> And the lack of evidence in this area suggests to me that the safety benefit is indeed negligible. It should only suggest to you that it is prohibitively exp

> And the lack of evidence in this area suggests to me that the safety benefit is indeed negligible. It should only suggest to you that it is prohibitively expensive to control for all of the different variables in order to provide a remotely reliable quantification.