to determine whether they had the coronavirus but did not become ill. The data show that the low-dose vaccine regimen was about 60% effective at reducing asymptomatic infections, but it is unclear whether the standard dose significantly reduced them at all. . . .
Trials of the two leading RNA vaccines have not gathered data on asymptomatic infections, but the vaccines have been more than 90% effective in preventing symptoms of COVID-19.<<<
The Oxford–AstraZeneca partnership is the first major developer to publish detailed data from phase III trials.
www.nature.com
Good point. I've been struggling with press reports that the Pfizer vaccine is 95% effective while no definition was provided about how that efficacy number was determined. That's the main reason why I read that FDA report on the Pfizer clinical trial. Pfizer did evaluate trial subjects regularly for disease. See the FDA report on the Pfizer trial,
https://www.fda.gov/media/144245/download.
Page 13:
5.2.3.3. Vaccine Efficacy Analyses
Study C4591001 is the pivotal (and only) efficacy study. Efficacy was assessed based on confirmed cases of COVID-19, for which the case onset date was the date that symptoms were first experienced by the participant and the cases met evaluable criteria. For participants with multiple confirmed cases, only the first case contributed to the VE calculations. Evaluable cases consisted of a positive virological test plus at least one COVID-19 symptom as defined below.
Only the first primary endpoint was analyzed in the planned interim analysis. All primary and secondary efficacy endpoints were planned to be analyzed in the final analysis of at least 164 cases.
Primary Efficacy Endpoints
Study C4591001 has two primary endpoints:
- First primary endpoint: COVID-19 incidence per 1000 person-years of follow-up in participants without serological or virological evidence of past SARS-CoV-2 infection before and during vaccination regimen – cases confirmed ≥7 days after Dose 2
- Second primary endpoint: COVID-19 incidence per 1000 person-years of follow-up in participants with and without evidence of past SARS-CoV-2 infection before and during vaccination regimen – cases confirmed ≥7 days after Dose 2.
Page 14
For the primary efficacy endpoint, the case definition for a confirmed COVID-19 case was the presence of at least one of the following symptoms and a positive SARS-CoV-2 NAAT within 4 days of the symptomatic period:
• Fever;
• New or increased cough;
• New or increased shortness of breath;
• Chills;
• New or increased muscle pain;
• New loss of taste or smell;
• Sore throat;
• Diarrhea;
• Vomiting.
For a secondary efficacy endpoint, a second definition, which may be updated as more is learned about COVID-19, included the following additional symptoms defined by CDC (listed at
https://www.cdc.gov/coronavirus/2019-ncov/symptoms-testing/symptoms.html):
• Fatigue;
• Headache;
• Nasal congestion or runny nose;
• Nausea.
For another secondary endpoint, the case definition for a severe COVID-19 case was a confirmed COVID-19 case with at least one of the following:
• Clinical signs at rest indicative of severe systemic illness (RR ≥30 breaths per minute, HR ≥125 beats per minute, SpO2 ≤93% on room air at sea level, or PaO2/FiO2 <300 mm Hg);
• Respiratory failure (defined as needing high-flow oxygen, noninvasive ventilation, mechanical ventilation, or ECMO);
• Evidence of shock (SBP <90 mm Hg, DBP <60 mm Hg, or requiring vasopressors)
• Significant acute renal, hepatic, or neurologic dysfunction;
• Admission to an ICU;
• Death.
The actual method used to calculate the percent efficacy is not directly presented, and would seem obscure to someone not acquainted with clinical trials. It involves statistical analysis of survival, often called Kaplan-Meier analysis. This is far from a straightforward division of the number of people in the vaccine arm who did not get Covid-19, divided by the total number of people in that arm.
If you are not familiar with Kaplan-Meier analysis and survival curves, and you love statistics, or can't sleep at night, check these links out:
An introduction
towardsdatascience.com
In 1958, Edward L. Kaplan and Paul Meier collaborated to publish a seminal paper on how to deal with incomplete observations. Subsequently, the Kaplan-Meier curves and estimates of survival data have become a familiar way of dealing with differing survival ...
www.ncbi.nlm.nih.gov
Kaplan-Meier estimate is one of the best options to be used to measure the fraction of subjects living for a certain amount of time after treatment. In clinical trials or community trials, the effect of an intervention is assessed by measuring the number ...
www.ncbi.nlm.nih.gov
en.wikipedia.org
Most survival curves show each death of a patient on the trial, as a single event, or step, on the curve.The graph I posted (
post #3966 above) of the two arms of the Pfizer vaccine trial is basically an inverted survival curve, where each event is someone getting sick with Covid-19. It climbs up, instead of dropping like a survival curve. That graph clearly shows the difference between the vaccine and the placebo groups, and you don't need statistical analysis for that. But you can also crunch the numbers to come up with percent efficacy estimates for the vaccine at a given time point. The example above shows a time frame of 2 months. This trial may eventually get efficacy numbers for as long as 2 years.
This is a long complex answer, but I hope it helps. If I understood statistics better, I could give a much shorter answer.
