More on chicken soup ...

More on the bizarre statements from the Canadian Public Health Officer David Butler-Jones:

"'In a pandemic, we would see potentially 50,000 deaths in Canada. We see approximately that number of deaths in every year related to tobacco.'

A new pandemic likely wouldn't resemble the famous killer flu of 1918-1919 whose victims were often in the prime of youth, he said. More likely it would look like the epidemics of the 1950s and '60s.

That could mean having 30-40 per cent of the population sick during flu season, compared with the normal level of about five per cent.

'The individual risk - in other words my risk as a 51-year-old asthmatic during a pandemic if I get sick - my risk of dying is the same as my risk of dying every single year during influenza season."

What he is essentially saying is that any mortality risk from a pandemic avian influenza would be entirely compensatory. In other words, he's assuming that any increase in the average mortality rate across the Canadian population resulting from a pandemic influenza strain would be compensated entirely by a decrease in mortality rates from other causes. For example, assume any given person has a 5% of dying in a year (or 5% of the population will die in a year) - this is the baseline mortality rate. Then assume the mortality rate associated with a pandemic influenza is 1% (1% of the population will die of influenza). What Butler-Jones is essentially saying in this news report (assuming he was quoted correctly) is that the base-line mortality rate will go down in response to an increase in mortality from influenza, leaving the overall mortality rate essentially the same.

The simplist way to describe this assumption is that he is assuming that whoever dies of a pandemic strain, would have died from another cause. Alternatively, each death from pandemic influenza allows an equal number of people to survive. This assumption justifies this fantastic statement:

"The individual risk - in other words my risk as a 51-year-old asthmatic during a pandemic if I get sick - my risk of dying is the same as my risk of dying every single year during influenza season."


This is hogwash. All three pandemics of the 20th Century caused increases in mortality rates (e.g., see here), and there's no reason to think the next will be any different.

I can play with some numbers, but I hope no-one who reads this takes these seriously as they are totally back of the envelope.

If you take the mid-point of Butler-Jones estimate of 30-40% of the population getting pandemic influenza (35%) and multiply that by the current Canadian population of 31,946,000, possibly 11,181,100 people could become infected (this ignores age-related variation in susceptibility and regional variation in transmission rates so is certainly wrong). No-one knows how many people who gets this hypothetical pandemic influenza will die, but the Spanish Flu killed about 5% of the people who became infected. If that was the case, that would be over 500,000 deaths. In perspective, in the last year, only 233,087 people died in Canada.

This means that there isn't enough baseline mortality in Canada to compensate for an increase in 500,000 deaths!