Yes. I mean, look, there's flare-ups out there, Texas, Florida, California, South Carolina, around the world. I mean Europe and certainly Asia is pretty much back up to not quite 100%, but not that far. The main issue here is, in some parts of the United States - but, again, we don't disclose exactly which sites are where, but you should know that it's not going to be that material, these flare-ups in the U.S. to our - we have over 100,000 sites globally. And we've become very adept at transitioning to remote monitoring. So, I think - yes, I mean, we now have been able during the second quarter, to adapt, and we have a number of workarounds using our remote capabilities. And so, we feel that these assumptions that we made for the third quarter, I think, are, in a way, better educated than the ones that we made three months ago, when we were just learning to adapt to the crisis. I should point out that these assumptions that we made in April 2020, many folks out there thought that we were getting ahead of ourselves and we’re maybe too optimistic about those assumptions. And the fact is, the progression of site accessibility as a metric was exactly as we predicted. And we base that on the course of the disease in China and other geographies that were more - that had been ahead of the rest of the world in terms of the disease. And we base that also on our own internal data and modeling that we are using to project our business. So, once again, while people thought we were overly optimistic, it turned out we were exactly on target with respect to site accessibility and, in fact, not so optimistic. So, I think, I'm not suggesting that, because we were right three months ago, we are right now for the next three months. But, I think, we - if anything, I believe, we are a little bit more educated now, and our models are even more precise than they were three months ago.