By Gideon M-K
-the changes we are seeing from day to day are truly enormous.
However, despite this, there seems to be one simple thing we can’t get over during Covid-19: the question of influenza.
Virtually every time you see an article published about the deadliness of coronavirus, the author has chosen to compare it to the flu, because if there’s one thing that we need more of, it’s analogies.
The problem is that comparing Covid-19 to the flu has serious issues. Covid-19 is a new disease, which has only been studied for just over half a year, which makes studying things like the number of deaths it causes overall somewhat challenging. On the other hand, we’ve been studying influenza for decades, and know an enormous amount about the disease.
Now, this doesn’t make the comparisons entirely a waste of time, but it does make them somewhat problematic. Take the death rates for both diseases – many people have argued that Covid-19 is only about as deadly as the flu, because influenza kills about 0.1 per cent of the people it infects and some studies have shown that the true infection-fatality rate – the proportion of people who die from the disease divided by the total number of infections, including asymptomatic cases – of COVID-19 is around the same ballpark.
The problem is that, while this comparison seems reasonable at first, it’s actually deeply flawed.
The first thing to do is to figure out the infection fatality rate of both diseases. For influenza, that’s relatively simple – just go to the CDC’s website, and they have easy-to-find estimates right there about how many people get the disease and die each year.
If you divide those two numbers, you end up with a rate of between 0.1 per cent and 0.2 per cent – or between 100 and 200 deaths per 100,000 infections.
So, with influenza out of the way, let’s look at Covid-19. There are dozens of estimates out there, but a colleague and I conducted a systematic review and meta-analysis to answer this very question and came up with an overall number of 0.64 per cent – or 640 deaths per 100,000 infections.
Right off the bat it seems like Covid-19 is worse than the flu –0.1 per cent compared to 0.6 per cent makes it about six times as deadly. But some people have argued that Covid-19 is much less fatal than this aggregated estimate would suggest, with an infection-fatality rate of just 0.1 per cent. What about then?
Let’s go back to the CDC’s estimate of influenza fatality. This is the most widely cited number, and is used as the benchmark for flu deaths by people across the world. How is this number calculated?
Firstly, there’s a problem – the CDC doesn’t account for asymptomatic cases in their calculation. This 0.1 per cent number is actually the CASE fatality rate, which is the rate of deaths in people who have symptoms of the disease – not the INFECTION fatality rate that we’re interested in.
According to CDC estimates, the rate of asymptomatic infections in influenza may be as high as 50 per cent, so straight away we need to reduce this figure by quite a bit.
Moreover, it turns out that the 0.1 per cent figure is more complex than you might’ve guessed. It is very hard to accurately know when people have died from influenza, because often they are old and frail anyway, and so the cause of death is listed as pneumonia, or simply an underlying condition that contributed mostly to their death.
This is particularly a problem because we don’t always test for flu – if someone has symptoms, people are often diagnosed without a test. For example, if someone catches the flu at 95 in an aged care institution and eventually passes away from pneumonia, the cause of death might be attributed to the pneumonia and not the flu.
Because of this, the CDC does something clever. They take all of the disease codes – a standardised record of the diagnoses someone had – across the US, combine this with known data on excess deaths and cases, and use a statistical model to estimate the number of deaths that have been caused by influenza every year.
Now, this is a big problem. We’re comparing this number to an infection-fatality rate calculated by dividing official Covid-19 deaths by infections determined from antibody testing. We haven’t modelled the number of people who are likely dying from Covid-19 but aren’t being reported officially – but, according to a number of sources, that figure is extremely high.
So what happens if we compare apples with apples? A 2014 systematic review into influenza looked at infection-fatality rates calculated as deaths as a proportion of infected people estimated from serological testing – the main source of data for our Covid-19 estimate of 0.64 per cent – and found that between one and 10 people died per 100,000 influenza infections.
This gives an infection-fatality rate of between 0.001 and 0.01 per cent, which is quite a lot lower than even the lowest estimates for Covid-19.
In fact, if we take a reasonable range from most of the published research, it looks like Covid-19 has a fatality rate roughly between 50 and 100 times higher than influenza. In other words, between one and 10 in 100,000 people who get the flu will die, while it is between 500 and 1,000 in 100,000 people getting Covid-19 who will.
This puts into context all of the cheerful statements this year that Covid-19 is just the flu so we can safely ignore it. Yes, influenza is a serious disease, but the evidence shows quite strongly at this point that Covid-19 is far more of an issue. When we take the most appropriate figures on both diseases, Covid-19 comes out as deadlier in every aspect.
Ultimately, the comparisons probably aren’t going to go away. Analogies are helpful for us to understand reality, and there will always be people who want to minimise the coronavirus and make it seem as routine as the flu. Sadly, the evidence shows that Covid-19 is far deadlier than influenza, and just overall a very nasty disease.
We all wish that coronavirus were mild, as that would make our lives so much easier. Unfortunately, the evidence shows that it is not.
Gideon M-K is a Health Nerd epidemiologist.