Because most philosophies that frown on reproduction don't survive.

Sunday, March 29, 2020

Flu vs the Coronavirus

This is the week that things got bad in several key hot spots for the COVID-19 pandemic in the US, particularly New York City. Congress passed the largest relief package in US history, at more than two trillion dollars. Much of the US population is under lockdown, with various state or city level orders for people to remain at home other than getting food or participating in essential work. And there are stories of swamped hospitals and paramedics starting to come out of the worst hit areas. And yet, for a lot of people, things still look relatively normal As a result, there are still a fair number of people asking, "Is this all an over-reaction? Don't more people die of the flu?"

I can't take credit for sounding an early alarm on the coronavirus. I'd heard of it early on, because the company I work for has offices in China and Taiwan, and so I knew about how it had affected our employees and customers. But I had assumed that it would not become a major problem in the US, not for any rational reason, but simply because there had been various pandemic scares before which had not turned out to be a major problem in the US. However, being a data person and someone who responds to big events by reading a lot of news, I have tried to do my own research over the last few weeks and it seems very clear to me that COVID-19 is worth worrying about. Yes, the isolation rules are having a huge effect on the economy, and I clearly recognize the risk of people trying to stop a massive pandemic and landing themselves in an economic depression in the process. I'm not going to address that question here -- though I will point out that an unchecked pandemic would probably cause a major recession anyway. My goal here is to show that the "what about the flu" comparisons miss how worrying the trends look.

The CDC lists Influenza & Pneumonia as the 8th leading cause of death in the US, with ~55,000 deaths per year, though it's well behind the leading causes, Heart Disease and Cancer, which account for 600,000 per year each. As people trying to argue that concern about COVID-19 is overblown have pointed out, we haven't had anywhere near 55,000 deaths due to the coronavirus yet, and flu also has a season during which cases increase exponentially.

The CDC provides data on deaths due to flu and pneumonia, and there are various sources that have been compiling the data on COVID-19. Let's start by looking at the data for New York State, which is currently the worst COVID-19 outbreak. I've pulled data on deaths due to flu and pneumonia by week from the CDC site here and I've compared it to data on New York State deaths due to COVID-19 here.

As you can see, pneumonia (which people often get as a result of the flu, but also results from other causes) is relatively constant through the year, ranging from 120 deaths a week in the summer to 160 deaths a week in the winter. Flu deaths range from zero a week during the off season to highs of around 20 a week during January and February.

In the last three weeks (coronavirus data is being updated faster than flu data right now, so I don't have flu data for the last two weeks) New York had zero COVID-19 deaths the week ending March 14th, 44 deaths the week ending March 21st, and 684 deaths the week ending March 28th. This virus is hitting New York much harder in the last week than the flu ever does.

If we look at similar data nationally, you can see why it's possible for people in other areas of the country to still wonder if we're overreacting. Just a week ago the number of COVID-19 deaths was lower than the number of flu deaths. But this last week it was well above flu and closing fast on pneumonia. (US COVID-19 data source here.)

The CDC also provides data on the hospitalization rate for flu and pneumonia in terms of the number of hospitalizations per 100,000 people. I compared this to the hospitalization data for COVID-19 for New York State. In terms of cumulative hospitalizations, there haven't been as many cumulative COVID-19 hospitalizations in the last three weeks as there have been during the 25 weeks of the flu season for flu and pneumonia:

However, compare the numbers for individual weeks you can see why they're concerned about hospital capacity:

We're still in the early stages on the COVID-19 pandemic, and we hope that the stringent measures that have been taken will slow the spread of the disease and eventually allow us to isolate it and let it die out. Because it's so new, and it's still somewhat regionally isolated in hot spots, the total numbers aren't yet as large as those for flu and pneumonia. However, when we look at the number of deaths per week and the number of hospitalizations per week, especially in the hard hit areas, it's easy to see why people are worried and taking major steps to stop this pandemic from getting worse.


Julia said...

Commenting from NYC here, where another statistic is staring us in the face: people fixate on the fact that only 20% of cases are serious. However, that needs a bit more definition. Last week 15% of cases required hospitalization. The reported death rate would be many, many percentage points higher without the infrastructure to support people for 10-25 days on ventilators or oxygen.

Granted, that's not 15% of all cases, because NYC will not test unless you are a) famous or b) have significant preconditions that put you at high risk (my 21yo's lab mate has an immunodeficiency disorder and was tested), or c) are already in such bad shape that you are probably in need of hospitalization. Hence the actual incidence of COVID-19 is probably at least 1/3 higher. But the frequency of serious illness is high enough that a group of friends and I discussed the possibility yesterday of setting up a Google spreadsheet to keep track of prayer requests.

Anonymous said...

It's also worth noting that in most of the US (and most of the world), COVID-19 is panning out to be far less severe than the flu. In Georgia, for instance, cases are still growing exponentially (as testing grows exponentially), but the death rate hasn't grown in over 2 weeks (despite only getting serious about social distancing in the past week or so), and the deaths are hovering around (still tragic, but negligible in a population of 10 million) 10 per day.

It seems we should focus our efforts on understanding the combination of factors that have led to disaster in Italy, Iran, Spain, and New York City.

Darwin said...


Good points. I stuck with graphing deaths and hospitalizations in the post because they seemed a bit more objective and some people have been spinning elaborate scenarios about how many un-detected cases there might actually be.

I hope you guys are all doing well and staying safe!


I take your point. Ohio, where we live, is also not a hot zone, though Gov. DeWine and his health department have been taking lots of measures to see that things stay that way.

Even so, the weekly deaths are climbing in a way that is worrying compared to flu. For instance, it looks like Georgia had 55 deaths in the just completed week versus 14 in the week before that. For comparison, it looks like the CDC shows around 7 deaths a week over the last three reported weeks from flu and around 65 deaths a week from pneumonia. So it's already almost as bad as flu and pneumonia, and if the number of cases each day are a leading indicator, it may be worse soon.

It does sound hopeful, however, that areas which are overall less dense are more resistant to the spread of the disease. If we end up with a dozen urban hot spots around the country and the others not as hard hit, the rest of the country will be able to help the hard hit areas.

Julia said...

You might find this good reading:

Julia said...

The WSJ has a good article titled Why Doesn't Flu Tank the Economy Like COVID-19, if you have access Published 4/10.