Skip to main content

Safety and efficacy of the ChAdOx1 nCoV-19 vaccine (AZD1222) against SARS-CoV-2: an interim analysis of four randomised controlled trials in Brazil, South Africa, and the UK

Our take —

This combined interim analysis of four large-scale blinded randomized controlled trials including 11,636 participants shows that an adenovirus-based SARS-CoV-2 vaccine delivered in a prime-boost strategy is extremely safe, and can provide moderate (63% in full dose prime) to excellent (90% in low dose prime) protection against symptomatic COVID-19 disease, though the fact that the low dose prime strategy resulted from a vaccine production error and only involved a subgroup of participants in a single site complicates interpretation. Protection against asymptomatic infection was only observed in the low dose prime group, suggesting this strategy may be the most beneficial moving forward and warrants further investigation.

Study design

Randomized Controlled Trial

Study population and setting

This is an interim report of results from four vaccine trials that examined the safety and efficacy of the Oxford/Astra Zeneca SARS-CoV-2 vaccine. The vaccine is made up of a replication deficient adenovirus-based backbone that expresses SARS-CoV-2 Spike protein, and has shown good safety and immunogenicity in previous phase 1 and 2 studies. This paper describes the combined efficacy results of two dose strategies of the vaccine (two full doses of 5 x 105 vaccine particles, or one low dose followed by a second full dose; in both strategies doses are given at least 4 weeks apart) versus a placebo control (meningococcal vaccine or saline) in a 1:1 fashion from two of the three sites included in the larger phase 3 study, United Kingdom and Brazil. The low dose followed by a high dose strategy was not originally planned, but later adopted for one group of participants in the UK due to an error in vaccine production; this also led to variation in timing of the second dose in that group. The primary endpoint of the trial was virologically confirmed, symptomatic COVID-19 among patients with at least one of the following symptoms: measured fever, cough, shortness of breath, anosmia or ageusia. In addition, UK participants were asked to self-swab weekly to evaluate differences between groups in asymptomatic shedding. Safety data from these trials, as well as earlier phase 1 and phase 2 studies in the United Kingdom and South Africa were also analyzed. 11,636 people aged 18+ years were included in the interim efficacy analysis and were analyzed according to the treatment they received.

Summary of Main Findings

The participants that received both full doses of vaccine were 62% less likely to experience symptomatic COVID-19 infection at least 14 days after the second dose of the vaccine versus the placebo arm (27 of 4400 [0.6%] vs 71 of 4455 [1.6%] respectively). In this group receiving two full doses, vaccine efficacy was highly similar in both the UK and Brazil participants, even though the populations were different, with the UK group being slightly older and less racially diverse. In a smaller group, participants in the UK who received a low-dose of the vaccine followed by a full dose were 90% less likely to experience symptomatic RNA-confirmed COVID-19 (3 of 1367 [0.2%] vs 30 of 1374 [1.7%]). Among 126,324 identifiable swabs from UK participants, there was no observable protection from asymptomatic infection for the vaccinated participants who received both full doses of the vaccine, whereas the participants who received a low dose followed by the full dose were 58.9% percent less likely to have an asymptomatic infection. There were 10 cases of COVID-19 that resulted in hospitalization, which occurred 21 days or more after treatment (two were classified as severe cases of which one died). All of these cases were in the placebo arm. Lastly, using data from all participants who received at least one dose from all four phase 1, 2, and 3 trials (74,341 person-months of follow-up), the vaccine appeared extremely safe with a very low incidence of severe adverse events that were evenly split between the two arms (84 events in 12,021 participants in the vaccine arm and 91 in 11,724 participants in the control group), with only three events being classified as being possibly related to the vaccine or placebo.

Study Strengths

This is the first published assessment of the efficacy of a COVID-19 vaccine, and one of the primary strengths of the paper is the large amount of detail the researchers provided to explain their results. Another strength of the study is that the researchers assessed protection from asymptomatic infections and viral shedding, which is not being measured by all COVID-19 vaccine trials. In addition, by combining results from the UK, Brazil, and South Africa the study analyzes both safety and efficacy in diverse populations facing very different epidemics. Given this diversity in the study populations, the similarity in results between the UK and Brazil increases the overall confidence in the effectiveness of the vaccine. Lastly, the authors clearly state the role of the funders in the presentation of the results, and a detailed listing of any conflicts of interest.


While the variability in the doses and the diversity of study sites and protocols combined in this report are a strength from a research point of view, it is also a limitation in the ability of the researchers to precisely determine the overall efficacy of the vaccine. While the authors reported that when combining all their data the vaccine is 70% effective, this number combines two different dosing strategies, used in different populations, and with varying timing between doses, and is therefore is somewhat problematic. Although, it should be noted that this strategy was discussed and approved by regulators prior to the analysis. The authors also report the results of the trial according to dosing strategy used, which is a more accurate way to present the data. In addition, due to an error in dosing the vaccine in early phase 1 and 2 studies, a portion of the participants in the UK cohort received an initial lower dose of the vaccine, which resulted in better protection for both symptomatic and asymptomatic infection. Those in this low dose cohort also received the second dose after a significant gap (<1% within 8 weeks of first dose). Additionally, although older participants (>55 years) were included later in the trial, the numbers are too small to make firm conclusions on vaccine effectiveness in this population. However, this research team has previously shown that the vaccine does induce a similar immune response in older individuals, suggesting similar efficacy would be expected in this population. Due to the nature of the pandemic, it is impossible to know the duration of protection beyond 6 months, but further analyses are planned to address this issue. Lastly, as this is an interim analysis, the efficacy results of the vaccine may change slightly as more follow-up data become available, though this is not expected to result in a significant change in efficacy.

Value added

This study provides the first detailed assessment of the safety and efficacy of an adenovirus-based SARS-CoV-2 Spike vaccine in multiple populations, and provides encouraging results that this vaccine can provide moderate to strong protection from symptomatic COVID-19.

This review was posted on: 9 December 2020