New York City hit a 3% positive test rate. Or did it?
By Joseph Goldstein and Jesse McKinley
It is the most important number in New York City right now, a critical threshold that triggers restrictions by state and local governments in response to the coronavirus. The mayor of New York shut down public schools at 3%. The governor said that a sustained 3% level in the city would result in banning indoor dining, closing gyms and hair salons, and placing a 25-person cap on attendance at houses of worship even as the holidays approach.
But as important as that number is, it seems the city and the state can’t agree on whether we’re there yet.
That’s the situation that has played out over the past week, with Mayor Bill de Blasio saying 3% has been breached, while Gov. Andrew Cuomo said it is well below that. Each relies on his own statistics, which are compiled and reported in different ways, varying even on which tests to include in the calculation.
The discrepancy can be striking: On Sunday, for instance, the city said its seven-day average was 3.09%. Cuomo’s office, however, put the city’s rate more than a half-point lower, at 2.54%.
In one way, of course, not all numbers are equal, as Cuomo’s statistics from the state Department of Health govern a wider array of activities and businesses in regions all across New York. But Cuomo granted local school districts the right to set their own parameters for school shutdowns, and de Blasio, who controls the school system, set 3% as that level.
And so, in the end, it was the city’s numbers that caused the temporary pause of in-person learning in the nation’s largest school system.
Why can’t they agree on whether the city is at 3%?
The cause of the discrepancy lies in both the tests that are included and the time frame in which statistics are reported, leading to the mayor and the governor giving different numbers each day.
It is the latest discordant message between two rivals that has played out over the entire pandemic, adding a level of dysfunction and confusion to the response.
On Sunday, Cuomo suggested the city’s calculation of its positive rate was “confusing and unnecessary” and also “irrelevant” because the state numbers would govern any broader restrictions.
The state and city health departments have different accounting rules for tracking the spread of the virus. The state treats a new case as arising on the day the test results came in. The city dates each new case to the day the sample was provided.
So if an infected person goes to a clinic to have his nose swabbed Monday, that sample is often delivered to a laboratory where it is tested. If those results are reported to health authorities Wednesday, the state and city would record it differently. The state would include it with Wednesday’s tally of new cases, while the city would add it to Monday’s column.
The 3% threshold is based on a seven-day rolling average. It matters what day a new case is registered.
Another factor contributes to the discrepancy as well, which has received little attention so far: antigen tests. New York state includes the tests in its official metrics. But while they are generally faster, they are less likely to detect the infection in people with a low viral load.
New York City, however, does not include antigen tests, preferring a more sensitive one known as a polymerase chain reaction test. The city only includes PCR tests performed in a laboratory in its count. That’s why the state — which is counting both antigen and PCR tests — may have a higher tally for overall cases in New York City but a lower percentage of positives.
PCR tests in a laboratory have long been considered the gold standard because they are so unlikely to miss any infections. But some public health experts say that much of the PCR testing for the coronavirus is too sensitive, resulting in coronavirus diagnoses for people who are carrying relatively insignificant amounts of the virus and are probably not contagious.
Antigen tests, which can be performed rapidly and cheaply, detect bits of coronavirus proteins. But they are more likely to miss cases, including people recently infected who have lower viral loads.
The difference in sensitivity between the two types of tests can contribute to a gap in the positivity rates between the city and state.
In fact, the positivity rate of the same group of people — in this case, New York City residents — can vary depending on how many receive antigen tests versus traditional PCR tests.
Antigen tests can miss some cases when the amount of virus is still low.
Say 1,000 people get tested. Let’s assume all have a PCR test and that 30 tests are positive, for a positivity rate of 3%. Now let’s assume half get an antigen test and half get a PCR test. Maybe only 25 tests come back positive, for a positivity rate of 2.5%.
There are other differences in how the city and state calculate the positivity rate that also help explain why the city’s rate is higher.
For instance, if a person tests negative repeatedly within a seven-day period, the city counts only a single test when calculating the positivity rate. But the state includes all negative tests from different days in its calculation.
On the other hand, when someone tests positive repeatedly, the city includes each positive test when calculating the positivity rate. The state, however, only counts the first positive test and ignores subsequent ones.
That last variable — whether repeated positives are counted or not — can have a big effect on the positivity rate. Gareth Rhodes, a member of Cuomo’s coronavirus task force, estimated that about 15% or 20% of positive PCR tests currently emerging from New York City were for someone who tested positive in the past.
Why does it matter?
If New York City hits 3% based on the state’s metrics, the governor has suggested it will enter into a so-called “orange zone,” the middle of three color-coded restrictions that the governor enacted in early October.
Under the governor’s plan, a so-called “red zone” is subject to the toughest restrictions, with bans on mass gatherings, indoor dining and in-person learning. Nonessential businesses would close, and religious services would be limited to either 25% capacity of houses of worship or 10 people, whichever number is lower.
“Orange zones” — which the governor said the city is in danger of falling into — allow slightly large gatherings and outdoor dining, but schools shut down, as do “high-risk” nonessential businesses like barbers and gyms. Precautionary zones — appropriately yellow — allow in-person classes but ramp up testing and put some less stringent caps on dining and gatherings.
Four of the city’s five boroughs — Queens, the Bronx, Brooklyn and Staten Island — already have yellow zones, and Cuomo said upper Manhattan may have one imposed later this week. Staten Island may also face intensifying restrictions.
These zones are reevaluated after two weeks, and the metrics for entering and exiting each of these zones involve a complicated mix of data and a dash of deliberation. Restrictions can be modified on “expert advisement,” the state says, and include determinations based on local hospitalization rates or whether outbreaks can be traced to a single source (like a prison, gathering or group residence).
Can the city act on its own?
Based on its metrics, New York City has already hit the 3% mark. Yet it may still be some time before businesses face any restrictions.
Not only do state statistics put the city’s seven-day rolling average below that number, but New York City must breach that threshold for 10 consecutive days before the city would enter the “orange zone.”
Technically, the city could use several workarounds to target businesses it decides are contributing to virus spread. For example, the city’s Department of Health could try to close any restaurant by declaring it a public health threat.
But the governor has broad emergency powers over a range of businesses, trumping de Blasio. At news conferences, the mayor sometimes sounds a resigned note, saying that certain decisions about closures must be left to the governor.
So, for now, New Yorkers will continue to wait for the city to hit 3% … again.