November 23, 2020

Oxford–AstraZeneca COVID-19 vaccine

Oxford and AstraZeneca announce preliminary AZD1222 efficacy results

The headline 90% rested on three symptomatic cases among 1,367 vaccinated volunteers — and the regimen that produced it began with an unintended half-strength dose.

The first interim readout

On 23 November 2020 Oxford University and AstraZeneca released the first interim efficacy analysis from their UK and Brazilian trials of the vaccine then called ChAdOx1 nCoV-19, or AZD1222.

Oxford reported 70.4% pooled efficacy against symptomatic COVID-19 across two dosing regimens: 90% in a low-dose then standard-dose subgroup, and 62% in the two-standard-dose subgroup.

AstraZeneca said the analysis covered 131 COVID-19 cases and 11,636 participants — 2,741 on the low/standard regimen and 8,895 on standard/standard. The data had been locked on 4 November, after the independent monitoring board recommended the analysis.

Where the 90% came from

The low first dose was not the intended regimen. It arose from a dose-quantification discrepancy, and everyone in that subgroup was aged 18 to 55.

Three symptomatic cases occurred among 1,367 vaccine recipients in the subgroup, against 30 among 1,374 controls. The peer-reviewed figures were 90.0%, with a 95% confidence interval of 67.4% to 97.0%, and 62.1% for two standard doses, with an interval of 41.0% to 75.7%.

This was a pooled analysis of trials using different regimens and intervals, not a head-to-head trial designed to show a half dose was better; the second injection came at least one month later, with intervals varying substantially.

What the numbers did not settle

No hospitalizations or severe COVID-19 cases had occurred among vaccine recipients in the analysis reported that day — encouraging, but too small a dataset to establish complete protection against severe disease.

Regulators did not adopt 90% as the vaccine's efficacy. The EMA authorized two standard doses and calculated about 60% efficacy from the standard-dose evidence.

Oxford said it would support AstraZeneca in submitting the interim efficacy and safety data to regulators in the UK, Europe, Brazil and elsewhere; the full peer-reviewed interim results followed on 8 December.

Sources

Researched 24 Aug 2026 7 sources not yet audited

How this was checked

Researched from the web into a fact sheet and rewritten from that sheet. This article has not yet had an independent model audit. How the pipeline works →

What the sources leave uncertain

The date is secure: 23 November was the public announcement date, although the analyzed data had been locked on 4 November.
The 70.4%, 90% and 62% figures were preliminary. They came from a pooled analysis of trials with different dosing regimens and intervals, not a completed head-to-head trial designed to prove that a half dose was superior.
The low-dose first injection arose from a dose-quantification discrepancy and was not the originally intended regimen. All participants in that subgroup were aged 18–55, and only three symptomatic cases occurred among 1,367 vaccine recipients versus 30 among 1,374 controls.
Later regulators did not treat 90% as the established efficacy of the authorized vaccine. The EMA authorized two standard doses and calculated about 60% efficacy from the standard-dose evidence.