As New York Governor Cuomo has said, “we need an army of contact tracers” to track down the people who have been in contact with someone who has tested positive for Covid-19. I’m rather sick of hearing it because until you get to the next level of detail, this is rather meaningless. If we could just call things into existence, then this wouldn’t be a problem but calling things into existence is magic which we’re not very good at doing. On the other hand, “it would be hard” is why a lot of doable things never get done. Is this a case of needing magic or is it something that would just be really hard?
To the degree we can believe anything coming out of China, there were 1,800 contact tracing teams of five people. By the way, there wasn’t any testing going on because it didn’t exist! In any instance, that’s 9,000 people. New York State is about twice the size of Wuhan so the governor needs about 18,000 people.
By way of comparison, the government hired 500,000 workers for the U.S. census.
Let’s do another check on the numbers. Massachusetts is going to hire 1,000 contact tracers. Massachusetts is about 1/3 the size of New York State so 1,000 seems low by maybe 5,000 people. Of course, it depends a lot on what people are tasked to do and how efficient they are in doing it.
“One person who has recently been diagnosed has been in contact with you,” the script told her to say. “Do you have a few minutes to discuss what that exposure might mean for you?” Forty-five minutes later, Ms. Cross hung up the phone. They had giggled and commiserated. Her file was crammed with information.
Well that’s not going to work! Oh, well. We’ll chalk this one up to start up problems but the basic issue is that this is a numbers divided by time game and phone tracing is going to have be really quick and efficient.
Each time a citizen tests positive, the results will be immediately shared with a case investigator through a secure database. Within two hours, the case investigator will aim to reach the patient by phone and compile a list of every person he or she had been in close contact with in the 48 hours before the onset of symptoms.
The names of the contacts — the expectation is 10 people per new case — will then be passed to contact tracers, who will attempt to reach each one by telephone within 48 hours, calling back three times in succession to signal the call’s importance. For now, tracers are not leaving messages or callback numbers.
The contact tracers, speaking from their prepared script, will inform the contact of approximately when they were exposed, and then take an inventory of symptoms, talk the contact through quarantine requirements, and help arrange assistance with food or housing if the contact cannot easily quarantine.
Here’s where I say “so what?” Let’s say I received that phone call, what would I actually do? Really nothing more than what I’m already doing. I might decide to wear a mask which may or may not do any good. And, by the way, you have to quarantine everyone in the house as well so it’s not necessarily just one individual.
Why was this so effective in China and Korea? Because they slapped a tracking device on you and if you left your house, you were arrested.
Ramping up testing will also be critical for effective contact tracing, as researchers will need to be able to test everyone with symptoms as well as everyone they came into contact with, and obtain results quickly. To support that effort and contain the outbreak locally, San Francisco may need to be able to test as many as 130,000 cases per month, Rutherford estimates.
San Francisco is currently doing about 60,000 tests per month but ramping quickly.
The United States is currently testing people at a rate of 150,000 per day.
This is generating about 40,000 positives per day which means contact tracers need to try to reach about 400,000 people per day. A good telemarketer can make about 10 calls per hour or 80 calls per day. This would mean about 5,000 telemarketers. It looks like we’re generally about 1/2 the capacity of testing we’d like so that would mean about 10,000 people.
What this tells me is that we need 5,000 people immediately ramping to 10,000 people in about three weeks.
There’s my take on it. I’m still not completely convinced that even with all the contact tracing that anything we’ll be able to do will accomplish much more than remind people to stay home if they’re sick. On the other hand, at less that half that of the number of people to take a census that doesn’t even ask if you’re a citizen, this isn’t the worst investment the country has ever made.
By the way, here’s my solution to the staffing problem — give it all to telemarketing companies! They already have the trained people as well as the phone bank technology and there are something like 511,000 in the United States. And, of course, they’ll destroy all the data! Well, maybe not.