> Aren't the efficacy numbers from the phase 3 trials based on giving periodic PCR tests to everyone that was participating in the trials?

For most trials, no. You have to look at each individual study, but most commonly, they only test after people show specific symptoms.