One of the most contentious statistics among the many contentious Covid-19 statistics, is the Infection Fatality Proportion (IFP). This is the proportion of people infected with Covid-19 who will die from the infection.
It is probably the most important statistic we need to know about because it determines how long it will take for a population to achieve herd immunity in the absence of a vaccine, and how many people are likely to die in the process.
• Coronavirus in Sweden: What life is like in a country without lockdown laws
• Premium - Cecilia Robinson: The truth about Sweden and its Covid 19 coronavirus choices
• Covid 19 coronavirus: Swedish expert says NZ faces years of quarantine for arrivals
• Covid 19 coronavirus: Sweden records its deadliest month in 27 years after shunning lockdown
Herd immunity is the nirvana in an epidemic and occurs when the proportion of a population who are immune to future infection reaches the point where the new infection can't spread any further and dies out. For Covid-19, about 60 out of every 100 people in a population need to have been infected or vaccinated to achieve herd immunity.
At any point in time, the IFP enables us to estimate the actual number of people in a population who have already been infected and therefore how close a population is to achieving herd immunity.
Unfortunately, the reported number of Covid-19 infected cases in most countries misses so many cases that it cannot be used for this purpose. While it is known that the IFP increases significantly with increasing age, estimates of the average IFP vary widely. Some estimates have been as low as 1 death for every 1000 infections, which is the approximate IFP for the flu, whereas others have been more than 1 death for every 100 infections.
A true value of 1 death per 1000 infections has radically different implications to a true value of 1 death per 100 infections.
The most recent evidence, based on just reported large-scale antibody studies in France and Spain, suggests that the average IFP for Covid-19 is about 1 death per 100 infections, which is at the high end of previous estimates.
As most countries already report the number of Covid-19-related deaths per 1 million people, if you assume the IFP is 1 death per 100 infections, you just multiply the number of Covid-19 deaths per million people by 100 to estimate the approximate number of Covid-19 infections per million people.
Spain, for example, currently has about 600 Covid-19 deaths per million people, so multiplying this death rate by 100 would suggest that there have already been about 60,000 infections per million people (or 6 infections among every 100 people).
This is a much lower number of people infected than many expected in a country that has already suffered terribly.
Unfortunately it means that herd immunity is a very long way off for Spain and there will be many more deaths unless the virus is eliminated.
Unlike almost any other high-income country, Sweden has deliberately chosen a relatively low-key approach to dealing with the pandemic. The level of restrictions has been sufficient to flatten the infection curve, so that health services have not been overwhelmed as they have in Spain and elsewhere, but not sufficient to stop the infection from slowly spreading through the population.
The only logical reason Sweden would deliberately choose this approach is that it must be aiming for herd immunity, although it sits on the fence when challenged about this.
Moreover, if it is aiming for herd immunity, it must believe that the IFP is closer to 1 death per 1000 infections than 1 death per 100 infections. Otherwise the costs, as outlined below, would be way too high.
Almost 4000 people have died from Covid-19 in Sweden so far, which has a population of 10 million (equivalent to 2000 deaths in New Zealand rather than 21). So the current death rate in Sweden is just under 400 deaths per 1 million people. If the true IFP was about 1 death per 1000 infections, then about 40 out of every 100 Swedes would have already been infected; it would be well on the way to the 60 out of 100 people required to achieve herd immunity, and could expect approximately 2000 more deaths in the process.
However, if the true IFP is about 1 death in every 100 infections, which is the current best estimate, the implications are hugely different. Instead of 40 out of every 100 people already being infected, it would only be about 4 in 100 people. This would mean that at the current rate of infection, it would take more than two years for Sweden to reach herd immunity and an additional 56,000 lives could be lost in the process, unless a vaccine or
effective treatments are developed.
This additional 56,000 deaths is not picked out of the air, but can be easily calculated. If 60 out of 100 people need to be infected to achieve herd immunity, then 6 million of the 10 million people in Sweden will need to be infected. If the IFP is 1 death per 100 infections, then 1 in every 100 of these 6 million infected people (i.e. 60,000 people) will die. As 4000 have already died, that's another 56,000 people who could still die.
If New Zealand, with half the population of Sweden, had taken a Swedish-style approach to Covid-19, as has been suggested by some, just divide the Swedish numbers by about two to work out the implications.
As stated above, instead of 21 deaths there would already have been 2000 deaths. Of
even greater concern, this would possibly rise to 30,000 before herd immunity would be achieved, sometime in 2022.
To add to this depressing scenario, despite relatively low level restrictions, the Swedish economy is expected to suffer almost as badly as its European neighbours in the short term and could be worse if ongoing restrictions are required to keep the spread of infection at the current rate.
For Sweden's sake, let's hope the true IFP is much less than 1 death per 100 infections. If not, it will have made a fatal mistake - and it will be a lot worse if infection does not ensure immunity.
After completing this article, a new study has reported that the proportion of people in Stockholm with antibodies to Covid-19 is only 7.3 per 100 people, despite a much higher death rate than the rest of Sweden.
This is unfortunately consistent with the worst-case scenario estimates presented here. The conclusion should now read: "Sweden has made a fatal mistake".
In light of this new evidence, Aotearoa New Zealand has clearly taken the only sensible route in the absence of a vaccine or effective treatment. Australia and other countries need to refocus efforts on the same elimination strategy.
• Rod Jackson is a Professor of Epidemiology with the School of Population Health at the University of Auckland.