I know of no variable we are trying to isolate with a 'white' vs 'everyone else' besides, perhaps, a hamfisted approach at trying to stratify over racism. Its very possible that the race of the healthcare provider has an impact depending on the race of the person receiving treatment, but I have no idea why they would make it 'white vs non-white' as the criteria.
Skimming the paper results, it might have been dredging for a significant effect showing that white people are less-badly affected. "The vaccine has racist side effects" is an eye-catching headline.