An early look at influenza in the US highlights that the south does not predict the north

Influenza (flu) hit Australia earlier in 2019 than it had for many years. It reached a very high peak of cases and the flu season lasted for longer than usual. But it was not a severe season. Now we turn to the northern hemisphere for a quick early look at influenza in the United States (US). Guess what? Flu in the south does not reliably predict that in the north.

Influenza hit Australia early and lasted until the normal season end

The data are pretty clear on Australia having had an early flu season and a big flu season.

How early? I call it 7 months earlier than normal. That’s because, from November 2018, confirmed numbers were consistently elevated and didn’t drop back down until the normal season end, October.

A bit on the methods I used to make that claim

When I say elevated, what do I mean?

Throughout the 2019 Australian flu season I’ve chosen to compare each month’s laboratory-confirmed flu total to the highest total – not the 5-year average which would have made 2019 look even worse – for the same month and I’ve looked at every previous year from 2000 to 2018 to find that month, using NNDSS data.[1]

The range was 1.4- to 9.2-times higher than the relevant highest month’s total confimred flu cases over those 7 months. That figure of 1.4 occurred in November 2018, and it didn’t drop back below that figure until August.

The figure of 7 months comes from November to May. I stopped at May because June is when cases usually start to rise above normal baseline and into the foothills of what will become the season’s annual peak. You can see all this below

When I say that Australia had an “early” flu season, in 2019, I’m talking about 8 months early! That’s based on the number of laboratory-confirmed total flu cases each month being at least 1.4 times higher than they were in any previous same month’s total, dating back to 2000. If you want to be generous (and this is already a pretty harsh way to calculate this), you could choose 3 times or more and that would give you season that started 3 months early. This all relies on June being the month when we usually see flu cases start to rise towards their annual peak; the foothills. Data from NNDSS

What’s happening in the United States?

Like Australia, the US is a big wide land of variability in climate, population density and travel activity. See the influenza-like illness map (this is not confimred flu, so grain-of-salt time) below. It shows how ILIs (whatever they’re caused by) have a range of activity levels across the US right now. They are firing up in the south.

The most recent US Centers for Disease Control and Prevention (CDC) weekly flu report provides data for Week 46 (week ending 16NOV2019). I’ve been tracking the clinical laboratory test data this year-all my graphs will use these numbers.

From those maps above we can see that some states have already become well-engaged by flu whereas have yet to see a lot of consistent cases. Overall, it is still very early in the season but it is in the season as the percentage of ILI cases among people visiting a health care provider (at 2.5%) just crossed the 2.4% baseline.

Which flu to rule them all – so far?

Keep in mind that the flu season’s ILI levels – at a national level in Week 45 – are still in the throes of kicking off. Only 4 of 10 regions are at or slightly above baseline, the rest remain below baseline.[2]

FluB rules the roost in Week 46 among clinical laboratory data, United States. Data: CDC weekly reports.[2]

In Week 46 – and for 14 of the last 15 weeks – FluB viruses have been dominating the US although not always by a large margin. For the past 5 weeks, their dominance has been growing. In particular those of the B/Victoria lineage, clade V1A.3. This is the same lineage and clade that dominated Australia’s FluB viruses in 2019. But overall, A/H3N2 viruses were our season‘s main virus, not FluB.

Recent US flu history

I threw together data lifted from the CDC website to show recent years of lab-confirmed flu data.[2]

I’ve blocked out – just by eye – the area which encompasses the peaks of flu activity (blue) from the past 8 years. The peak and foothills are quite broad compared to Australia’s.

Weekly clinical laboratory-confirmed influenza case graphs, United States. The lead-in, peak and lead-out period of the season between 2011/12 and 2018/19 is boxed in blue. Data: CDC weekly reports.[2]

Currently, there is no sign of a very early start to the US flu season.

While the number of US positives are elevated compared to this stage for previous years, this is happening at a time when other years have also started to climb towards their peaks (e.g. 2018/19, 2017/18, 2014/15, 2012/13).

There just is no comparison to the 3 or 8-month early Australian 2019 #flunami here.

But let’s zoom in on that a little more to make sure.

Weekly clinical laboratory-confirmed influenza case graphs, United States but zoomed into just Week 32-49. Confirmed Flu positives in 2019 have at times sat above the same Week’s totals for any of the previous 7 years, and they sit there currently.
2019/20-red boxes; 2018/19-yellow diamonds; 2017/18-blue triangles.
Data: CDC weekly reports.[2]

Yep. We made sure. The US season – from a countrywide perspective – is taking off at the same time as previous recent seasons have.

Not early, but more positives

While not earlier than some recent seasons, the US 2019/20 season, from Week44 to Week46, has seen as many or more flu positives among the clinical lab data as in any of the past 7 seasons.

If we compare the weeks so far (divide the total confirmed flu cases by the total for the same week of each of the past 7 seasons) we get a table of ratios (below).

  • Week 44’s total was equal in size to the total flu cases in Week 44 of the 2017/18 and 2014/15 seasons. But 2019/20 had more cases confirmed (ratio >1) than in any of the other 5 years compared. This was the first week in which 2019/20 season had a bigger total than any of the same weeks in a previous season back to 2012/13
  • Week 45 was at least 1.17-times bigger than 2014/15, and more than that for any of the other 6 seasons
  • Week 46 was at least 1.26-times bigger than 2014/15, and more than that for any of the other 6 seasons

More positives or more testing?

Is this pattern just due to more testing? This remains a big question – especially for some – around Australia’s 2019 #flunami.

Using the same approach we can say that yes, there has been more testing going on in the 2019/20 season so far than in the previous 7 seasons.

And yet, during Wk44-46, it isn’t as much more (1.00-1.18-times more than in the 2018/19 or 2017/18 season) as we saw above for the flu-positive ratios.

Crudely, this might suggest that there is just more flu around than is normal for this time of year.

So perhaps not an earlier start than in recent US history, but it does look like a bigger start. And remember, I’m talking about things at a national perspective. Jurisdictional experiences will vary in big countries where the “flu season” is really more of an overlapping group of separate epidemics.

Not the same as the Australian season

So far – and of course things might change – the 2019/20 US season is neither starting much earlier than normal nor is it dominated by the same viruses as we saw in the 2019 Australian season. These are the data we have right now.

Once again, Flu in the south does not reliably predict that in the north.

So please read this, and also read this earlier post, and realise that what happens with the southern hemisphere’s flu is not a reliable predictor of the start time, length, size or virus makeup of the following northern season.

Every country has its own flu ecosystem and you’re just going to have to make sure you have good surveillance in place if you want to understand what your season will most likely be like. Sorry but crystal balls belong in fantasy.


  1. National Notifiable Diseases Surveillance System (NNDSS)
  2. Weekly U.S. Influenza Surveillance Report

Views: 1619

1 thought on “An early look at influenza in the US highlights that the south does not predict the north”

  1. A bit of clarity is welcomed and you provided us well with your piece. US is seeing also in recent week a upsurge in H1pdm09 isolations, so things are just starting, we will see. Demographics might play a role in this, in a complex play of Original Antigenic Sin, age, co-morbidity, societal status, health insurance coverage and so on. Early reports here in Italy points also toward an interesting season: the elderly and the 0-5, 5-14 classes of age just started to be infected with flu, almost a month earlier that previous seasons. Weather conditions may be part of the scenario, since the wet climate now and a extremely hot summer prepared for good the human airways to inflammations and infections (to not say the droplets floating all around). Recently, another instance of vaccination campaign derailment has been attempted, with the circulation of rumors about a contamination of vaccines vials with latex microparticles. The rumors started at high level (from a source into Drugs Safety Agency), then dismissed by the same organ and by the Ministry o Health, but perhaps the damage was done. In 2015/16, one of the worst season experienced in this country since the pandemic seasons, similar rumors caused a strong decrease in vaccines intake among the high risk population with the result of a loss of life-expectancy (higher than all other OECD Members States). The rumors helped local health agencies to reduce the amount of doses ordered for the next seasons, and to shorten the vaccination campaign duration. This caused a plummeting coverage, now at around 50% or less in >65 years old population and toward coverage in all others.

Comments are closed.