twitter: @rolygate

R West - Ecig Fact and Faction - Update

Update Sept 9th
The previous blog post outlined Prof West's estimate of smoking vs vaping mortality risk. The numbers are for the UK but could be extrapolated for other locations:

There were some woolly edges in this, which invited comment from the ecig community and some of his peers. Prof West has now clarified his statement regarding the number of smokers who die from smoking and the number of vapers who might possibly die from vaping:

A quote from the update article, referring to UK numbers:

At current smoking prevalence of around 19%, the 9 million smokers will correspond to a long term death toll of 60,000 per year, or about 6,670 premature death per year for every million smokers. Even if e-cigarettes were to carry a significant risk of death, say 1/20th that of cigarettes at the upper end, this would reduce to 330 premature deaths per year for every million smokers saving more than 6,000 lives. This is a conservative estimate, as Professor West explains: “It's very conservative because the main toxins from cigarette smoke are either completely absent or in much lower concentrations than 1/20th.”
- R West, J Brown

To make these points clearer, what he is saying is:
- Smoking deaths will reduce anyway by 40% after the timelag required, from 100,000 per year today (the figure usually quoted for an estimate of smoking-related mortality, in the UK, in 2014, by public health sources) to 60,000 per year, which is 6,670 per million smokers per year (this is 0.67% smoker mortality per annum - the percentage of all smokers who die in any given year).

- This 40% reduction in the smoking death toll will (presumably) be due either to the reduction of the number of smokers to 9 million from a figure previously 40% higher, or West's opinion that the currently-popular figure of 100,000 deaths per year is too high; or a combination of the two.

- And then, further to that, for every million smokers who switch to ecigs, more than 6,000 deaths will be avoided.

- For every million vapers, there will be at most 330 deaths. West states this is likely to be an upper figure as there are few if any toxic materials of significant quantity in ecig aerosol. In fact it looks a very conservative figure indeed (i.e. higher than some others might give), as some other form/s of fatal disease/s would be needed to reach even this scale of mortality given the absence of high-quantity high-potency toxins.

Total switchover

Therefore if there were to eventually be 10 million UK vapers, there would be at most 3,330 deaths per year (cf 100,000 a year, the most common figure quoted for smoking currently; thus about 3.3% of the usually-quoted smoking mortality rate for 2014).

This is a useful figure for comparisons because we can see what West's estimate of the scale of mortality is for UK all-smokers or all-vapers; and we know from Brad Rodu that there are about 8 million US ST users, for which it may be possible to estimate an equivalent figure for comparison and which would be interesting as the group size is similar; and we can compare it with the mortality rate per million within Snus users in Sweden - my fag-packet calc is 1.6 million Snusers (Swedish gen pop = 10m and ~20% of adults are Snusers).

The mortality rates  for all these low-risk THR products are probably impossible to calculate accurately since:
- they are all going to be below reliably statistically-identifiable levels (1% or so);
- the effect of confounders will be greater for these very low numbers;
- low mortality rates are going to be intrinsically poorly-identified and reported;
- stats in many countries are political property and may be 'massaged' before release.

West's figure for ecig-related mortality is about 0.03% per annum, which is obviously far too small to identify by any statistical method, and could only be seen by very careful collation of individual death reports (the equivalent figure for UK smokers is about 0.75% per annum, if current numbers are given credence). Given this is his estimate of the highest likely number, it seems clear that ecig-related mortality is not, in his view, a big issue; compared of course to smoking, as all comparisons or caveats related to ecigs must be. You can either have smokers, or ST users, or vapers. You can't have zero - that is a fantasy.

Note about theoretical mortality rates
West's figure for possible ecig-related mortality is for a theoretical 'vacuum' situation where people have never smoked and where there is no timescale - in other words it applies after 20 or 30 years vaping, to never-smokers.

Note about timescales
When discussing smoking-related mortality it is necessary to point out that this must apply after about 25 years or so. Smokers who die prematurely as a result of smoking do so after 20, 30 or 40 years' smoking [1]. So a timepoint has to be chosen where you are going to apply any nominal effect in a statement such as, "smoking kills x% of smokers". The minimum realistic timepoint would be at least 25 years down the line.

I'm only pointing these things out because of the application to vaping. Because West states that vaping might possibly result in the premature deaths of up to 330 people per million per year, it needs to be pointed out that this is (a) a 'vacuum' theoretical situation, where no one ever smoked and there are no other factors affecting the mortality rate [2], and (b) after 25 years or so when possible diseases resulting from vaping have had a reasonable chance to operate. Vaping is not going to kill large numbers of people in the near future: after 8 years there has not been a single confirmed death.

[1] Doll showed that - statistically, that is to say, on average - smokers lose 10 years of life; and that smokers who quit by the age of 35 will - statistically, that is to say, on average - lose none of those 10 years (if they quit by 35 then on average they suffer no lifespan loss). This raises all sorts of interesting questions; and Doll is regarded as an unimpeachable source.

[2] As an example of this, oropharyngeal cancers (mouth cancers) associated with smoking are believed to be accelerated by drinking, and especially by heavy drinking. So, other factors can operate to promote / mitigate disease effects.