A transcript is as follows: 

HOST: We were talking about some of the models, whether it's from the imperial college guy in England or the University of Washington. Thank god some of these numbers are falling short of some of these catastrophic numbers. Tell me about the dynamic of the modeling and how it helps and influences decision making and then, when the reality comes in, how does the decision making transform?

DR. ROBERT REDFIELD: I think it's really important. First, models are only as good as their assumptions. Obviously, there was a lot unknown about this virus. The ability to actively make a lot of assumptions was much wider than if this was an Influenza B outbreak. Second thing, I will say from a public health perspective, to me, the real value of models is to have a model and then try to understand — if I invest resources here, what does that do to the model? If I invest in intervention strategies here, what does that do to the model? It's a way of beta testing how you're going to respond and what it does to the different models. And models should never be used to assume that we have a number. You saw those numbers are quite staggering. You've got 200,000 to 2 million Americans are losing their lives before the fall. That's a pretty staggering number.