Maintaining confidentiality of emerging results in COVID-19 vaccine trials is essential

[...]access to information on the timing of events—ie, the event rate by time after randomisation—in the pooled data could be informative about vaccine effects, since such a rate in the placebo group would be expected to be fairly constant over time. [...]information should not be used to alter criteria for defining primary events or for changing the time after the last vaccine or placebo dose when events would be included in the primary analysis. Protocol revisions to these crucial elements of a trial design should be prohibited after availability of trial data that may be informative—even indirectly—about vaccine efficacy, unless these potential adaptations were prespecified in the protocol.5 This issue is important since such revisions to a trial design, if potentially influenced by emerging trial data, would compromise the reliability of a trial originally intended to be confirmatory rather than hypothesis-generating, and since it would be difficult to conclusively show that emerging trial data had no role in decisions about these changes to the trial.2,5 Design changes during the conduct of a trial—eg, a vaccine regimen is altered or the demographic composition of a trial is changed to include more participants in subgroups in which early results suggest greater efficacy—can compromise the interpretability of the results and the generalisability of the trial findings. Influence of such motivation has been shown in comparisons of open-label investigator versus blinded central evaluations of disease progression in oncology trials.13 Early clues about preliminary trial results have the potential to influence trial enrolment and health-seeking behaviour, risk behaviours, or reporting behaviours of trial participants, which could alter the likelihood that cases are reported or that cases meet the actual case definition for the primary endpoint or for secondary severe disease endpoints.