During COVID. many epidemiologists found themselves explaining complex, uncertain data in real time - often without any training in communication. I wrote about what it was actually like: medium.com/@adrian.esterman/…

Bird flu update, SA: H5N1 now confirmed in 5 endangered Australian sea lions (Kangaroo Is. + the Pages), plus a fox in Adelaide. SA accounts for 318 of 583 national events. No poultry cases; human risk low. See sick/dead wildlife? Don't touch – call 1800 675 888.

Here are the current COVID notifications for South Australia for 2026. Relatively quiet at the moment. Note that we did not get the Winter (June) peak this year compared to 2025.

theguardian.com/us-news/2026… FDA recommends that updated vaccines target the XFG (Stratus) variant. According to @Mike_Honey_ , XFG is currently accounting for only 10% of sequences in Australia, so interesting to see what TGA do.

In just one week, the avian flu virus in Australia has gone from birds only to a seal, a fox and now a dolphin. That is the same path it took on every other continent. Thankfully, it has still has not reached our poultry farms and there have been no human cases

COVID-19 vaccination is about to undergo a major change in Australia. From 1 October, COVID vaccines move to the National Immunisation Program - but funding will no longer be universal. Free vaccination will be limited to: • people aged 65+ • Aboriginal and Torres Strait Islander people aged 50+ • adults with severe immunocompromise That’s a significant policy shift after more than 76 million doses were administered since 2021 under the universal program. The question is whether this narrower approach strikes the right balance between cost, access and preventing severe COVID-19 disease.

Friday epi explainer Misconception: a new study found it, so it must be true. Correction: if something truly has no effect, roughly 1 in 20 careful, honest studies will still come out positive - purely by chance. That's the threshold we work to. Run enough studies on enough things and flukes are guaranteed. And the fluke is the one that gets written up, because null results rarely make it into journals, let alone the news. That's why we wait for the pattern, not the paper. Evidence not vibes.

Friday Epi Explainer Most people who commit murder probably ate toast for breakfast that morning. Does that mean toast is a risk factor for committing murder? Of course not. Eating toast is common. So is eating cereal, drinking coffee, or brushing your teeth. Simply finding that two things occur together doesn't mean one causes the other. Epidemiologists call this an association. To show causation, we need much stronger evidence. Correlation is where we start. Causation is where we aim. #Epidemiology #PublicHealth #Statistics

Today, Moderna has announced that the first participant has been vaccinated in a Phase 1 clinical trial in Canada evaluating mRNA-1469, their investigational mRNA vaccine against Bundibugyo ebolavirus The study will evaluate the safety, tolerability and immunogenicity of the candidate in approximately 80 healthy adult participants across three Canadian sites. The program is is being advanced through Moderna's expanded strategic collaboration with CEPI.

This time last year, South Australia was climbing into a substantial winter COVID wave - cases peaked near 500/week in June 2025. This year? We're at the lowest point of the year, around 50 cases/week, and still falling. No winter wave. Likely broad hybrid immunity from past infection and vaccination, plus no major new immune-escape variant to reset susceptibility. Worth noting, COVID isn't behaving like a winter virus the way flu and RSV do. Its timing is driven more by new variants and waning immunity than by the season.

Friday Epi explainer A quick guide to reading any "vaccine injury study" you see online: How many people? If it's one patient, it's a case report. It can raise a question - it cannot tell you your risk. "Billions affected" never follows from n=1. Was there a control group? No comparison, no conclusion. Who published it, and where? Check the journal and who funded it. Evidence > vibes.

Duck hunting season is approaching. Worth flagging a risk that gets little attention: hunters handle and dress wild waterfowl, the natural reservoir for avian influenza, and often with bare hands. With H5N1 circulating in wild birds, simple precautions like wearing gloves, hand hygiene, and avoiding sick birds matter.

Nationally there are now 27 confirmed detections of H5N1: 10 in Western Australia, 14 in South Australia, 2 in New South Wales, and 1 in Queensland. A further cluster of suspect greater crested tern cases in SA , seven in the state's South East and four on Kangaroo Island, has been sent to CSIRO for confirmatory testing, and more detections in wildlife are expected. Every case to date is a wild seabird; there are still no detections in poultry, captive birds, pets or livestock, and no human cases.

H5N1 update for Australia🇦🇺 Six weeks after H5N1 avian influenza was first detected on mainland Australia, around 20 infected wild seabirds have now been found across WA, SA, NSW and Queensland. The good news is that there is still no evidence of spread to Australian poultry, and the risk to the general public remains low. Globally, human infections have been rare and have almost always occurred in people with close contact with infected birds or other animals. There is still no sustained human-to-human transmission. This is a situation worth monitoring closely, but not one that should cause undue concern.

The Bundibugyo Ebola outbreak in the DRC is the fastest-growing on record. So far, 2,536 cases, 1,033 deaths (~41% CFR) as of 23 July. No approved vaccine or treatment exists; Only licensed products target Zaire strain. Ervebo may be partially cross-protective against Bundibugyo, and CEPI is fast-tracking three candidates (IAVI, Moderna, Oxford).

Friday's epi in 60 seconds explainer. The most useful diagram in epidemiology. A disease is water in a bathtub. 🚰 Tap = new cases (incidence) 🛁 Water = existing cases (prevalence) 🕳 Drain = recovery or death Prevalence ≈ incidence × duration. Long COVID? The tap has slowed - but the drain is barely trickling. Fewer people falling in, hardly anyone climbing out. That's why prevalence stays high.

The US is at real risk of losing its measles elimination status. Elimination does not mean zero cases, it means no sustained transmission for ≥12 months. With thousands of cases and ongoing outbreaks, the question is whether chains of transmission can be stopped before that threshold is reached. A reminder that vaccination coverage is the foundation of measles elimination.

A reminder that COVID-19 has not disappeared. Taiwan has entered a new COVID-19 epidemic phase, with outpatient/ED visits rising 66.5% in a single week to 4,766. Severe disease remains concentrated among older adults and those with underlying conditions. Surveillance matters - even in the era of endemic COVID-19.

This adds to a growing body of evidence that long COVID is not a single condition but a complex syndrome with multiple manifestations. The finding that women report more severe neurological symptoms and greater impacts on cognition and quality of life is important, and deserves further investigation in larger, diverse populations. Understanding who is most affected is essential for improving diagnosis and care.