AstraZeneca and Oxford University have given conflicting accounts of how they came upon the most effective dosing pattern for their COVID-19 vaccine, a rare instance of public dissension between major institutions collaborating on a pivotal project.

The discrepancy, reported for the first time by Reuters, centres on the regimen administered to a smaller group of volunteers in the late-stage trials, of half a dose followed by a full dose. This diverged from the original plan of two full doses, given to the majority of participants.

The half-dose pattern was found to be 90 per cent effective, versus the 62 per cent success rate of the two-full-dose main study, based on interim data.

AstraZeneca's research chief told Reuters 10 days ago that the half-dose was given inadvertently as a first shot to some trial participants, and emerged as a stroke of fortune - "serendipity" - that scientists expertly harnessed.

This narrative was refuted by a leading Oxford University scientist, however, who told Reuters on Monday that the half-dose shot was given deliberately after thorough consultations.

Uncertainty over how the dosing regimen came about raises questions about the robustness of the data, according to some experts who said it risked slowing down the process of gaining regulatory approval for the shot and denting public confidence.

"These differing accounts are worrying," Eleanor Riley, a professor of immunology and infectious disease at the University of Edinburgh, told Reuters. "Trust is at a premium when it comes to vaccines and we must not do anything that might in any way undermine it."