1 in a 1,000 is definitely not accuraet. They’ve already had more 26,000 test positive out of a population of 20 million, which is 1.3 in a 1,000, and that wouldn’t include anyone who had it earlier or anyone assymptomatic or mild cases..
I'd count that as a rounding error, she was having casual conversation.
That said, the test rate is low. Using the roughly 25% of those tested, test poistive (compared with 10% most other places), about 100,000 tests done to get 26,000 positive. As more tests are done, more positives will certainly be found.
But I do agree that 1:1,000 figure is odd. Will have to go back and see what she was referring to.
Found the source of the statement ...
Remarks by President Trump, Vice President Pence, and Members of the Coronavirus Task Force in Press Briefing - March 23, 2020
DR. BIRX: ... Finally, to conclude: New York City. The New York metro area of New Jersey, New York City, and parts of Long Island have an attack rate close to 1 in a 1,000. This is five times what the other areas are seeing. There -- through the high-throughput lab investigations, we're finding that 28 percent of the submitted specimens are positive from that area, where it's less than 8 percent in the rest of the country.So to all of my friends and colleagues in New York: This is the group that needs to absolutely social distance and self-isolate at this time. Clearly the virus had been circulating there for a number of weeks to have this level of penetrance into the general community.
In this case, the attack rate is the product of number of tests times the rate of positives (28%) divided by the population. It has no bearing on the actual infection rate, which is unknown and not really sought yet. We have close to zero data on the asymptomatic.