To pile on to what @Tickerguy wrote, above, not only did the trials not do background surveillance testing of all participants, participants who developed symptoms were to report them to the trials staff to get a recommendation whether or not to get tested.

The folks making testing determination were not blinded to whether or not the participant had been vaccinated. Thus, they may have been more likely to not recommend testing for someone who was vaccinated, because they may have suspected less severe outcomes/lower chance of harm to that person, or because they were deliberately attempting to skew results.