Why Do NYC And NY State Report Different COVID Test Positivity Rates?

After days of suspense for many parents, the closure of New York City schools was ultimately decided on a razor's edge. City health officials on Wednesday reported that the seven-day average positivity stood at precisely 3%, the threshold above which Mayor Bill de Blasio had said he would shut down schools.

Around the same time, Governor Andrew Cuomo announced that the city could face greater restrictions, including some business and school closures, if the seven-day average positivity rate were to hit 3%.

Both events illustrated one inescapable and painful truth of policymaking during a pandemic: never has the fate of the city been so tightly bound by one single public health metric. Or at least, one metric in principle. Because while de Blasio announced an average positivity of 3%, Cuomo stated the city's positivity as being 2.5%.

So which one is right? As with many matters involving Albany and the city, the diverging numbers involves "a difference of opinion," as Cuomo might say.

Here's what you need to know about the two positivity rates.

State and city health officials employ different methodologies for calculating positivity.

Test positivity is defined as the percentage of positive tests out of the total tests conducted on a certain day. While the mathematical computation is the same, health officials are faced with a choice: do they calculate the number based on the day tests are reported or on the day the tests are performed?

State health officials have elected to report positivity based on all tests reported to the state on the previous day. This allows Governor Cuomo the ability to announce a positivity that reflects the most recent tranche of test outcomes.

But their counterparts in New York City have chosen to base their calculation of positivity on the day the tests are performed. This means that they report on a positivity rate from two days ago in order to allow time for those results to come in.

So which method is better?

It depends who you ask. Corinne Thompson, a director of epidemiology with the New York City Department of Health, recently told Gothamist that no one method is considered right or wrong. The state's method has a sense of immediacy. In a perfect world where tests had a rapid turnaround, this presumably gives epidemiologists an accurate portrait of the most recent infections.

But as Thompson noted, labs routinely experience delays, so it is possible that the tests being reported by the state reflect infections from many days ago. In the summer, for example, some New Yorkers witnessed delays of up to two weeks.

With that in mind, the city's approach prioritizes accuracy over recent data, a point that the city's Health Commissioner, Dr. Dave Chokshi, made Wednesday during the mayor's press conference.

"We're always striving to find the right balance between providing the most recent data as well as the most accurate data," he said. "If we were purely going on having, you know, the utmost accuracy, then that would mean that we would be reflecting the reality from several days ago."

Two outside experts contend that the city's positivity is a better snapshot of a particular point in time, one that gets revised, becomes more accurate as more data becomes available.

Wafaa El-Sadr, an epidemiology professor at Columbia University, stressed that in general, the seven-day average positivity is the best indicator, as opposed to the daily positivity, which takes only one day into account.

"This avoids the swings that can happen due to delay in reporting as one example or big variations in testing results," she said.

In terms of which way to calculate, she said it should be the day test is performed

Dr. Denis Nash, an epidemiology professor at CUNY, said that while the drawbacks of the city's approach is that health officials must wait several days, "it may be worth the wait, since accuracy and meaning are critical."

He added: "The state approach will be more timely and day X‘s value won’t change on day X+1 as new data come in. But it will be harder to interpret the values from day to day, since a given daily result mixes negative and positive results from specimens collected on different days, maybe wildly different days."

How much can the positivity rates vary?

A lot. One instance that stood out to some astute New Yorkers came on September 29th, when New York City reported a that the daily positivity rate rose to 3.25%, a headline-grabbing jump that came on the first day of in-person school.

The state on the same day reported a positivity rate of 1.27%, meaning that the city's figure was more than double that of the state's.

Over time, the discrepancies can get smaller since city health officials will update the positivities as more data comes in. But the gaps are likely never be closed. For example, the 3.25% positivity from September has since been only slightly downwardly revised to 3.10%.

So which one should I follow?

Both. As New York battles a second wave, the governor and mayor have, for better or worse, leaned on the positivity rate to determine policy that has and will effect the lives of city residents. In each instance, they are relying on the data generated by their respective health departments. Such a scenario can result in bad optics like schools getting shut down before indoor dining and other businesses. Both rely on positivity triggers of 3%, but because they are not calculated the same way, the timing of city and state restrictions won't coincide.

Some experts have warned against relying too much on positivity altogether, given that it can vary wildly depending on who in the population is showing up to get tested.

Dr. Wan Yang, an assistant epidemiology professor at Columbia University, is part of a team that produces epidemiological models predicting deaths and hospital capacity that are used by New York City health officials.

"The test positivity rate is very confusing," she told Gothamist. "Sometimes people get it mixed up with the number of people having the virus which is really not the case."

"Personally I think it’s really not an informative metric in terms of how to get at the infection rate," she added.

Instead, Yang focuses on hospitalizations, deaths, new cases, as well as other data.

In terms of shutting schools, she said the mayor should have based his threshold on the random testing being performed at city schools rather than that within the general population, which is skewed by the many sick people seeking out tests.

Nonetheless, she said she sympathized with policy makers, who have found themselves scrambling to come up with ad hoc rules that the public can understand and accept.

"We do have to keep that in mind. They are in a very difficult position," she said.