February 2022

The Nation Reviewed

Australia’s world-leading COVID guidelines

By Elizabeth Finkel
Illustration by Jeff Fisher
When it came to setting guidelines for clinical treatment during the pandemic, Australia got it right

With Australia running at more than 100,000 COVID-19 infections per day, rising numbers of deaths, and people waiting days for a PCR test result or desperately hunting for rapid antigen tests, it’s hard to feel upbeat about the national management of the pandemic.

Commentators hold state and federal governments responsible for trading public health for political gain. But public health experts are themselves in open discord over how to deal with the latest wave.

It’s a far cry from what we’ve seen with the medical advice on how to treat sick COVID patients. Clinicians are just as fractious by nature, yet, throughout the uncertainties of the pandemic, they’ve spoken with a single voice, setting weekly guidelines and updating them as more data comes in.

Australia got it right clinically when many other countries didn’t.

In March 2020, for instance, the United States authorised the emergency use of hydroxychloroquine, a drug commonly used for autoimmune disease, to treat COVID. Two months later, Peru placed hydroxychloroquine and the worming drug ivermectin on its national guidelines, with the army leading the rollout of both medications. Peru ended up with the highest COVID-19 death rate in the world. Patricia García, a former health minister of Peru, suspects these drugs are part of the explanation. Indeed, hydroxychloroquine was later shown to be harmful and the inappropriate use of ivermectin put many people in hospital.

It’s hard to be critical of countries recommending unproven treatments in the face of mounting deaths. It would take months for properly controlled studies to tune the signal from the noise.

So how did we get it so right?

Australia was fortunate to have in place one of the world’s most advanced tuning systems – the National COVID-19 Clinical Evidence Taskforce. Most of us have never heard of it, but it has served the nation extremely well. Its guidelines to clinicians covered everything from the use of oxygen for hospitalised patients to emphatic recommendations against the use of hydroxychloroquine and ivermectin, other than within a clinical trial. Our vaccine guidelines committee – the Australian Technical Advisory Group on Immunisation – used a somewhat similar process to distil fast-changing evidence on vaccines.

The Clinical Evidence Taskforce was also world leading. The World Health Organization and the United Kingdom’s medical guidelines institute, the National Institute for Health and Care Excellence, took a leaf from the Australian model. “Australia’s taskforce was ahead of the game”, says Justine Karpusheff, who leads the development of COVID-19 guidelines at NICE.

The taskforce employs a streamlined process for harvesting and grading medical evidence. Panels of experts then debate results and a guideline is delivered.

Traditionally, reviews of a medical treatment – so-called systematic reviews – take one to two years, while new medical guidelines typically appear every five years. Reviewing hundreds of studies is anything but straightforward because the results may differ and reviewers need to weigh the reliability of the studies. (In the case of ivermectin, several studies turned out to be fraudulent.)

Australia’s taskforce compressed that entire process into a week. Moreover, each guideline was delivered with 100 per cent consensus from panels composed of a dozen feisty clinicians – a remarkable achievement in itself.

There’s no doubt that necessity played a role. “There was a sense of a call to arms,” says Rhiannon Tate, a one-time intensive-care nurse and principal adviser to the taskforce on strategy. But the fact the taskforce had state-of-the-art systems ready to roll out is thanks to the efforts of a frustrated HIV clinician.

Back in 2002, Julian Elliott had relocated from working with Central Australian Aboriginal communities to Cambodia, then the epicentre of Asia’s HIV epidemic. Elliott treated HIV patients and advised the Cambodian government on a national treatment program. But whether it was for treating patients or policy advice, he struggled to find the information he needed.

It was the conundrum of 21st-century medical research. Since the 1980s, the flow of studies had turned from a trickle to a flood, but the processes for distilling the findings had not kept up. The result was a yawning gap between the latest evidence and medical practice.

In 2006, Elliott returned to Melbourne’s Alfred hospital to carry out trials for new HIV treatments. In his spare time, he turned his mind to narrowing the evidence gap. Rather than leaving new data buried in a flood of information, he conceived of “living evidence”, a system that would continuously update evidence as new research was published.

He developed techniques incrementally and then, in 2014, together with trauma surgeon Russell Gruen, he founded the not-for-profit company Covidence – a name that captured the notion of collaborating on evidence. (Six years later the name would take on a new and entirely coincidental meaning.) Covidence also attracted a heavyweight partner: Cochrane, the UK-based institute that wrote the rule book for how to conduct rigorous and transparent systematic reviews. One of the first medical bodies to road-test living evidence was the Australian Stroke Foundation. By 2019, they had telescoped the updating of guidelines from seven years to three months.

COVID-19 reached Australian shores in late January 2020 and within weeks, like everywhere else, there was an explosion of sometimes contradictory medical guidelines. “We thought the last thing clinicians need is a cacophony,” says Elliott. In mid March, he spoke to Anne Kelso, the head of the National Health and Medical Research Council. With government support, as well as philanthropic funding, he established the National COVID-19 Clinical Evidence Taskforce, with a remit to issue a single set of national medical guidelines and update them every week. The first, on April 1, advised on the use of oxygen in critically ill patients.

Initially, Elliott had no idea how the taskforce he headed would meet the weekly deadline. “It was like flying the plane while building it.” After a crazy two weeks of late-night calls, Rhiannon Tate ultimately mobilised 200 of Australia’s elite clinicians drawn from 32 national peak bodies representing everyone from nurses to infectious disease and intensive care physicians.

Then they had to “build the plane”. What they built bears more resemblance to farming than flight.

First, all the relevant information on a particular clinical question needs to be harvested. For instance, all the clinical trials that dealt with the effects of administering ivermectin to COVID patients as compared to an equivalent set of patients, randomly assigned to a control group, who did not receive it. This is known as a randomised controlled trial, RCT.

The next step is to sort the grain from the chaff. Fewer than 1 per cent of the thousands of studies harvested typically end up being relevant.

Winnowing typically requires at least two people to independently read through the studies to decide which are relevant to a particular clinical question and reconcile their disagreements – a process that takes some 300 hours. But Elliott employed crowdsourcing and also collaborated with machine-learning expert James Thomas at University College London to teach computers to do the winnowing, cutting the average time to as little as 100 hours. Some of this work had been done by Cochrane with the sorted RCTs deposited in their database, so the taskforce could shortcut their process.

Next comes a process akin to grading the quality of grain. In the case of RCTs, quality is all about eliminating different types of subtle bias that can skew the result. For instance, were the researchers careful to ensure that the process to sort patients into treated and untreated groups was truly random? Were they “blind” to which group had received the treatment versus a placebo? Did they conceal any of the results?

Finally, the quality-tested papers need to be baked into a guideline that clinicians and decision-makers can use. Here, Elliott collaborated with Per Vandvik, a professor of medicine at the University of Oslo and the lead developer of another platform known as MAGICapp.

MAGICapp facilitates the writing of guidelines, laying out the steps for reviewers to decide whether the recommendation to treat or not to treat is strong, medium or weak. Crucially, by clicking buttons, an inquisitive doctor can look back at all the steps that led to the recommendation.

Still, setting guidelines is hardly a matter of clicking buttons. “This is where cat-herding comes in,” admits Tate. Panels comprising around 12 experts in a particular field argued over where to place the guidelines and their strength. “Fierce debates occurred and were encouraged.” Resolving them relied on thoughtful software and skilful humans.

MAGICapp structured the debate by asking the reviewers to consider particular criteria. Was the effect size large and consistent? Was it seen across different populations? Was there a high risk of bias? Meanwhile, taskforce epidemiologists Tari Turner and Britta Tendal Jeppesen facilitated the discussions to ensure every panellist’s viewpoint was fully heard.

Besides the speed and reliability it offers, living evidence can also save money by avoiding needless duplication by different groups of researchers. The evidence base, after all, is the same.

A proof of concept is that the UK’s NICE initially used the evidence reviews performed by Australia’s taskforce to develop its COVID therapeutic guidelines. Later, it reciprocated with patient management guidelines for treating “long COVID”. That collaboration has now expanded to include Canada and France. “The model of the Australia–UK relationship has inspired and led to an international effort to share evidence, particularly around long COVID,” Karpusheff says.

In mid December, Elliott took his mission to modernise knowledge to researchers across the world. Together with international colleagues from fields as diverse as climate policy to maternal health, they published a manifesto in Nature calling on “researchers in all scientific fields and their funders to test the living evidence model”.

So, could Australia’s fractious public health experts take a leaf from the living evidence model when they give advice to governments?

There’s no doubt it’s a more challenging space. As former deputy chief medical officer Nick Coatsworth put it, “Once you get into the realm of public health policy, it’s so inherently social and political, it starts to defy the model Elliott set up.”

But Elliott is working on it. He’s part of the Global Commission on Evidence to Address Societal Challenges, co-led by public health expert John Lavis, from Canada’s McMaster University. Its report, released in late January, seeks to capture the lessons of the pandemic.

Lavis says that decision-makers “can either build on what works in using evidence to inform decision-making, or carry on with what doesn’t”.

Elizabeth Finkel

Melbourne-based Elizabeth Finkel is an erstwhile biochemist who switched to telling the stories of other scientists. She is the former editor of Cosmos magazine.  

From the front page

Composite image showing John Hughes (image via Giramondo Publishing) and the cover of his novel The Dogs (Upswell Publishing)

A dog’s breakfast

Notes on John Hughes’s plagiarism scandal

Image of Erin Doherty as Becky Green in Chloe. Image supplied

App trap: ‘Chloe’

‘Sex Education’ writer Alice Seabright’s new psychological thriller probing social media leads this month’s streaming highlights

Pablo Picasso, Figures by the sea (Figures au bord de la mer), January 12, 1931, oil on canvas, 130.0 × 195.0 cm, Musée national Picasso-Paris. © Succession Picasso/Copyright Agency, 2022. Photo: © RMN - Grand Palais - Mathieu Rabeau

‘The Picasso Century’ at the NGV

The NGV’s exhibition offers a fascinating history of the avant-garde across the Spanish artist’s lifetime

Cover image of Paul Dalla Rosa’s ‘An Exciting and Vivid Inner Life’

‘An Exciting and Vivid Inner Life’

Alienations and fantasies of escape unify the stories in Australian author Paul Dalla Rosa’s debut collection

In This Issue

Image of Anthony Albanese

The repair man

Anthony Albanese and the task at hand

Image of woman on train platform

The first victim

Does the legal system put people through unnecessary trauma?

Illustration by Jeff Fisher

Roosting

The trials and transcendence of packing and unpacking in order to move house

Image of René Block installing 7000 Oaks on behalf of Joseph Beuys as part of the 1984 Biennale of Sydney

From little things: 23rd Biennale of Sydney

The story behind the 40-year-old tree centring ‘rīvus’, José Roca’s Biennale


More in The Nation Reviewed

Illustration by Jeff Fisher

Red rover

The Australian astrobiologist leading NASA’s search for life on Mars

Illustration by Jeff Fisher

Where did all the bogongs go?

The drastic decline of the bogong moth could have disastrous ecological consequences

Illustration by Jeff Fisher

OnlyFans and the adults in the room

The emerging OnlyFans community offering training and support to adult-content creators

Illustration by Jeff Fisher

ANAM Set and music in lockdown

The project that commissioned 67 Australian composers to write for each of Australian National Academy of Music’s musicians in lockdown


Online exclusives

Composite image showing John Hughes (image via Giramondo Publishing) and the cover of his novel The Dogs (Upswell Publishing)

A dog’s breakfast

Notes on John Hughes’s plagiarism scandal

Image of Erin Doherty as Becky Green in Chloe. Image supplied

App trap: ‘Chloe’

‘Sex Education’ writer Alice Seabright’s new psychological thriller probing social media leads this month’s streaming highlights

Pablo Picasso, Figures by the sea (Figures au bord de la mer), January 12, 1931, oil on canvas, 130.0 × 195.0 cm, Musée national Picasso-Paris. © Succession Picasso/Copyright Agency, 2022. Photo: © RMN - Grand Palais - Mathieu Rabeau

‘The Picasso Century’ at the NGV

The NGV’s exhibition offers a fascinating history of the avant-garde across the Spanish artist’s lifetime

Composite image of Sydney Morning Herald editor Bevan Shields (image SMH/supplied) and actor Rebel Wilson (AP Photo/Chris Pizzello)

Two sides of the same Shields?

Editor Bevan Shields’ attempts to handle the backlash over his masthead’s treatment of Rebel Wilson points to the dismal and fragile state of news media