How many pieces would be enough to guess the full picture and how many could you cope without?

In a way, this puzzle analogy works to help explain what we know about the true spread of coronavirus.

Because we only have data from people who are tested, we’re working with an incomplete picture.

There are many people who are infected with COVID-19 but never tested — they are known as undocumented cases and contribute to a silent, stealthy spread of the virus.

How would you react if you knew there were many more infections in Victoria than we knew about, say a daily rate of 900 rather than 300?

The consequences of these undocumented cases are serious. We can’t accurately predict the spread of the virus or the true infection fatality rate, that is the chances of dying compared to the total number infected.

So can we take an educated guess at how much coronavirus is really out there?

11 times more

Dr Nir Menachemi, who is Professor of Health Policy and Management at Indiana University in the US, tried to do just that.

He and a team of researchers tested a random sample of Indiana residents for active COVID-19 infection as well as the presence of COVID-19 antibodies.

More than 3600 people over the age of 12 were tested between April 25-29.

From the results, Menachemi concluded that the true number of infections in Indiana was 11 times greater than the number of known cases at the time.

Menachemi said he wasn’t surprised many cases were going undetected.

“Almost everything we know about the disease comes from people at the top of the pyramid, the people who have died, the people who have been hospitalised, or the people who are relatively symptomatic,” he said.

“We really just never knew the size of the pyramid, [or] the ratios of each of the groups to each other.

“[But] it’s the people who are out in the community with little or no symptoms that are probably responsible for most of the new infections.”

The researchers concluded that the infection fatality ratio in the state was therefore 0.58 per cent, which is within the range of current estimates (between 0.26 per cent to 0.65 per cent according to the US Centre for Disease Control).

In Australia, there’s less puzzle pieces missing

Australia has some of the best testing rates in the world. Our World In Data shows Australia tests an average of 110 people per 1,000.

Ahead of us are the United Arab Emirates (390 for every 1000), Bahrain (360), Denmark (200) and Russia (150).

“This gives us a reasonable level of confidence that we are detecting most of the COVID-19 cases across the country,” says Dr Adam Kamradt-Scott, who specialises in pandemic preparedness at the University of Sydney.

Epidemiologist John Kaldor from UNSW, who is currently working to find the ratio of hidden infections in Australia, agrees we have a fuller picture than other countries.

“We pride ourselves on very high levels of testing so we would expect a lower level of undetection in Australian than some overseas settings,” Professor Kaldor said.

“The speculation has been one or two additional cases for every detected case.”

Kaldor and his team are analysing blood samples taken for other reasons to look for antibodies and estimate how many undocumented cases we’ve missed.

The serosurveys (blood testing for antibodies) will first be conducted on samples from the Red Cross in NSW, with a focus on Sydney, before being rolled out nationally.

The testing is anonymous so the person’s identity is never disclosed, just their age, sex and postcode.

“The serological findings will give us an insight into how we can enhance our testing strategies to get better coverage of cases in the community,” Kaldor said.

“It’s very important to know not just how much we’re missing but where we are missing cases; whether it’s in certain age groups or postcodes.”

Missing asymptomatic cases

The majority of undocumented cases will be among infected people who have no symptoms or mild symptoms they don’t take seriously.

The science is not yet definitive on the proportion of COVID-19 cases that are asymptomatic, but the estimates range from around 18 per cent to 80 per cent.

Back to that research out of Indiana, Dr Menachemi says 44 per cent of those who were infected had no symptoms during the two weeks before testing.

Professor Raina MacIntyre, head of the biosecurity program at the Kirby Institute, says the substantial body of evidence tells us that asymptomatic and presymptomatic infection is common with COVID-19.

“Studies of the Diamond Princess found about two-thirds of passengers were infected, with a high proportion asymptomatic. Studies in aged care and other outbreaks have also found 50 per cent or more of all positive cases are asymptomatic.”

Silent carriers are thought to be largely responsible for the rapid spread of COVID-19 in China.

Professor Jeffrey Shaman from the Columbia University analysed the data and found 86 per cent of all infections were undocumented prior to January 23.

Even though these undocumented infections were half as contagious, they were the source of two-thirds of known infections.

“For every case you find, you should multiply it by 5 to 10…even 12 or 15 to get the total number of infections,” Shaman said.

“[But in Victoria] the multiplier would be a little lower, maybe one in five cases are possible rather than one in 10.”

Professor MacIntyre says the World Health Organisation has been slow and reluctant to accept the prevalence of silent transmission.

“Until testing of every at-risk person during outbreaks is conducted, asymptomatic cases will be missed and outbreaks will occur.”

Victoria’s position remains that people without symptoms should not be tested unless they have special circumstances or are requested to do so by contact tracers.

“If a person without symptoms is tested and the result is negative, it does not mean that they have not been infected, as they might still be incubating the virus,” the state health department says.

– ABC / Paige Cockburn