Peak height, number of cases, deaths, severity – these are all terms that are used to give you an idea of how big a flu season has been. Here’s another one – the “area under the curve”. It doesn’t look at the height of the highest number of cases or laboratory detection; it looks at the entire epidemic curve, even the low bits of those “mountains”.

The epidemic curve

An epidemic or ‘epi’ curve shows the number of infections, detections, or cases of disease over time. It has a start and an end.

I’m going to show some graphs from the European Centre for Disease Prevention’s (ECDC) Guidelines for presentation of surveillance data to get you in the groove.

There are lots of ways to present these numbers. The simplest is called a histogram – totals represented by bars that are in contact. Their height is measured against the x-axis (vertical) and is usually the number of detections, cases or infections. A series of related values – like age groups, days or weeks are presented along the y-axis (horizontal).

If you have detailed data, you can show what’s happening ‘inside’ each bar using a case plot, as shown below. Each box represents a case and includes additional details indicating whether it’s a confirmed or probable case of disease. All those who developed symptoms in the same week are stacked on top of each other. Good for small outbreaks or the start of an epidemic.

You could overlay a line, with the midpoint of each column of cases being the connection point (top mixed graph below). By itself, it shows a little less detail, but you can replace those cases with two lines (second graph below), which may more clearly show the rise and fall of confirmed and probable cases during that time.

Epidemic flu in Australia

Let’s have a look at flu detections by month across a quarter of a century.

Blergh. That’s a lot of flu epidemic peaks. You can see that some peaks are higher than others, and your gut may tell you that means those were the bigger years. But your gut’s wrong. Let’s zoom in and see if I can show you why.

Below, I’ve zoomed in on the 2017 to 2026 period. I started in 2017 because it was Australia’s ‘biggest year on record’. At the time. Then 2019 was. Then 2022 was…

You can see that 2022 had the highest peak in detections, followed by 2024 > 2017 > 2025-6 > 2019.

I took that graph and manually added four coloured shapes (polygons), each covering the area of an entire epidemic peak for four years: 2017, 2019, 2024, and 2025-6. Each shape includes the area under the curve between the month with the lowest case numbers – the nadir – on either side of the annual peak.

For 2025, that peak (red) extends into summer and is not yet complete.

I’m sure maths professionals could do this more accurately, but my imperfect attempt should still help illustrate another way to define a big flu season.🙂

Next, I went over to SketchandCalc and got it to calculate the area for each of those shapes.

Which year was the biggest in terms of flu detections?

Even though 2024 and 2017 have higher peaks, 2025-6 had the biggest area under the curve.

• 2025: 43.8 cm²
• 2024: 28.6 cm² (65% of 2025)
• 2019: 26.9 cm² (61% of 2025)
• 2017: 20.9 cm² (48% of 2025)

2024 was only 65% of the area under the 2025-6 curve, and 2017 was just 48%. These were big years, but 2025 was really big.

This example shows that a pointy peak may mislead you into thinking it’s a huge season when it’s just a huge pointy peak! The timing of the season’s start and end, and the speed of its rise into and fall out of that peak, define the season’s size. 2017 started normally and rose and fell very quickly. 2019 started very early and rose and fell a little more slowly. 2025 started early, rose slowly, and fell slowly because an entirely novel flu variant suddenly travelled from the US, dramatically prolonging the entire Australian season.

Epidemic peak height is not the best measure of season size.

How severe the season is is a separate issue again.

2025: lots of cases, sure, but lots of harm as well?

If the vaccine’s effectiveness, its time to production, and its uptake stay about the same between years, then more flu cases will generally mean more accompanying harm overall – apart from acute influenza, there’s the stress on the heart and head.

A truly severe year accounts for a higher-than-normal risk of hospitalisation and death. So 2019 may have been a #flunami, but, as I’d said at the time, it wasn’t a severe season; it was howevere, a really big season!

“Hospitals have been strained, not because of exceptionally severe flu cases, but due to the sheer number of cases at an unexpected time.”

In 2025, Australia had a record-breaking number of flu detections. But the long season, driven well into an unseasonal summer epidemic by an unusual emergence of a novel variant, didn’t look unusually severe.

The 2025 flu season was just as deadly and harmful as a ‘normal’ flu season, just scaled up due to many more cases, leading to more hospitalisations and deaths.

It’s noteworthy that in one jurisdiction, Queensland, 82% of cases were unvaccinated. This is despite the 2025 season’s flu vaccine producing protection at low-to-moderate levels comparable to past seasons, despite the K variant’s changes. Probably a message that came too late for most. We could each do more to prevent severe disease. Better communication is one of those things. Getting vaccinated is another.

So the area under the curve can highlight the entire size of a flu season, but not its severity. Just like the peak height of the epidemic, although the thw area method is better at telling you about the entire season’s size than a single peak is. 2017 looked huge based on peak size, but it was half the size of 2025. Of course, you have to wait until the season is over to get a final figure (or you can get an estimate earlier if you have no patience like me).

Severity is more about the number of cases (infected people) who test positive and end up very ill or dead than about the total number of cases.

Just because a season has a high peak or a huge area under the curve, that doesn’t mean it’s more severe than normal. Just that healthcare will be overwhelmed by the total number of sick people.