(Image: AAP/Bianca De Marchi)

As the spread of the COVID-19 virus wreaks havoc, the question of how Australia is faring compared to other countries — and what might lie ahead — is being much discussed.

But with the numbers changing by the day, and with countries fighting and tracking the disease in different ways, the figures can, at best, only tell us so much.

At worst, the data has the potential to mislead.

On March 30, Prime Minister Scott Morrison suggested the stringent measures that had been put in place to slow the spread were finally starting to work.

“It is clear the epidemiology curve is beginning to flatten,” Morrison said in a news release.

“But it is too early to determine whether such movements will be significant or sustained.”

What can the data usefully tell us? How does Australia compare? And are we starting to flatten the curve? RMIT ABC Fact Check takes a look at the numbers, and the assumptions underpinning them.

What is the COVID-19 curve?

We don’t know the actual number of COVID-19 cases, only the number of cases confirmed by testing. These are two very different things.

Countries are often compared by tracking the spread of the virus from the time of their 100th confirmed case.

This can be useful, because it places different nations at the same starting point, just when the virus is starting to take hold, providing a reference point from which to compare the spread of the disease.

But relying on confirmed numbers, as opposed to actual numbers, can be fraught. More on that later. First: the raw data.

The following chart compares the number of confirmed cases by country. It uses data from the European Centre for Disease Prevention and Control (ECDC), an agency set up by the European Union in 2004 to track and monitor infectious diseases.

In Australia, the number of confirmed infections has steadily risen.

As of April 5, there were 5,687 confirmed cases in Australia.

According to estimates released by Oxford University-based organisation Our World In Data, the number of cases in Australia is now doubling every 10 days.

This compares to a doubling every six days in the US, where confirmed cases have risen to over 369,000 (April 7).

The number of infections has risen more slowly elsewhere.

In Japan, for example, there are 3,654 confirmed cases, with infections currently doubling every 8 days, while in Singapore there are just 1,375 confirmed cases, with the number doubling every 9 days.

Infections as a proportion of population

To state the obvious, some countries have much bigger populations. The raw numbers tell us little about the proportional impact.

Combining the ECDC figures for the number of confirmed cases with population figures published by the United Nations, it is possible to track the number of confirmed cases per million people, by country.

On this measure, Spain has been proportionately hit hardest, with 2,668 cases per million people, followed by Switzerland (2,367 cases per million people) and Italy (2,061 cases per million people), as at April 5.

In the US, there are about 943 cases per million, while in the UK the figure is 617 cases per million.

The Scandinavian countries also have high numbers in proportional terms, with 1,016 cases per million people in Norway, 704 in Denmark and 638 in Sweden.

Australia currently has recorded about 223 cases per million people.

By international standards, that’s comparatively low, but still higher than China (57 cases per million), Singapore (203) and Japan (26).

What about deaths?

Another way to analyse the impact of the virus is to compare mortality, by country. Again, there are several ways to carve up the figures.

Below is a graph comparing the total number of confirmed deaths by country.

By this measure, Italy has been the hardest hit country with 16,523 confirmed deaths, followed by Spain, the US, France, the UK, Iran and China.

In Australia, the number of deaths has been comparatively low, with 43 fatalities recorded (as at April 7).

What about the rate? This is where things get more complicated. There are a number of ways to analyse the death rate.

The first is to look at the number of confirmed deaths, expressed as a proportion of population.

On this basis, Australia is still — at this early stage — ahead of many other countries, with about 1.3 deaths per million people.

That compares to 254 confirmed deaths per million people in Italy, 251 in Spain and 116 in France. The US has recorded 26 deaths per million, compared to 2.3 per million reported by China.

There is a second type of death rate. This is sometimes referred to as the “case fatality rate”.

This tells us the number of people who are dying as a proportion of confirmed cases.

This measure has received a lot of attention, particularly in the media.

Case fatality rates appear to vary considerably between countries.

For example, in Australia, at this stage, the case fatality would appear to be relatively low, at 0.6% (April 5).

This compares to a case fatality rate as high as 12.3% in Italy, 9.4% (Spain), 10.3% (UK), 4.0% (China) and 2.7% (US).

There has been much speculation as to why case mortality rates differ so markedly between countries.

Why is the case mortality rate, for example, relatively low in Germany (1.5%) and Norway (0.9%), but 5.8% in Sweden and 11.0% in France?

What we don’t know about mortality and infection rates

The reality is that, as a measure, the case mortality rate should be treated with an abundance of caution.

Crucially, the case fatality rate is not the same as an actual mortality rate.

It can only tell us the number of deaths as a proportion of confirmed cases, and at a particular point in time.

First, the number of actual cases may differ wildly to the number of confirmed cases, particularly if testing rates are low.

As the World Health Organisation defines it, a “confirmed case” represents a person with laboratory confirmation of COVID-19 infection, irrespective of clinical signs and symptoms.

A recent article published in the science journal Nature cites estimates suggesting that asymptomatic or mild cases could represent about 40% to 50% of all infections.

The point is, at this stage we simply don’t know how many people are infected with the virus.

So, countries with limited levels of testing might appear to have higher mortality rates — particularly if that limited testing is being restricted to those patients with more severe symptoms.

Data relating to testing can be sketchy and difficult to compare. But what is clear is that test rates between countries can vary dramatically.

At this stage, Australia appears to have one of the highest testing rates in the world.

According to the Department of Health, Australia has now conducted more than 268,000 tests as at April 1.

That implies a cumulative testing rate of more than 10,551 tests per million. According to other figures complied by Our World in Data, the testing rate in the US is now 3,345 per million, the UK (2,126), Italy (8,573) and South Korea (7,971).

Although these figures are not directly comparable due to different testing methodologies, the data shows how widely testing varies between countries.

In an article in The Conversation Sanjaya Senanayake, an associate professor of medicine and infectious diseases physician at the Australian National University, says variations in case fatality rates between countries “may partially be explained by whether hospitals have been overwhelmed or not”.

Even with relatively high testing rates, in countries severely affected by the crisis, such as Italy, it is probable that testing is confirming the more severe cases while failing to identify less severe ones.

This might explain Italy’s alarming high case fatality rate (currently 11.7%).

There is a second important reason why comparing case fatality rates could be fraught.

Case fatality rates tend to change over time, often dramatically. So, taking a single rate, at a single point in time, and using it as a basis to compare countries or to estimate the likelihood of dying can be misleading.

Different countries are at different stages in terms of the progression of the virus, as the following chart demonstrates.

One recent study published by medical journal The Lancet, estimated it takes an average of 17.8 days from when a person first experiences symptoms of COVID-19 to when they die.

The World Health Organisation’s Report of the WHO-China Joint Mission on Coronavirus Disease 2019 (COVID-19) found in fatal cases it took patients between two and eight weeks to die.

The implication is that many people who are currently recorded as infected will eventually show up in the fatality statistics. This could cause the case fatality rate to rise over time.

As another article published in the American Journal of Epidemiology put it: “Simple estimates of the case fatality ratio … can be misleading if, at the time of analysis, the outcome is unknown for a nonnegligible proportion of patients.”

This is what happened during the 2003 severe acute respiratory syndrome (SARS) outbreak.

As has been pointed out, during the early stage of the SARS outbreak, the case fatality rate was estimated at 3% to 5%. By the end, the estimates had been revised up to 9.6%.

During the SARS epidemic, the rise in the case fatality rate initially led some people to wrongly conclude that the disease had evolved and was becoming more deadly over time.

But as medical studies later established, this was not the case. Rather, it simply took time for some people to die.

The same could conceivably be true of the COVID-19 pandemic.

The bottom line

In proportional terms, Australia is, at this stage, being hit less severely by the virus than many other countries.

And, as Morrison has noted, there are some early promising signs that the daily rate of increase is slowing, though the total numbers are still rising and the situation is changing rapidly.

In reality, relying on figures such as confirmed case numbers and case fatality rates is likely to paint a misleading picture.

Australia and the rest of the world won’t be able to properly assess the disease until we have a sound understanding of the final numbers of deaths and actual infections.

That will require both time — potentially months, even years — and extensive research.

Principal researcher: Josh Gordon, economics and finance editor

[email protected]