Australia's coronavirus death rate is half South Korea's death rate, despite both countries gaining fast control of the pandemic, a puzzle that researchers suggest might relate to different strains of the virus or perhaps antibiotic resistance.
Other possibilities might be health hangovers from the Korean War among people in their 70s and 80s, or differences in hospital treatment.
Australian National University infectious disease expert Peter Collignon and Melbourne economist and statistician John Beggs have co-authored a paper, yet to be peer-reviewed, on the differences between the countries.
They chose South Korea for the similarity in its pandemic success to Australia. Both had extensive testing and tracking programs, neither had had any significant stress on their hospitals, and both had relatively few ongoing new cases, they said.
"This means the data from these two countries are close to a complete picture of a first wave of the epidemic," he said.
But Australia's death rate was 1.4 per cent of confirmed cases and South Korea's was 2.4 per cent. Taking age differences into account, if Australia had South Korea's death rate, 2.6 per cent of Australians with coronavirus would have died, close to twice the South Korean number.
That would have seen 265 deaths by May 9, instead of the 97 recorded.
"Since both countries have advanced high-quality health systems, and since the epidemic in both countries was quickly contained, and each country's hospital infrastructure was not overwhelmed, it remains unclear what accounts for this difference in fatality rates," Professor Collignon and Dr Beggs comment.
Australia's death rate in the first few weeks of the pandemic, at 3 per cent was much higher than it became later, an outcome he puts down to missing cases and an older cohort in the early days. Professor Collignon estimates that Australia has missed roughly 20 per cent of cases, which would put the true total at more than 9000, rather than the 7641 cases confirmed by Saturday.
It would also put the true death rate at about 1 per cent. By comparison the flu has a known case death rate of 0.1 per cent, with the real death rate likely to be a lot lower, making COVID-19 well over 10 times more deadly than the flu.
He suggests South Korea's death rate could be partly explained by the season - South Korea was in winter - which makes him concerned for what the next few months will bring in Australia.
Viruses and bacteria such as pneumococcus might spread more easily in winter and cause more lung complications. In the Spanish flu epidemics, death rates varied depending on secondary infections and immunity to bacteria and viruses.
In South Korea also, people in their 70s and 80s might have been carrying health problems related to nutrition and other effects of the Korean War.
South Korea had higher rates of antibiotic resistance than Australia, so if people were dying from secondary bacteria infections that could also explain the difference.
The authors also speculated that differences in the strains of the virus could impact mortality. Korea's big initial outbreak was from virus strains from China, whereas Australia had a larger proportion from the United States and Europe.
"This might be of significance if strains have changed in their pathogenicity in different regions with time," they write.
But Professor Collignon said it was suggested as a possibility only, with no convincing evidence yet about different mortality from different strains.
In Australia, the youngest death was a 42-year-old male Filipino cruise ship crew member who died in Perth. Two other deaths have been recorded in people in their 50s.
Professor Collignon says the death rate for people in their 50s in Australia is two in 1000 (0.2 per cent). South Korea's death rate in the age group is four times higher, at eight in 1000.
Among people in their 60s, Australia's death rate is seven in 1000 (0.7 per cent), compared with South Korea's 27 in 1000 (or 2.7 per cent).
The numbers jump dramatically from the age of 70 upwards, with 4 per cent of people in their 70s dying in Australia and 11 per cent in South Korea.
Over 80, COVID-19 kills 20 per cent or one in five of its victims in Australia, and 25 per cent, or one in four, in South Korea.
In Singapore, the death rate overall is just 0.1 per cent. Professor Collignon said that was mainly because most of its cases were in relatively young migrant workers.
The death rate in Italy is 14 per cent and Britain 14.3 per cent.
"We have seen a major difference in the mortality rates when comparing Korea to Australia both in the disease hazard curve and also in the simple age adjusted fatality rates," Professor Collignon and Dr Beggs comment in their paper.
"It will be important to better explore difference in the virus strains, variations in carriage of bacteria (e.g. pneumococcus) and/or how healthcare is delivered to try and unravel what are the most important factors."