The Gothamist/WNYC newsroom is using statistics to shape our daily coverage of the COVID-19 epidemic. This article explores COVID statistics in New York City, then looks at New York State, and finally compares New York to other regions and looks at a projection of the future.
Please send any questions or comments to firstname.lastname@example.org.
A note on dates: New York City’s most-current data is released on a 3-day lag. This is because the Department of Health assigns every case, hospitalization, and death to the date it occurred, and reports take time to come in, be collated, and assigned. They also revise the data for older dates as new data comes in, so numbers for each day may change over time.
This set of charts reviews major COVID statistics in New York City over the last 90 days. After months of low numbers of infections, the city’s positivity rate began to increase in September, driven by clusters in Brooklyn and Queens. After a lull in October, positivity really began to rise in earnest at the beginning of November, due mainly to wider community spread. Throughout the period, the number of tests conducted in the city continued to rise, which also contributed to better detection of new cases. Hospitalizations and deaths also increased during this time, but to nowhere near the record numbers the city saw in April.
This chart helps explain the progress of the epidemic geographically during the last 90 days. Though all boroughs rose during this time, you can see the first peak was concentrated mainly in Brooklyn, Queens, and Staten Island, while the second peak has affected every borough, and was particularly concentrated in Staten Island.
This map shows the current positivity rates for each NYC neighborhood over the last week, to give a better sense of where COVID is currently concentrated geographically in the city. For total cases over the entire course of the epidemic, see the map below in the Positive Cases section of this article.
We use this scatterplot to help us identify ZIP Codes of concern each week. Look in the upper-right quadrant: those are the ZIP codes that have high positivity and have shown growth in their positivity over the last week: ie. they have an active and growing outbreak.
Positive Cases and Testing
Testing began in earnest in early March, but for weeks was limited in availability to the most symptomatic cases. This led to very high positivity rates in April, which gradually declined as the epidemic was brought under control by social distancing measures, and more people were able to get tests. Testing continued to grow every month since, which allowed the city to spot the beginning of the second wave in September, when it was still confined to a few neighborhoods.
Queens and Brooklyn have larger populations than the other boroughs, so they tend to have more cases, but when normalized for population, Staten Island and the Bronx have often seen more positives each week. This is particularly true during the second wave, with Staten Island outstripping the other boroughs.
Starting on 4/1, the New York Department of Health started to release positive cases by Zip Code information. You can see totals for the entire epidemic, as well as possible demographic associations, at our larger version of the map.
We’ve charted the positive case zip code data in a scatterplot. On average, the lower income, older, and more diverse a neighborhood is, the more positive cases it will have. Asthma rates and larger household size also show a positive association with cases. Population density shows a negative association with cases, mainly because Manhattan, where ZIP codes are densest, has had lower case rates so far in the epidemic. Median age of a neighborhood shows no clear association, probably because median age is insignificant compared to the other factors. You can examine individual neighborhoods and various demographic factors on our larger chart.
Over time, serious cases of COVID will put patients in the hospital, and once they’re unable to breath on their own, into the Intensive Care Unit. After a desperate struggle to build emergency hospital beds in April, it turned out that the city’s existing beds were mostly adequate, and in the months since, spare capacity has hovered around 20%.
During the first wave of the epidemic, the Bronx, Queens, and Brooklyn had more new hospitalizations per capita, but because Brooklyn and Queens are larger, they had more total new hospitalizations.
The best measure of a hospital’s ability to respond to additional COVID cases is spare ICU bed capacity, because the most serious cases require treatment only available in the Intensive Care Unit. The map above tracks capacity available using “staffed adult” ICU beds, ie. those adult beds that are both available and have staff to attend them. It is not unusual for a hospital to aim to keep most of its ICU beds full, because staffing empty beds cost money, but rates close to 100% can signal a hospital is close to being overwhelmed with cases.
COVID has a fatality rate between 0.5% and 1%, and New York City has suffered more deaths than any other city in the country. On April 14th, the NYC Department of Health began reporting “probable deaths”- people who had COVID listed as a cause of death on their death certificates, in addition to deaths of people with confirmed COVID tests. This raised the number of COVID deaths in the city by about 40%. This still may not include all COVID deaths, as more seemingly unrelated deaths may eventually be classified as caused by COVID. From February 1 through October 2, the CDC estimates there have been about 26,000 excess deaths in the city, compared to historical averages.
During the first wave, the Bronx had a higher per-capita death rate than the rest of the boroughs, but because Brooklyn and Queens have more population, they had the greatest death totals. During the second wave, deaths have so far been quite low compared to what happened in April, but Staten Island has been slightly above the other boroughs on a per-capita basis.
On May 18th, the city released data on deaths by zipcode. Neighborhoods with the most cases tended to have the highest numbers of deaths, but the maps of cases and deaths do not coincide exactly.
We’ve created a set of scatterplots exploring the correlations between deaths and demographic factors, across the NYC zip codes. In general, ZIP codes which are poorer, more diverse, or have larger households or higher asthma rates, have higher rates of COVID deaths. Population density is negatively associated with deaths, mainly because denser ZIP codes in Manhattan have less deaths. A ZIP code’s median age does not seem to have a strong correlation with deaths, probably because other factors outweigh that association. You can read more about our findings in a post on Gothamist.
Each day the New York City Department of Health releases demographic data on COVID-19 deaths. The majority of those who die of COVID are aged 65+ and/or those with pre-existing health conditions, which the DOH defines as: “Diabetes, Lung Disease, Cancer, Immunodeficiency, Heart Disease, Hypertension, Asthma, Kidney Disease, and GI/Liver Disease.” Note: these demographics include “confirmed” COVID cases only, and exclude “probable” deaths.
Older people, poorer people, men, and African-American and Latino New Yorkers, have been diagnosed, hospitalized, and died at much higher rates than other groups.
After a person has survived an encounter with coronavirus, their body produces antibodies, which are detectable on a test for a period of weeks or months, but which fade over time. After the big outbreak in April, and with limited antibody testing, rates were very high; as more people were able to get testing and antibodies from earlier exposure began to fade, the positivity rate settled in at around 20% of all antibody tests performed in New York City. We do not know if antibodies confer total or partial immunity to the virus, or if so, how long that immunity lasts. The human immune system is complex, and may build other defenses to coronavirus once exposed, such as T-cell response, which is not measured in an antibody test.
This map shows the total results of all antibody tests conducted by ZIP code, with neighborhoods that experienced the highest numbers of positive cases generally having the highest antibody rates.
COVID in New York State
About 42% of New York State residents live in New York City, so COVID stats for the entire state will tend to mirror the city’s stats. This is true for positivity, which shows the same large increase in April when testing was limited, and for total tests, which have risen steadily since then. However, when looking at positive tests, you see that during the second wave, the state suffered a more pronounced rise than the city, because of large second-wave outbreaks in upstate counties. Similarly, the increase in hospitalizations and deaths in New York State is more pronounced than in New York City, because of these upstate outbreaks.
Note on dates: in contrast to New York City, New York State releases its data with no lag, by assigning every test, hospitalization, and death to the day the report came in, which means that because of reporting delays, for the state data, each day in the data blends cases, hospitalizations, and deaths that actually occurred over a period of recent days.
Sources: for this New York State section, we are getting our data from the New York State Department of Health, partially via the COVID Tracking Project.
This chart allows you to look at positivity rates, new cases, and total tests for each county in New York State. Note that during the first wave, counties in and around New York City showed pronounced upticks, but during the second wave, we’re seeing more counties upstate with lines moving in the wrong direction.
This map shows the current situation in each of the New York State counties, colored by per capita population, to allow comparisons between counties.
Comparing New York to Other States
New York City was the epicenter of the outbreak in the United States in March and April. Since then, every other state has had its own outbreak, and some states have begun to approach New York in total per capita deaths. These numbers include both confirmed and probable deaths.
What Will Happen Next?
There are many models that predict the future of the outbreak, but Governor Cuomo has repeatedly praised the Institute for Health Metrics and Evaluation’s COVID-19 model. In their 2/25 update, the model predicted falling deaths through June, with total deaths of 54,121 by June in New York State.
You can find several other models at the CDC’s website.