Can Burna AI run across multiple sites in the same trial?

Yes. Every site in a trial grades on the same engine against the CTCAE version the protocol names, so the same event is graded by the same criteria wherever it occurs, with the criterion and source line attached. Each site's records stay scoped to its own organization. The sponsor reads graded safety data for its own trials in a separate sponsor view, including inter-rater agreement per site, measured as Cohen's kappa over a rolling four-week window, so a site drifting from the others shows between monitoring visits rather than at them. Blind grading, where raters grade the same case without seeing each other's grades and an adjudicator resolves disagreements, gives a site or a CRO a direct measure of grading consistency across its team.