I understand where you are coming from and I had the same concern, but doesn't the p value not need to be as low for a completed trial as it has to be for an interim to stop the trial?

Please correct me if I am operating under the wrong assumption.