Health Canada smoking statistics misleading?

A month or so back (September 24, 2008), I noted that Health Canada was reporting a significant decrease in smoking related deaths on their website. Smoking related deaths were reduced by roughly 21% (10,000 deaths) annually with the stroke of a pen. Health Canada was apparently responding to criticism that CPS-II data used to generate their previous estimates was “not generalizable to the entire US population.” They also observed that “direct application of a large US survey to the Canadian population may not be appropriate.”

But, last week, I noticed the following claim on their website.

“Overview of Health Risks of Smoking
More than 37,000 people will die prematurely this year in Canada due to tobacco use. Unless they quit, up to half of all smokers will die from their smoking, most of them before their 70th birthday and only after years of suffering a reduced quality of life.”

I found it a little strange that all 37,000 deaths allegedly due to smoking were “premature”. So I did a few rough calculations using the StatCan mortality tables from 2002 on which the latest Health Canada study was based. Then, I sent an e-mail to Health Canada which read in part: “I found these claims to be somewhat misleading. After a few very rough calculations, it’s quite obvious that well over 25% of these deaths occurred in the segment of the population over 80 years of age. I hardly think these deaths, which occur beyond average life expectancy, qualify as “premature”.

In addition, the claim that “most” smokers will die before their 70th birthday also tends to mislead. A quick review of the data reveals that roughly half of smoking attributable deaths from all causes occur after the age of 70, and does not support this claim."

I received a computer generated response the same day which read: “Your message has been received by the Web site administrator and is being forwarded to a subject-matter expert for consideration and a timely response.”

That was on Friday, October 17 and I haven’t yet heard from their “subject-matter expert.”

But, over the last week, I’ve had a chance to look a little closer at the data provided in the most recent study. In fact, over 31% of the smoking related deaths estimated by Health Canada occurred in the segment of the population over 80 years of age. It’s difficult to understand how Health Canada, or anyone else for that matter, could suggest that these individuals died prematurely.

For example, of the 40,000 plus Canadians who died from IHD (Ischemic Heart Disease) in 2002, over 21,000 deaths occurred after the age of 80. Yet, only the 1,400 deaths attributed to smoking were considered premature. Given that almost 20,000 non-smokers in that same age bracket died of the same disease, presumably as a result of old age, isn’t it just as likely that the smoking attributable deaths also occurred as a result of old age?

Attributing these deaths to smoking defies common sense and appears to be little more than a blatantly dishonest attempt to inflate the number of deaths associated with smoking for propaganda purposes. Claiming that deaths that occur at 80, 85 or 90 plus years of age are premature is simply bizarre.

The claim that most smokers will die before they reach the age of 70 is also suspect. Although I used a figure of 50% in my e-mail to Health Canada, a closer investigation of the data shows that over 57% of deaths attributed to smoking in 2002 occurred after the age of 70. The claim that most smokers will die before that age is simply not supported by the data provided by StatCan.

Deliberate distortion of the statistics is to be expected from the fanatical members of the anti-smoker brigade. But this type of propaganda should not be condoned from a public health body such as Health Canada.

And, while we’re on the subject of inaccurate claims, I should point out that the claim that 90% of lung cancer deaths in Canada are a result of smoking also appears to be inaccurate. The correct figure is 78% (13,401 of 17,188) for 2002. But, what’s 10 or 12 percentage points here or there?

It’s not as if anyone is going to hold them accountable for the misleading information they provide to the public.