I was talking about the rate of tests done and % of those tests going positive. If the percentage is staying the same/falling and the testing is sufficently high (+3k) we could think that there may be a trend. Now, we need a truly randomized trial to eliminate bias, but i dont think that's feasible