Tuesday
Apr072009
The politics of secondhand smoke
Tuesday, April 7, 2009 
I have been reading the comments (101 so far) on last Friday's post. I have nothing to add except to refer readers, especially our friend Rollo Tommasi, to the 2005 Forest report Prejudice and Propaganda: The Truth About Passive Smoking.
The 50pp document, which features a lively introduction by Lord Harris of High Cross, our late chairman, is divided into three sections: Questions and Answers about ETS, Epidemiology and ETS, and Examining the Evidence on ETS.
Nothing that has been said or written since prompts me to change my mind that the dangers of secondhand smoke have been greatly exaggerated for reasons that have far more to do with politics than health.
You can download the full report HERE.
Reader Comments (156)
Some of the comment about this article will centre on the guideline that relative risks of less than 2 shouldn't carry much weight. The anti-tobacco lobby counters this by saying that it applies only to individual studies, and not meta-analyses.
It should be noted that in an ideal but unachievable scenario, a meta-analysis is, statistically speaking, one (large) test, so, in general it would seem that the same principle should apply. For example, if you suspect a coin is biased and toss it 10 times today and 10 ten times tomorrow, conduct two statistical tests and then pool them, you will arrive at the test based on tossing the coin 20 times in one go. Conducting a meta-analysis results in a smaller confidence interval, which may not then include 1.0, but it doesn't eliminate bias. Depending on the individual tests, various confounding biases will possibly remain and a bias running through all or most of the tests will certainly still show up. If there are any professional statisticians out there, please comment on the general problem of pooling confounding biases - and please, not by quoting some tobacco control paper spcific to this situation. There is a long list of potential confounders, many with relative risks far higher than ETS.
The report doesn't mention publication bias - where studies showing no effect aren't published. There has been work done on this - some funded by the tobacco industry and a paper by John Copas and his research student, funded by the EPSRC and published in the BMJ, predictably attracting interest in the response section. I attended a seminar by Copas and I don't recall him slipping outside afterwards for a smoke. The conclusions estimate that the claimed relative risk should be reduced by about a third. This isn't surprising, given the nature of medical statistics and academia; and the most prominent anti-tobacco activist, Stanton Glantz, actually admits to it. He has been quoted as saying words to the effect that "if it helps us, we use it: if it doesn't, we don't".
Perhaps sensing that possibly increasing ones lifetime chance of dying from lung cancer from 3 in a thousand to 4 in a thousand by living with a smoker for 20 years wasn't going to cut the mustard with Joe Public when arguing for total smoking bans, including private smoking clubs, the anti-smokers turned their attention to heart attacks. To me this is is even less convincing because of the association, through social class, of passive smoking with so many other possible risk factors for heart disease.
Although the statistical questions are interesting, those who really understand them (and I'm not including myself) and who have no agenda, tend not to get involved, so I think you'll have to settle for another 101 rounds with the mysterious Rollo Tommasi. My bet is that he is American (explains the smoking obsession and American phrases), living in Edinburgh, possibly something to do with formulating policy in some kind of think tank. His expertise seems more in politics and propaganda rather than medicine or statistics. I am intrigued as to why he needs to remain anonymous, but I'm sure we'll eventually find out.
My argument against total smoking bans isn't statistical; it is that adults, once informed of a risk, should be allowed to themselves analyse the costs and benefits. I wouldn't ride horses or become a scaffolder.
Finally, Richard Doll was not misquoted and he was not speaking as a 90 year old to whom it didn't much matter. I listened to that Desert Island Discs. Sue Lawley was surprised at what he said. I think she asked him twice. I thought at the time we hadn't heard the last it.
The problem with people like Rollo is that they fail to understand that the studies are junk. Really.
Then they further fail to understand that if you mash 8 junk studies together to produce a meta-analysis, you haven't just produced more junk, you have made the original junk worse.
A car analogy, if you will indulge me.
You begin with a car that won't start. But you tell everyone that the car is fine. No-one is really convinced. So, after a while, you find four other cars that seem okay but they aren't running perfectly. So you take the gearbox from one, the exhaust from another, the pistons from a third and the battery from a fourth and make a new car. The new car performs extremely badly but you now put this car on the centre of your forecourt and advertise it as in "Excellent Condition". It looks good, it is nice and shiny and clean, but once you take it for a test drive, you notice that the gears make a hellacious noise, it sounds like a tank, the pistons are knocking like mad and most of the electrics are iffy.
But, you paid top mechanics to fix this wreck, and you really, really don't want to admit that you got it badly wrong, so you change the rules to suit. "Noisy cars that fail to start now and then are fine. Really. Trust me".
I always found it odd that other chemicals/substances with much higher Relative Risks are simply ignored. Second hand smoke, the zealots say, is more toxic than Sarin Gas, Zyklon B and nuclear waste, combined.
Hard cheese, whole milk, and olive oil ALL have higher RR's for lung cancer than SHS.
When will we see those things banned?
Simon - Thank you for at least responding even if, as I expected, you have shown you are only interested in criticising the use of extreme claims when they are made by one side in this argument.
That Forest report, which you hold so dear, is a complete work of fiction. Let’s look at some of the techniques the Forest boys and girls used.
1. Take an issue which potentially has some relevance (in this case the risk of misclassification and the existence of confounders). Then exaggerate the likelihood and effect of this risk beyond all reasonable levels, without any supporting evidence.
2. Take a very selective look at evidence and don’t be afraid to use completely outdated material if it helps your case. So, for instance, spend all sorts of time complaining about the 1992 EPA report. But completely ignore the fact that much more recent overall studies have been undertaken but you don’t know how to criticise the substance of these (which take account of a lot more evidence) – such as Law et al (1998), Hackshaw et al (1998), IARC (2002) and SCOTH (2004).
3. Apply your own completely distorted version of scientific reasoning, and never mind that no serious professional scientist would back up what you have to say. So for instance, argue that the test of passive smoking depends on how many individual studies show a statistically significant increased risk. Okay, no professional journal will tell you that this approach reflects solid science, but that doesn’t matter, does it?
4. Take some perfectly reasonable quotes by distinguished professionals (preferably off-the-cuff and unscripted words used in interviews, which are unlikely to be carefully phrased, or old statements which are difficult to access and therefore hard to prove or disprove). Remove unhelpful parts of their statements to make it seem that they are actually saying something else. So, when Sir Richard Doll was saying that he was no longer worried about the effects of passive smoking on his health given he was already 89 years old, make it appear as if he was suggesting that nobody should be concerned by passive smoking. And when people say that studies which show new or surprising findings should not be views as conclusive unless they show a relative risk of 2 or more, make it seem as if they were actually saying that NO study is valid unless it demonstrates such a high relative risk. Okay, it’s downright dishonest – but you’re trying to sell an argument here, aren’t you? Why let accuracy get in your way??
In short, the Forest report may appeal to the suckers who have allowed themselves to fall unquestioningly for deceit sponsored by Big Tobacco. We’ve seen plenty of them on your blog. All very keen to show how much they disagree with what I have to say. But very few of their posts have attempted to challenge on any basis of evidence. To them, pro-smoking arguments based on homespun theories and personal attacks on people with whom they disagree trump hard evidence. What a strange perspective that is.
But it is no surprise that your arguments fell on such deaf ears among people who actually know about the subject. And, if that is the quality of argument which Forest continues to display, it is hardly likely to win many future political debates either.
Of course, the irony is that you have just applauded someone else (Michael Siegel) for condemning the tobacco control movement for using extreme claims which they cannot justify. What a pity that you do not have the decency and courage to challenge deceit and distortions from your own side.
Rollo a nice try again, the IARC study is in fact the 1998 WHO Boffetta study which I shall quote to you again.
"“Conclusions: Our results indicate no association between childhood exposure to ETS and lung cancer risk.” And a comment by Dr John Dale Dunn “(Note: a relative risk of 1.16 is not proof of anything. It must be 2.0 or more, the equivalent of a 200 % increase in risk. This study shows less than 20% increase in risk, not valid in an uncontrolled population study."
SCOTH we have covered before as politically motivated rubbish and admitted the RR "did not reach statistical significance."
Hackshaw's study is deeply flawed. It uses Confidence Intervals of 90% when the gold standard is 95%.
"This bias estimate is little affected by using an alter native method for misclassification adjustment or by using updated data from 47 studies rather than data from the 37 studies used by Hackshaw et al. The bias is increased if strong evidence of much higher misclassifi cation rates in Asian women is taken into account and could then explain about half the observed association. Misclassification correction has not previously been at tempted for dose-response data."
http://ibe.sagepub.com/cgi/content/abstract/10/6/384
'... you are only interested in criticising the use of extreme claims when they are made by one side in this argument.'
Haven't noticed any criticism from you of the pro-ban studies Rollo. Is that because they never do anything wrong?
So, when Sir Richard Doll was saying that he was no longer worried about the effects of passive smoking on his health given he was already 89 years old, make it appear as if he was suggesting that nobody should be concerned by passive smoking.
That wasn't what Doll said. But then, with the transcript of the programme no longer available, But reports of it are. The Times:
The crusade against passive smoking is one area where their assertions often seem to run ahead of the evidence. Yet those questioning the orthodoxy are likely to find themselves out in the cold. Indeed some frowned upon Doll because for years he seemed relatively unimpressed by the epidemiological evidence on the effects of passive smoking, telling Sue Lawley on Desert Island Discs that “the effects of other people smoking in my presence are so small that it doesn’t worry me”. .
University of Ottawa:
He told the presenter Sue Lawley something that came as a surprise to those who knew him well. He said the effect of someone lighting up a cigarette in his presence "is so small that it doesn't worry me", a comment which some interpreted as a denial of the impact of passive smoking.
Nothing there about being "89 years old". Nothing about "no longer". Forest did not misquote him. You have.
Rollo Tommasi wrote: In short, the Forest report may appeal to the suckers who have allowed themselves to fall unquestioningly for deceit sponsored by Big Tobacco. We’ve seen plenty of them on your blog. All very keen to show how much they disagree with what I have to say. But very few of their posts have attempted to challenge on any basis of evidence.
Would you care to point to a bit of deceit sponsored by Big Tobacco. I haven't heard from them for years. But perhaps you are better informed, and will have the relevant links at hand.
And I have challenged the entire validity of epidemiological studies which can't even measure the ETS levels which are the object of their concern. That's like having studies of road traffic that can't measure vehicle speeds. And that's not science. And you have not answered me.
in short, the Forest report may appeal to the suckers who have allowed themselves to fall unquestioningly for deceit sponsored by Big Tobacco. We’ve seen plenty of them on your blog. All very keen to show how much they disagree with what I have to say. But very few of their posts have attempted to challenge on any basis of evidence. To them, pro-smoking arguments based on homespun theories and personal attacks on people with whom they disagree trump hard evidence. What a strange perspective that is.
Rollo seems to forget that some of us are not interested in the evidence, I am well aware of the risk to my own health through smoking, less conviced of the shs argument, totally dismiss the 3rd hand theory.
But be that as it may, I have been de-normalised by a society that cares nothing for my own rights and wellbeing. (See, The social impact of the smoking ban - on the Forest website).
All I am asking for is CHOICE, for smokers and non-smokers, Personal Attacks Rollo? I have had my personal lifestyle ripped apart by zealouts like you. If that is not an attack on me then I do not know what is, so forgive me for being angry - I don't have to 'trump' you with hard evidence, The evidence of what the ban has done to my life is enough!
Amen to that GraemeJ
No surprise that my comments attracted some responses. People trying to tell me that me whole argument is wrong, even though very few have actually tried to challenge my arguments.
As for the few attempts to challenge what I said, they don’t really add up to much.
DaveA: Sorry mate, but I don’t accept your points. There is no reason why Forest relied on a 1992 report, when more recent and comprehensive alternatives had been published. You have no comments to offer on the Law et al and SCOTH reports at all, while your comments on the Hackshaw et al and IARC reports are well off the mark.
You say “the IARC study is in fact the 1998 WHO Boffetta study”. Where on earth did you pick that up from?! Have you actually read the IARC monograph? Have a look at it: http://monographs.iarc.fr/ENG/Monographs/vol83/index.php. It is NOTHING LIKE the Boffetta study: http://jnci.oxfordjournals.org/cgi/content/short/90/19/1440.
Hackshaw’s work WAS based on a 95% CI. Hence the report states the results showed “ The excess risk of lung cancer was 24% (95% confidence interval 13% to 36%)”. Don’t know where your 90% figure came from. You refer to Lee’s study. He takes a distinctive approach to deciding what amounts to “misclassification” but even his revised calculation made little difference to the overall level of risk. In fact, what difference there is in the figures is largely explained that Lee’s assessment was based on the “fixed effects” model, while Hackshow used the preferred “random effects model” and a more sophisticated sensitivity analysis (http://www.advisorybodies.doh.gov.uk/coc/ETS.PDF).
Idlex claims Forest has not misquoted Sir Richard Doll. He offers quotes which supposedly prove his point. Only they don’t. Mick Hume in The Times just repeats Forest’s accusation, while the Uni of Ottawa quote simply notes that “some interpreted [Doll’s quote] as a denial of the impact of passive smoking.” It does not try to argue that this was the message he was trying to convey.
So what was Doll’s true stance on the impact of passive smoking? I leave the true position to Lord Walton of Detchant (http://www.publications.parliament.uk/pa/ld200506/ldhansrd/vo060619/text/60619-05.htm):
“In Grand Committee, a noble Lord referred to the fact that Sir Richard Doll was quoted some time ago as saying that active smoking was harmful, but that second-hand smoke did not worry him. He said that as a throwaway remark when being interviewed on Radio 4's "Desert Island Discs" in February 2001. I knew Richard Doll very well and, some years ago, I had the privilege of succeeding him as warden of Green College, Oxford. In February 2005, just a month or two before he died, he wrote: "We first established the causal link between smoking and lung cancer in 1950, but the tobacco industry spent decades arguing that our results did not justify our conclusion . . . Now tobacco companies are using the same technique to undermine the conclusion that passive smoking causes fatal disease. The evidence that it does is clear", and incontrovertible. He continued: "As a responsible citizen, I believe that nobody should have to work in an atmosphere polluted by other people's smoke".”
So there you have it. In Doll’s own words: "As a responsible citizen, I believe that nobody should have to work in an atmosphere polluted by other people's smoke".”
As for the other examples of misquoting in the Forest paper, I see nobody has sought to refute that charge. DaveA did quote John Dale Dunn and, to be fair, he does appear to believe that a relative risk of 2.0 or more is required in any study. But his professional colleagues – even in the American Council for Science and Health, where he works – profoundly disagree. Elizabeth Whelan tore to shreds an article by Robert Levy and Rosalind Marimont which sought to make the same claim as Dunn - http://www.acsh.org/publications/pubID.498/pub_detail.asp.
She said: “Relative risks less than 2.0, while small, can indeed be statistically significant and reflect a causal relationship. Given the pervasiveness of a risk factor, such as smoking, and the prevalence of some of the diseases it causes, small relative risks can, and do, represent serious threats to public health. ….Levy and Marimont's assumptions regarding small relative risks VIOLATE BASIC PRINCIPLES OF EPIDEMIOLOGY" [my emphasis].
Both Levy and Marimont and this Forces paper misquoted the National Cancer Institute. Here is what Elizabth Whelan had to say about that: “The authors mislead readers by misrepresenting a quotation from the National Cancer Institute, which qualifies relative risks, as the agency's "own guideline." In fact, the NCI has no such guideline about relative risks, and the quotation cited is taken from a 1994 NCI press release on abortion and the risk of breast cancer. Given its proper context, this so-called guideline makes a much different point from what the authors suggest.”
So yet another damning indictment of the type of cynical misquoting in which this Forest paper engaged.
Belinda:
If Simon Clark wants to gloat about Michael Siegel criticising the tobacco control movement, he should have the decency and consistency to attack extreme claims, whichever side makes them.
As for me, I have stated on several occasions that I am not persuaded by the evidence about thirdhand smoke or the acute effects of changes in exposure to secondhand smoke (e.g. the Pell study). I've still to meet a pro-smoker who is as prepared to challenge and criticise the evidence on both sides.
Sorry if that makes me harder to paint as an anti-smoking zealot, though it hasn't stopped some of your cabal trying anyway.
Cut the crap Rollo.
In answer to your much earlier comment that I haven't read any of the ETS studies conclusions, you couldn't be more wrong. It is the conclusions from these trash studies that have led me to see the ridiculous claims of the anti-smoking crusade in all their glory.
Not long ago I read a study that claimed eating more than two portions of fries a week would lead to an increased risk of breast cancer. It claimed this with a familiar methodology, and with again the same 'unquantified and retrospective data'.
The glaring fact here was that portions could NOT be measured, and therefore 3 portions for some could well be less than 2 for others, which brings us to the rational conclusion, that the findings could be turned on their head, giving creedence to a protective effect from eating 3 portions a week.
Does this seem familiar? Now where have I seen this moronic garbage masquerading as science before?
Idlex claims Forest has not misquoted Sir Richard Doll. He offers quotes which supposedly prove his point. Only they don’t. Mick Hume in The Times just repeats Forest’s accusation, while the Uni of Ottawa quote simply notes that “some interpreted [Doll’s quote] as a denial of the impact of passive smoking.” It does not try to argue that this was the message he was trying to convey.
The University of Ottawa repeats what the Times wrote that Doll had said, which is the matter at hand. Were they also simply repeating Forest's accusation? What about Tobacco.org?
He quoted only from his figures and was no absolutist. When questioned recently on second-hand smoke, he exasperated the anti-smoking lobby by replying: “The effects of other people smoking in my presence is so small it doesn’t worry me.”
He was also wont to point out that the idea that smokers cost the taxpayer millions in hospital fees is in fact a myth, stating that smoking efficiently killed its adherents before they could retire or become old — and that the habit might actually be of economic benefit to the country. It was, as one journalist pointed out, “an argument that only Richard Doll could get away with airing”.
The same link (Washington Post) also provides the following gem:
"I don't mind in the least if someone in the room lights up a cigarette,'' he told the Journal of Addiction in 1990. "It's their decision and their life, not mine.''
House of Lords Written Evidence:
As recently as February 2001, Prof Sir Richard Doll said: "the effects of other people smoking in my presence are so small that it doesn't worry me."
Included in this are the views of Ernst Wynder, who also published a paper on smoking and lung cancer in 1950.
Dr Ernst Wynder, the first major researcher to link cigarette smoking with lung cancer, said in 1974 that he did not believe that "passive smoking really hurts the health of somebody who sits next to you." Ten years later, in 1984, Dr. Wynder concluded: "Should lawmakers wish to take legislative measures with regard to passive smoking, they will, for the present, not be able to base their efforts on a demonstrated health hazard from passive smoking."
To counter what so many different sources repeat, you find a quote from someone who said he knew Sir Richard Doll, and that Doll had written something somewhere or other (in a private letter or email?) about the dangers of passive smoking. That's hearsay. It's not evidence at all.
The simple truth of the matter is that neither Ernst Wynder nor Sir Richard Doll, who were the twin pillars or post-war smoking epidemiology, believed that passive smoking posed any significant threat. Both said so repeatedly. Doll also said, in respect of the dangers of passive smoking, "the quantitative relationship is very weak."
If he changed his mind, he certainly had not done so by the time of the BBC Radio broadcast in 20001.
Im so sorry i'm going to have to upset your values on debate.
I think we need a comment from someone who is sick fed up of phoney puritanically inspired anti smoking statistics .
Here's my comment , **** ***.
I'm not stupid .
Your the sort of person who moves in next door to a pub and complains about the noise are you not.
You know who, you are ,to quote the charlatans "your a wierdo".
In my eyes thats the way I see it.
In your eyes you probably think your extremes are normal concensus, and everybody want's to hunt down smokers in packs like a screaming mob from a "Frankenstien", film noir,
demanding they be vilified demonised , punished, and "CURED" (REMEMBER crooked evangalists,("HEAL,HEAL,HEAL").
The sad old salvation of the soul thing.
Yes "punished" broken on the wheel because they are so vile and do not "FIT", into your twisted ideal of a happy jogging carrot crunching world
I tollerate "gits" like you.
Tolerate me git.
Then maybe i will forget you exist.
And tollerate you.
All my friends and workmates dont give a toss whever I smoke or not.
Or whether they smoke or not.
Only "nutters" make an issue out of it.
Im so sorry my diagnosis is .......
Your "NUTS".
Happy "nutting", the world would be a far less stressfull place without you.
The world needs Tollerant people now, not gits like you.
It seems studies such as Helena, Pueblo and J Pell came to an incorrect conclusion.
What were the benefits of a smoking ban? Reduction in heart attacks? Other disceases?
A new study (discussed at Dr Siegels) shows that bans do not have any affect on heart attacks or indeed any discease. This is the largest study so far and confirms the Kuneman/McFadden study 5 years ago.
Rollo, have you viewed yet the (not so) astonishing results of the biggest, most comprehensive study to date, by RAND Corp, Stansford University and others, showing finally, definitively, that smoking bans had zero effect on heart attack admissions and mortality trends across the US.
http://www.tobaccoanalysis.blogspot.com/
You be the judge of the results.
We here don't find the results astonishing (they should be unsurprising to the casual researcher), but astonishing for the reason that at last it seems some honest research and proper reporting has taken place.
I don't suppose that the Independent newspaper, that screamed on its front page a few months back, that the smoking ban in the UK had saved 40000 lives (or was it 400000?), will be making a retraction? C'est la vie. We can only calmly and firmly stand up to engage with the occasional reader like Rollo, and try to encourage critical thinking and a dose of common sense to this important issue of smoking and smoking bans.
I understand that you find cigarettes revolting, Rollo, but I confess that I simply do not understand your attitude to smoking. Do I take it that you are involved professionally in the anti-smoking movement by the way? With respect, if you are, I must tell you that I simply do not understand your motives. I mean, if you don't like smoking, why don't you just not smoke and discourage your children from smoking. Isn't that 'job done'? Why the crusade to rid the world of smoking? Do you really think it right to impose your own picture of paradise on everyone else?
Anyway, back to that study...do you think that it is right to continue the push to harden the line on smoking bans (if you personally are) whilst simultaneously being party to the results of world's largest study, which trashes the concept that 'smoking bans save lives'?
Incidentally, I mentioned 'paradise' above. To throw this into the mix...you may as well know that smoking is paradise to me, so the only reason I am engaging in blogs like this is to defend something that I know to be something that gives me a lot of pleasure and creative zest. I feel impelled to stand up for my right, or opportunity, to enjoy life to its fullest. To be frank, I am sick and tired of people, whom I don't know, engineering events such that my enjoyment of life is threatened.
Imagine there was a powerful, government-backed organisation lobbying persistently for your annual holiday abroad to be banned (or for you to 'give them up') for good, whether you liked it or not. If a foreign holiday is not particularly for you, then choose whatever you like - music, wine, cheese, pillows - as an exemplar of something you enjoy. Are you getting to where we are coming from yet? That's why we're here - we suspect that people like you want to remove cigarettes from our lives, and yet we know they're the most enjoyable thing on the planet.
So please just bear it in mind that, far from being desperate, denialist addicts as you would no doubt prefer to think of us here, we tend, on this site, to be articulate, passionate people who happen to have a different viewpoint towards tobacco (one where we happen to think it's brilliant). And we're not too inept at understanding the limitations of epidemiology, nor spotting the widespread use of weasel words in the antismoking propaganda campaigns.
Happy Easter Rollo. I was bound to show up eventually. Forget all the surveys and questionairs, the guestimates and the epidemiologicalisms. Let us look instead at observation and common sense. I have found that just about all my non smoking friends and aquaintances have never really given a second thought to the passive smoking scaremongering, for why? - lack of evidence. They have not experienced or seen other people dropping like flys. This includes friends and relations in the nursing profession.
Here is something for you to work out Rollo. I will give you a bona fide calculation based on statistics from the CRUK website. 0.1% of smokers, I repeat, 0.1% are diagnosed with lung cancer every year. It also informs us that most are over 65. That is not that they are 66 either. It would mean that many are in their 70's and 80's + . So here is your question. Going off that statistic, how many non smokers get lung cancer from passive smoking - and no cheating ie trying to say that a lit cigarette is more dangerous to non smokers, that is not allowed.
Rollo, if you look at the IARC link you provided you can see very clearly that the majority of the evidence they provide for a link between active smoking and lung cancer is from questionnaires. Questionnaires do not represent scientific quality and hardly represents what we are told constantly is indisputable facts that smoking kills.
How very interesting. Several more responses have come through. But of those, only one (Idlex) has tried to defend the Forces submission which Simon Clark holds so dear. Simon – it looks like the Forces document is too radioactive for even your pro-smoking buddies to line themselves alongside.
Idlex’s comments dealt with Sir Richard Doll. Everyone agrees that he said on Desert Island Discs that “the effects of other people smoking in my presence are so small that it doesn’t worry me”. But nobody seems to know what he said immediately before and what he said immediately after. That contextual information is crucial to knowing whether Sir Richard considered passive smoking to be a minor risk to anyone or only to himself. Since it is Forest who are using his quote, the onus is on them to demonstrate that they have faithfully reproduced the message he wanted conveyed then.
Idlex didn’t like the quote I offered him (although he’s quite happy to accept unsourced quotes if he likes what they have to say). Try this sourced quote which Sir Richard gave in 2002 when announcing the findings of the IARC evaluation of the evidence on the links between passive smoking and lung cancer (http://www.dailymail.co.uk/health/article-124168/Passive-killer-proof.html; see also http://www.independent.co.uk/news/science/greater-risks-from-passive-smoking--and-extra-cancers-645917.html).
“Sir Richard Doll, whose research 50 years ago in the UK first linked smoking to lung cancer, and a member of the IARC group, said hundreds of thousands more cases of the disease worldwide could now be attributed to smoking.
'It is the first time there has been a formal evaluation by scientists of the carcinogenic risks of environmental tobacco smoke and it has concluded that it causes lung cancer,' he said.
'Environmental tobacco smoke that people experience at work or at home is definitely a cause of lung cancer.
'That has been discussed for a long time but this is the first time a group of independent scientists has reviewed all the evidence and said that there is no question.'”
So Idlex, you lose either way. Even if Sir Richard had meant what you think he meant in 2001, his views by 2002 were clear. Either way, the Forest report cynically tried to misrepresent Sir Richard’s latest position on passive smoking – and that was downright deceit on their part.
I’m going to be enjoying an Easter break, and may not return to this board, at least not for a few days. If I do find time to return here, I am going to keep to the subject of this post – about the hypocrisy of Simon Clark for not criticising the deceit of the pro-smoking lobby and about the tawdry deceit running throughout the Forest paper which Simon himself referred me to. Don’t expect me to respond to “off-topic” comments.
Rich White claims "Questionnaires do not represent scientific quality". Where on earth did you pick up that ludicrous idea from?
You're doubting the use of questionnaires, rather than epidemiology per se. Well, epidemiology professionals view questionnaires as an entirely valid and indeed important part of their science. There appears to be a lot of work going on in the epidemiological world to ensure the standard of questionnaires is as high as it reasonably can be, e.g.: http://www.dundee.ac.uk/iea/EuroQuests1.htm.
So there you have it. Questionnaires DO feature in studies of scientific quality.
Rollo, that they feature does not mean they should feature. Questionnaires are rife with problems, especially if they are retrospective. No causative link can be made with questionnaires, only statistical. Plus, they are fully dependent on what questions are asked which means the researchers have to include all possible factors AND provide room for elaborate answers. Then this all has to be quantified into some semblance of responsible conclusion.
Please. If smoking was such a deadly killer then raw data would prove it - clear cut, no if's and no but's. But that's not the case is it Rollo? No. With the attack on smoking being what it is, we should expect unequivocal and irrefutable facts, but instead we are left with 'evidence' that is wide open to critique. As Timbone has stated above, even CRUK admit only 0.1% of smokers are diagnosed with lung cancer each year, and most are elderly. Looks like it comes back to smoking being a disease of the elderly....
And, to repeat what Timbone said because you failed to answer it, if the actual stats of smokers themselves developing cancer are so low, how is it that SHS can be an instant killer?
Btw, you mention Siegel - I'd like to point out we don't all love him. I actually criticise him more than once in my book because he used methodologies and tactics that he so openly criticises today. I do not know enough about him personally to make any accusations on a personal level, so I will refrain from doing so. But let me be clear in saying that most of us are well aware of Siegel's tainted hands, so to speak - in fact, when I was first informed of his blog I was highly skeptical of his change of heart. I am still slightly confused by why he used those fundamentalist tactics, developed bogus studies, now attacks anti-smokers, yet still raves on about the negative effects of smoking. I am sure you will say 'because they're true!', but let's not go down that path....
I've been following the posts of Rollo, this anti-smoking junkie, not only on here but other sites, and it's come to the point where I read the first couple of lines and by-pass him, reading his posts is like watching repeats on TV. I bet he has the same reports & studies in front of him every time he switches on his computer.
But it's quite obvious his knowledge of genuine science is abolsutely zero, but that said he can be quite entertaining at times, even to doing a double act with the wife.
Rollo, or the anti smoking propaganda movement are NOT interested in facts & truth, these hold no meaning for the them.
Perhaps the pro-choice movement should commission a couple of dozen eminent scientists to study the brains of 100 anti-smoking zealots, see what makes them tick, see if the brain cell for common sense & truth is damaged, if they're alright they can always be paid to produce junk reports, lie.
Rich - Guess what? Epidemiological studies do have to be conducted with care to deliver scientifically robust results. And they are - not just in relation to passive smoking but in many other areas too - including exhaust emissions, vaccine testing, etc.
As far as I can see, evidence about the dangers of smoking is at least as clear as evidence relating to the dangers of exhaust emissions (other than carbon monoxide) and radon gas.
As for Tim's lung cancer stat, I couldn't find it on the CRUK website. What I did find was that 34,150 people in the UK died of lung cancer in 2006. Total number of annual deaths in E&W was around 500,000, so let's estimate 570,000 total UK deaths. So lung cancer accounted for around 6% of all deaths that year. Hardly insignificant, is it?
Right, I am now finishing. Have a good Easter. We may resume later. Perhaps by then Simon Clark will either find the reasoning to justify the comments in the paper he so warmly endorsed, or have the courage to acknowledge the deceits peppered within it.
Rollo, it's all well and good to have faith in epidemiological studies. Sure, they have the potential to produce good results, but if you've ever actually looked at them you'll see they're riddled with serious, serious flaws. Not just ETS, but active smoking too - Doll, Wynder etc. The 1964 SG Report was slammed by a statistician, as was Doll's work.
Just out of curiousity, have you ever emailed Lauren Colby? I'm sure you must have heard his name flying about so I wonder if you ever spoke to him
Sir Richard Doll was a smoker and lived to 93. OK so he quit at some stage and said that if you quit for five years you will be as healthy as a non-smoker. Is he saying that smoking does not affect your health, providing you quit sometime?
The use of questionaires is valid but only to produce a suggestion. When you have the suggestion you bring hard science to play. The suggestion from epidemiology is that smoking causes lung cancer, so you move on to animal studies to produce evidence of causation rather than correlation, which is what you start with.
That is indeed what they did fifty, sixty years ago. The smoking animals lived longer than the controls, which was hardly the outcome expected. That situation remains to this day, though these days they use genetic variant animals prone to developing cancer and then point to the smoking animals developing larger tumours than their control cousins. The pharmaceutical and tobacco control industries lap that up, seemingly oblivious to the fact that once a tumour has started, it is the body's ability to nourish that tumour that affects the tumour's size. There are clearly elements in tobacco smoke that increase vascularisation rather than what one would imagine which is to constrict blood vessels and reduce vascularisation. Ergo, provided cancer doesn't arise in the first place (which is what the early, healthy animal studies showed) tobacco acted as a beneficial agent.
Studies that show that tobacco smoking leads to an increase in a host of biochemical agents beneficial to health, particularly anti-oxidants such as glutathione reductase (on a longer-term, background basis, rather than those studies that detect reduced anti-oxidants immediately after intake of tobacco), back that hypothesis up. Oh, I wonder why we didn't read about those studies in the newspaper
It's the same thing with blood pressure: smoking a cigarette increases blood pressure we're told, like we should be worried about it, and it is indeed correct that there is an acute increase when one has just smoked but it's the background blood pressure, over time, that counts.
Just look at the life expectancy statistics of nations having at least some parity in genetic stock, wealth, diet, climate and so forth and wonder why those nations that smoke significantly more than other nations don't show the tell-tale signs of the 'decimation' caused by smoking. Note, for example, how the women of Spain, of whom over thirty percent smoke, live a little longer on average than Portuguese women, of whom just 9% smoke. Yes, it's possible that factors unknown account for this, and I'm not going to produce this statistic as proof that 'smoking is good for you'...it's merely 'suggestion'?
Rollo Tommasi wrote: Idlex’s comments dealt with Sir Richard Doll. Everyone agrees that he said on Desert Island Discs that “the effects of other people smoking in my presence are so small that it doesn’t worry me”.
Well, thank goodness for that! At least we can now dispose of that particular issue.
But nobody seems to know what he said immediately before and what he said immediately after.
We don't know because we don't have the transcript of the interview in front of us, because the BBC has removed it. We are instead using reports about that transcript which reproduce just one line from it.
And so I really don't know what he said just before or just after. However, I did read the transcript a couple of years ago, and it seemed quite clear to me that he simply didn't regard passive smoking to be a danger at that time. He did however think that active smoking was dangerous, and caused lung cancer. And I think that was made quite clear in the transcript also.
That contextual information is crucial to knowing whether Sir Richard considered passive smoking to be a minor risk to anyone or only to himself. Since it is Forest who are using his quote, the onus is on them to demonstrate that they have faithfully reproduced the message he wanted conveyed then.
I don't believe for a moment that Doll was suggesting that passive smoking posed no risk to himself alone, as if he had some sort of natural immunity (supposing that any 'immunity' is needed). That would contradict the entire thrust of his life's work, which entailed treating people as being equally at risk regardless of who they were.
And Forest faithfully reproduced what Doll said. It's true they did not reproduce the context in which these words were spoken. How big a context do you want? A lifetime's work in tobacco epidemiology? In Shakespeare's Richard III, the king at one point cries out:
"A horse! A horse! My kingdom for a horse!"
Are we to be forbidden from quoting this line, without setting it in its context? Do we really have to tell the reader that he uttered these words just after he had been defeated at the battle of Bosworth? Do we really have to make it perfectly clear that this was not how he felt about horses most of the time, and that he said this just once and not on many occasions?
You next produce another quotation from Doll dating from 2002, in which he unambiguously asserts that passive smoking causes lung cancer. And then you say,
So Idlex, you lose either way. Even if Sir Richard had meant what you think he meant in 2001, his views by 2002 were clear. Either way, the Forest report cynically tried to misrepresent Sir Richard’s latest position on passive smoking – and that was downright deceit on their part.
I'm quite happy to agree that Sir Richard Doll's views in 2002 were different from his views in 2001. He seems to have changed his mind. He seems to have spent most of his life - the first 90 years or so - believing and saying that passive smoking posed no risk to anyone, and the last 3 years of it saying the opposite. Did he change his mind because he was convinced by the evidence? Or did he change his mind because he was subjected to strong pressures from his medical peers to do so? I really don't know.
How is it best to describe the attitudes of someone who thinks one thing for most of their life, and quite the opposite for the rest of it? Should we say, "Well, for 90 years Sir Richard Doll thought this way, and for the next 3 years he thought that way, so he was 96.7% of the belief that passive smoking was harmless"? Or should we say, "For 90 years Sir Richard Doll thought this way, and then he saw the error of his ways, and began to think that way"? Or might we be also perfectly entitled to say, "For 90 years Sir Richard Doll thought his way, but then he was nobbled by his irate colleagues, and forced to think their way"?
He would not have been the first. Galileo was forced to recant by the Inquisition. So what were Galileo's true views? Did he believe what he had spent most of his life saying, that the Earth went round the Sun, or did he believe what he spent the last few years of his life saying, which was the opposite? Do we have any evidence that the Inquisition really did 'force' him to change his mind? I'm sure that if the Grand Inquisitor, Cardinal Oleo Tommaschi, were to make an appearance on this thread, he would say, "It is entirely false that Galileo was subjected to any form of coercion. We had a most pleasant conversation, during which I was able to demonstrate to him the motion of the Sun using a ball and chain that happened to be hanging on the wall behind me."
In conclusion, I don't see that Forest or Simon Clark have anything to apologise about in respect of Sir Richard Doll. It seems fairly clear that, very late in his life, he changed his mind about passive smoking. Is it unfair to report what he was very publicly saying for most of his life? And given Galileo and many others, should we not take thse sorts of deathbed conversions with a pinch of salt?
I should add that I am not a member of Forest, and I am at times a rather trenchant critic of Simon Clark. Not on this occasion, however.
The use of questionaires is valid but only to produce a suggestion. When you have the suggestion you bring hard science to play. - Adeimantus
Rollo is far more optimistic about questionnaires. Over on the other thread he wrote
Survey recollections. I can’t pretend to assume I’d be able to remember what happened 40 or 50 years ago. But, when it comes to typical daily conditions in the home, at work and in the car, I am confident most people would have reasonable recollections over 5, 10 or even 20 years. They are not going to say precisely how many cigarettes were burned in their presence on each particular day. But questionnaires do not need that level of sophistication..
In all honesty, I have to report that I really have no idea how many cigarettes I smoked yesterday, never mind 20 or 30 years ago. And I know that I don't know, because every now and then I try to keep a tally of them. But I always lose track after a few hours.
Why is that? Because I smoke roll-ups, and I have a rather uncanny ability to roll them automatically, without noticing that I'm doing so. Put a tin of tobacco, and some cigarette papers, and a lighter in front of my hands, and they will in the next instant magically produce a roll-up.
Furthermore, I don't know how I roll them. Over on Forces last year, I set out a long and detailed explanation of how to roll a cigarette. Then I just sat watching my hands as they went through their automatic actions, and realised to my horror that I had described it completely wrongly. I had described what I thought I did, rather than what I actually did.
I have discovered that memory is a strangely illusory thing. I keep a diary every day. And this has provided me with a record of exactly what I was doing over several decades. Last year, thinking back to how life had been in the late 1970s, and wishing to find out what I had been doing then, which I thought I remembered, but wasn't quite sure about, I dug out the relevant diary of that time, and began to read it. And got a most tremendous shock. My life in 1978 had not been AT ALL like I remembered it. Over the years, that tumultuous year had come to seem like it was just like every other year back then. And it was not. I had completely forgotten a whole raft of events. And yet I knew that my diary did not lie, so my memory had to be lying. I had been, in all sincerity, telling myself whoppers.
So, no, I don't think I can remember what happened 10 or 20 years ago. And furthermore I know that I don't know, because I have seen the evidence that proves it. Rollo may have 'confidence' in his memories. I have next to none in mine.
And if I had to fill out one of these questionnaires, these days I feel obliged to write, "I DON'T KNOW".
Great points Idlex. We must also remember that questionnaires are fully dependent on the questions asked, if they're closed-ended then participants are forced to answer one way or the other and it may be entirely different.
There's also recall bias. Suppose I'm 65 and I have lung cancer. I know that I've been smoking sice I was, say, 13, and I know my smoking habits changed. But i won't remember that i smoked just 5 at 13 and was on 30 a day at 25, i also wouldn't know if my habit fluctuated here and there, if i smoked more on nights out or whatever. there are a lot of variables. But, i would be likely to increase the number i thought i smoked if i had cancer because i would think 'well, i have cancer so i must have been a heavy smoker'.
Some years ago I worked in 'market' research as an interviewer. In some studies we used laptops to capture data and, inevitably, sometimes the programme crashed, in which case we'd ask respondents the questions again. I can reliably inform Rollo that within the space of five minutes the same respondent would give a different answer to the same question.
Genuine and honest research companies take great pains to minimise bias. They know only too well that respondents can easily be manipulated by the phrasing of a question and the order in which questions are posed as well as the tools used to measure responses to questions that gauge attitude. In addition there are strict rules for interviewers to minimise their influence (never explaining the meaning of a question, never venturing their opinion, never interviewing people known to them, and so on).
Questionnaires may be used on a stand alone basis when sampling the population for consumer research (eg buying behaviour, response to car ad campaigns). When it comes to more complex issues, they'd be used to highlight areas of interest which would be explored in more depth using qualitative methods. I doubt if any reputable research company would claim that its methodology complied with the rigour demanded of pure science and, as such, its findings and their interpretation would be treated with caution.
As far as I can see, the ETS studies are no different from market research projects in methodology, except I'm not convinced that the ETS studies attempt to minimise bias. They then, quite cynically, interpret their findings in a way that is erroneous, dishonest and designed to mislead.
Quesionnaires do not and never will minimise the bias.
Short scientific studies to prove a point won't either.
The government now has egg on their faces.
Are they so naive that they didn't realise that the truth always comes out? Or were they so naive to be brainwashed by such ridiculous lies?
They have been extremely naive one way or the other and it is their choice to which route they admit to.
I know who I'd blame if I was in their shoes.
The truth will always prevail unless they intend to shut up (by some means) many scientific researchers.
I'm interested in how they intend to actually do that.
I care nothing for Nu-Lab anymore. They dug their grave, they can lie in it.
RichW wrote: We must also remember that questionnaires are fully dependent on the questions asked
Indeed. And there's also questionnaire interpretation errors.
I've spent some while imagining how some doctors might have misinterpreted the only question asked in Doll and Hill's British Doctors Study, when it landed on their desks one morning, given they had no way of clarifying the questions.
...the doctors were asked to classify themselves into one of three groups - namely, (a) whether they were, at that time, smoking; (b) whether they had smoked but had given up; (c) whether they had never smoked regularly (that is, had never smoked as much as one cigarette a day, or its equivalent in pipe tobacco, for as long as one year).
These aren't quite as simple and straightforward questions as they were no doubt intended to be. They are all open to misinterpretation in a variety of ways. The third question is particularly convoluted, and even now that I must've read it several dozen times over the past few years, I still find myself wondering exactly what it means.
One problem of the first question is the ambiguities in "at that time" Does that mean right now, this minute? Or today? Or these days? And one problem with the second question is about "given up". Given up what? And the second question is really two questions; "Have you ever smoked? Have you ever given up smoking?" And the third question is positively labyrinthine. If you smoke 360 cigarettes in a year, does that count as less than one cigarette per day, even if they were all smoked in one week? And what if you never smoke a whole cigarette, but always stub it out halfway through? What is the equivalent to a cigarette in pipe tobacco?
And while the first two ask positive questions - "Do you smoke?" and "Have you given up smoking?" - the third asks a negative question: "Have you never smoked regularly?" The true answer to the first question is "Yes, I smoke", and to the second question is "Yes, I've given up smoking". But the true answer to "Have you never smoked regularly?" would be "No, I never smoked regularly." So the true answers to each question are Yes, Yes, and No. If the doctors had been asked "a) Are you a smoker? b) Are you an ex-smoker? c) Are you a non-smoker?" the true answers to each question would have been Yes, Yes, and Yes, and the set of questions would have been consistent. As it was, the questions asked were not consistent with each other. The third one is the odd one out, by being negative (as well as convoluted).
One can imagine some of the doctors' responses to the questionnaire:
Smoker doctor 1 peers at questionnaire: "Let's see. A, Am I smoking right now? Nope. I stubbed out the last one a couple of minutes ago. B. Have I smoked and given up? Yes, many times. C. Hmmmm... Complicated. Well, I don't smoke a pipe. And I never smoke as much as one cigarette, because I always stub them out half-smoked. So, no. So that's No, Yes, No, So tick B. Now where are my cigarettes?"
Doctor 2 looks at questionnaire: "A. No, I've given up. B. Yes, I've given up. C. No, I've smoked more than that. So tick B. Good thing this questionnaire didn't come yesterday, before I started my latest attempt to give up smoking."
Non-smoking doctor 3: "A. No, I don't smoke. B. No, I've never given up smoking. C. No, I've never smoked even as much as one cigarette ever. Oh dear, that's three Noes! Which one should I tick? Oh, tick B."
Non-smoking doctor 4: "A. No. I don't smoke. B. No, I've never given up smoking. C. No I've never smoked as much as one cigarette a day for as long as one year. Oh wait! I smoke a large Havana cigar every Christmas day. That must be the equivalent of 5 or 6 cigarettes in one day. I'll have to classify myself as a smoker. Tick A."
Cigar-smoking doctor 5 sees nothing about cigars, so ticks C, as he never smokes either cigarettes or pipes.
And so on.
I can't help but think that a fair number of the doctors simply wouldn't have understood the third question, because it was just too complicated, and so relatively few of them would have confidently ticked it. And the result would have been that there would have been that non-smokers were under-reported.
One way of testing out the ambiguity of the questions would be to take a group of people whose smoking habits were already established, and asked them to answer the questionnaire, and then seeing how many answered the questions correctly, given their known smoking status. A well framed set of questions would be expected to produce accurate answers. A badly framed set of questions would have induced interpretation errors, the source of which would perhaps be found by asking the participants why they had given wrong answers.
And the Doll and Hill questionnaire seems to have been a rather poor one, easily misinterpreted.
I must say that the more I think about the Doll and Hill studies (1950, 1954) the worse they seem to be.
The 1954 British Doctors study is generally regarded as the flagship study. It began in 1954, and was still generating results in 2004. But there's a strong case to be made that the earlier 1950 London Hospitals study was at least as influential. It was midway through this earlier study, as the results were still coming in, that Richard Doll gave up smoking.
On the face of it, the results of the first study were pretty damning. Out of 649 lung cancer patients, all but 2 of them were smokers. Is it any wonder that Richard Doll gave up smoking when he saw those sorts of figures emerging? And these figures are the ones that appeared in Sir Richard Doll's obituary 60 years later.
But if you dig a little bit, you find a rather strange thing, which is that most of the control group of patients without lung cancer were also smokers. Only 27 of the 649 patients in the control group were non-smokers. 97.7% of the patients in the entire study were smokers. In the subsequent British Doctors study, 87% of the doctors were smokers. It would seem that more or less everybody smoked back then.
Now, if you have a population in which 98% of people are smokers, then if the null hypothesis applied - that lung cancer was entirely unrelated to smoking -, one would have expected to find that 98% of lung cancer patients would be smokers. Just like if 56% of the population had regularly eaten custard, one would have expected 56% of lung cancer patients to have been regular custard eaters.
As it was, it turned out that 99.7% of lung cancer patients were smokers rather than the expected 98%. And this would appear to be well within the margin of error of the study, which would never produce exactly 98%.
From this perspective, the London Hospital study was entirely inconclusive. At best it showed a possible elevated risk of lung cancer among smokers, which a subsequent study might have elucidated. Sir Ronald Fisher, perhaps the leading statistician of the time, and also a friend of Bradford Hill, used the results from this study to show that inhaling tobacco smoke actually had a protective effect against lung cancer.
So how could Richard Doll have ever jumped to the conclusion that smoking caused lung cancer, given such meagre results? .
The answer perhaps is that Richard Doll was not at that time a mathematician or a statistician. He was a young doctor.who'd just spent the war aboard hospital ships, and who had some mathematical aptitude. Perhaps it isn't really very surprising that he got fooled by the numbers he saw emerging from the study.
He was very much the junior partner to Bradford Hill, who was a considerable statistician. What did Hill make of it all? Did he stop smoking as well? Hill seems to have been largely silent about the controversy.
But Richard Doll seems to have been convinced by this study, and his conviction seems to have been rapidly communicated to the medical profession, including to George Godber, who probably wasn't any sort of mathematician either, yet who was to many years later become the government chief medical officer, and driving force behind the secondhand smoking scare.
I've now begun to wonder whether a gentlemanly argument between two statisticians, Bradford Hill and Ronald Fisher, was being played out in the field of cancer research in the late 1940s. This was the playground upon which their arguments happened to be being conducted. They were both far more interested in statistics than they were in medicine. Richard Doll was just a pawn of Bradford Hill's, employed to gather the data for a new study which had been designed by Hill to press the limits of statistical analysis. I can almost see Bradford Hill meeting up with Ronald Fisher (who was a smoker) at some gentleman's club and saying to him, "Well, Ronald, what's wrong with this new study of mine then, eh? I think I have you outwitted this time, dear chap!"
What neither of them knew, or could possibly have guessed, was that they had set in motion a statistical juggernaut, a 20th century Frankenstein monster, that was going to still be gathering momentum long after they were both dead.
For if one reads Fisher's writings on the matter, they come across as the elegant wit of a considerable mathematician rather than the screams of an embattled smoker.
They were perhaps like those nuclear physicists at Los Alamos, who were more interested in protons and electrons and neutrons than they were in the nuclear weapons that they had accidentally produced, and who wandered off still arguing among themselves about spin-orbit couplings, and strong and weak forces, while their progeny was leading to an arms race that was to terrify the entire world.
Just read "the use and abuse of statistics" by reichmann.
Questionnaires are not a valid method FULL STOP
I AM a mathematician
Idlex, Doll was not a man of his integrity. I suspect he knew perfectly well what he was doing, given he was accepting thousands of pounds from chemical companies to 'prove' their products were safe - products including Agent Orange
RichW, I know about Doll's rather murky history. But I've yet to come across any suggestion that anyone bribed him to fix his tobacco research. The usual suggestion is that he was a legitimate researcher until he got 'bought' by Monsanto et al.
Instead I'm moving towards the view that both the Doll and Hill studies (1950 and 1954) were just extremely bad pieces of research. They just look worse and worse. To me anyway.
People take these studies on trust. They hardly ever read them themselves. They mostly just go along with what everyone else thinks.
I think while many smokers (and non-smokers) are critical of passive smoking studies, most smokers wouldn't question the Doll and Hill studies of active smoking. They aren't able to question it. And yet the passive smoking scare is really built upon the active smoking scare. But my growing conviction is that the whole war on tobacco is garbage from beginning to end, every last bit of it.
John Holmes, what's the gist of Reichmann's critique of questionnaires?
I agree with you Idlex, people accept the active smoking story because it is so engrained, but accept the SHS theory - forgetting that to the new generations it, too, is engrained.
i don't actually know the dates of when Doll got paid off to tell the truth, but i know he got funded by more than Monsanto. I have previously taken the view he used smoking as a scapegoat, but you are making me question that belie - you may be completely correct, that he was a serious researcher who just simply didn't know the shortcomings of his research. That being said though, he made huge errors in his Doctor's study and it was corrected before the 80s, yet he never apologised. In fact, the original study he did showing inhalers got less cancer than non-inhalers was criticised by, I believe, Fischer. Rather than answer him, he removed the inhaling question from the questionnaires! Then he responded 5 years ago Fischer died. I am certain he was not an honest researcher
RichW wrote: I have previously taken the view he used smoking as a scapegoat, but you are making me question that belie - you may be completely correct, that he was a serious researcher who just simply didn't know the shortcomings of his research.
Well, nobody has produced the smoking gun that proves that Richard Doll was paid to show that smoking causes lung cancer. The improprieties came later, when it emerged that he had been taking money from Monsanto while claiming that Agent Orange was harmless stuff. And his defenders say that it was quite normal at that time to not declare such conflicts of interest. Although it never stops them from screaming blue murder if anyone is found to have the faintist whiff of tobacco money backing their research.
My own supposition has been that the young Dr Richard Doll, who was very much the junior partner to Bradford Hill, began to genuinely believe that smoking caused lung cancer as the results of the hospital study began coming in, and the lung cancer patients were overwhelmingly smokers.
But it may also be that Richard Doll had already begun to believe this some years earlier, because he had access to the German Nazi era research. Muller's 1939 paper is cited among the references of the 1950 study, so he must have got hold of that. We also know that Doll attended lectures by SS radiologist Hans Holfelder in Germany in 1936. Perhaps there are other links. It could well be that Doll already 'knew' that smoking caused lung cancer, because he had read a considerable body of the Nazi research. But this was now tainted by its association with Nazism, and so he felt that he had to repeat their studies. And as soon as the results started coming in, confirming the German studies, his suspicions became convictions. And maybe Bradford Hill was of the same mind.
What puzzles me is why Hill selected Doll as his co-author. Doll was a doctor, not a statistician. He had considerable mathematical skills, yet had failed a mathematics examination which would have admitted him to university to study mathematics. Perhaps Hill thought that a statistical study in the medical field needed a doctor on board, and preferably one with some mathematical aptitude.
Either way, the young doctor Richard Doll seems to have been transmuted into a statistical expert by his participation in the studies, and to have built his reputation as that rather than as a doctor. And it would appear that later in life he may well have cashed in on this reputation.
Either way, I can't see that Doll's improprieties later in his life negate his early work. They simply raise question marks about it. It seems to me that the 1950 London Hospital's study should be taken at face value, and criticized for what it is, rather than for any subsequent misdemeanors by any of its authors.
And what bothers me is why Richard Doll should have supposed that he had found a link between smoking and lung cancer in this study. Yes, it's true that 99.7% of the lung cancer patients were smokers. But what was the frequency of smoking in the population? In the whole study, 98% of the participants were smokers. So, whatever disease was being looked at, it was always going to be the case that 98% of the patients would be smokers, just like 98% of the patients would probably be Londoners, and white skinned, and ate bacon and eggs at breakfast. I'm genuinely puzzled that any credence whatsoever was placed in this study.
That being said though, he made huge errors in his Doctor's study and it was corrected before the 80s, yet he never apologised. In fact, the original study he did showing inhalers got less cancer than non-inhalers was criticised by, I believe, Fischer. Rather than answer him, he removed the inhaling question from the questionnaires!
The original study was the 1950 London Hospitals study. It was this study that asked patients whether they inhaled or did not inhale. Sir Ronald Fisher, who was a considerable statistician, showed using data from the study that inhalers got less lung cancer than non-inhalers. He set out his reasoning here. I believe it was with considerable difficulty that Fisher got hold of the data from the study to perform this statistical analysis.
And I have yet to find a copy of the questionnaire that was used in this study. By Doll's account, it asked a number of questions, and covered 3 pages. But it seems clear that something like one of the three pages was devoted to some fairly detailed questions about smoking, of which the inhaling question was one.
There was supposed to be a paper published reporting on the other non-smoking results. But it seems it was never published. Why not? Even if it showed no correlations, surely it should have been published too?
The subsequent Doll and Hill British Doctors study is more famous than the 1950 study, and ran for 50 years. But I've come to believe that the 1950 study is actually the more important one, because it appears that it was during this study that Doll became convinced that smoking caused lung cancer. And Sir George Godber (of later passive smoking fame) was to write that it was known in 1950 that smoking caused lung cancer. The first paper from the British Doctors study didn't appear until 1954. So the conviction or 'knowledge' that smoking caused lung cancer had already taken root in 1950, after the publication of the Doll and Hill London Hospital study, and Wynder and Graham's paper in America.
There are all sorts of unanswered questions floating around this business. It seems that the 1950 paper was fairly thoroughly trashed by Sir Ronald Fisher and others, but their demolition of it has been largely lost in history, and only the unquestionable dogma that smoking causes lung cancer has survived to the present day.
True Idlex, it is a very sorry state of affairs. Thanks for your detailed posts though, I am going to edit a bit of the Doll section of my book to state that he probably was genuinely convinced of his 1950s work, despite it's problems.
Also of relevance is Sir Richard Doll's own recollections of the 1950 study:
By the time we had data on several hundred patients it was obvious that the principal difference between the patients with and without lung cancer was their smoking habits, and we had to make up our minds whether the association was due to chance, bias, confounding, or to cause and effect. The evidence that led us to conclude that it was due to the last (and which led me to give up smoking in 1949) is described in our first paper and the lines of thought we followed were subsequently written up by Hill and have become widely known as "Hill's guidelines." (emphases added)
It's rather hard to see how it was "obvious" that their smoking habits could have been "the principal difference" between the study group and the control group, given that while nearly 100% of the study group were smokers, nearly 96% of the control group were smokers too. (See Table 4 of the 1950 paper).
If half the control group had been smokers, and the rest non-smokers, I'd have to agree that it was the principal difference. But not given the actual figure of 96%.
Once again, I wonder what other differences there may have been between the two groups. If 100% of the lung cancer patients lived in East London, and only 96% of the the control patients lived in East London, might this have been another "principal difference" as well? We don't know, because the other data was never published.
Perhaps another quote from Sir Richard Doll might give some insight.
In 1973 he addressed the Mediacal Journalists Asscoiation about his forthcoming Nuffield lecture on "Monitoring The Health Service". He said (as reported by the Daily Telegraph)
"Old people over 65 should be prepared to accept death and should not be thinking of how to preserve their life a few more months, a 60 year-old professor of medicine said yesterday. It was their social responsibility to 'live dangerously' and not expect the Health Service to spend time and money on research into prolonging their lives."
Maybe he took his own advice when it came to Desert Island Discs?
My first time on this site, so please excuse me if I state 'the bleeding obvious', as Basil Fawlty, once said. There are so many issues surrounding the smoking debate and the loss of so many people's freedom that need to be brought up, mulled over and debated, and this site should be the perfect place to do exactly that. So why then, are so many of you throwing everything you've got into a fake debate with a fake debater?
Ask yourself this; what will the outcome of such a 'daft' debate bring to the table? Will it provide any sort of solution to the anti-smoking law? Will it bring about an amendment of such a law? Will it make politicians sit up and take notice of what smokers are saying? The answers to all three questions are a big NO!
There seems to be so many very intelligent and sensible people on this site who should be directing their energies and thoughts in another direction instead of wasting time in pointless arguments with fictional characters.
I was directed to this site by a friend, who also pointed out another similar site to me. I will not name the second site, which at first glance did look an excellent site, but when I delved into the level of intelligent conversation on there, I found it to be quite pathetic, and as far as I could see, even your fictional poster doesn't go there. He comes to this site. And the reason he comes to this site is because he likes the idea of grabbing your attention and making you all so angry that you will spend hour after hour of your valuable time in pseudo debate with him.
I would love to come back here and debate with all of you, but I am not going to while so many of you waste time on pointless arguing which leads absolutely nowhere, except who has the most time to look up bundles of 'facts' from other Internet sites.
Send this person packing. Ignore him and start getting things organised. If he wants to argue like a demented schoolboy, let him go elsewhere where he will have the freedom to argue, moan and talk silly as much as he likes.
Hi Sally,
We are 60 years behind the curve here. It is vital for all of those defending the freedom to smoke to know the facts. The interweb can be a bewildering place to many people and here, as with all other sites, the readers outnumber the posters by many thousands to one.
The antis rely on disinformation and forums/blogs like this allow us to hone our arguments. I never fail to learn from other posters, and I have spent thousands of hours reading studies, beefing up on the science, and committing a million facts to memory. And I still only know 1% of what I need to know to defeat these cultists.
Debate, such as you see here, does three things:
1. It helps the poster perfect his/her debating skills.
2. It spreads more information that we can all use.
3. It corrects the ban-fans who think the that the "debate is over".
I am certain other, more astute, posters can think of different reasons to demonstrate that it is useful.
But tell me, what would you have us do?
Do you have plans to "get organised"?
I think it provides a central place where people can come together to discuss such issues without the need to organise meetings where not everyone can attend / afford.
Sally, first may I congratulate you on a very well presented comment.
I would second Colin on his reference to the fact that many people read blogs and websites such as these. It is a fact that not only surfers visit, but they are trawled regularly by journalists, MPs and those who these comments are attacking. I think I know the other site you were referring to, and it is the Front Page of that site which is read by many people of all persuasions. Your remark about the forum was a little unkind. May I remind you that that is a forum not a blogsite.
As for the person you refer to as a fictional poster, he is very much alive, no more fictional that timbone, idlex or west2. Rollo Tomassi is a well known supporter of the SBE, whose only problem is that he regurgatates the same comments and has never moved on. Many of us have come across him in several places, including Scottish online newspapers and Labour MP Paul Flynn's blogsite.
Rollo serves a purpose. He provokes rebuttal, including well researched people like idlex, DaveA and others. There is nothing wrong with repetition, as you never know who is reading it, and after all, is this not the method of anti smoking campaigners with their illustrious leader ASH?
After some furious debate on more than one of Paul Flynn's blogs, which of course included Rollo, even the pompous, self opinionated, rude and arrogant Flynn himself admitted to me that writers, mainly from F2C, had presented a fair argument which was in no way inferior or uninformed.
If that is the effect on Flynn, then think of the effect on other readers, many of them silent.
Sally G wrote: He comes to this site. And the reason he comes to this site is because he likes the idea of grabbing your attention and making you all so angry that you will spend hour after hour of your valuable time in pseudo debate with him.
Actually, this is has been the first time (that I know of) that Rollo Tommasi has showed up here - and he shows up all over the place. And I believe he came here with the intention of getting Simon Clark to agree that he had been misquoting Sir Richard Doll, rather than get into an argument with us. Simon Clark did in fact quote Doll accurately, and so has no need to agree with Rollo.
And, to give him credit, Rollo is a one of the more courteous and reasonable of antismokers. Or at least he has been here. Most of the antismokers I've encountered restrict their remarks to saying that smoking stinks, and hoping that I get lung cancer.
And I've learned a lot from him. I've learned that he will take a number of statistically non-significant studies which prove nothing, bundle them all together, and get a statistically significant study which proves something. In such manner is base metal turned into gold.
I hope everyone had a good Easter. I see some of the posters have been busy while I was away. Unfortunately, beyond a couple of comments about what Sir Richard Doll said, nobody has tried to mount a defence of Forest on any of the points I made. Let me remind you of the fundamental flaws in Forest’s statement.
1. It exaggerated the likelihood and effects of misclassification and the existence of confounders), without any supporting evidence.
2. It went into depth in criticizing an out-of-date EPA report from 1992, but completely ignored more up-to-date and relevant overview studies which they are unable to criticise – such as Law et al (1998), Hackshaw et al (1998), IARC (2002) and SCOTH (2004).
3. It applied Forest’s own completely distorted version of scientific reasoning, and never mind that no serious professional scientist would back up what you have to say (e.g. by arguing that the test of passive smoking depends on how many individual studies show a statistically significant increased risk).
4. It took some perfectly reasonable quotes by distinguished professionals (preferably off-the-cuff and unscripted words used in interviews, which are unlikely to be carefully phrased, or old statements which are difficult to access and therefore hard to prove or disprove). Then Forest removed unhelpful parts of their statements to make it seem that they are actually saying something else. Richard Doll was one example. So were the various quotes which supposedly claim that a relative risk of 2 or more is required in all cases. And as for Sir George Godber, does ANYBODY have a full transcript of what he actually said that day in 1975????
I’ve also not seen any effort by Rich or Tim to confirm that CRUK did in fact state that “0.1% of smokers, I repeat, 0.1% are diagnosed with lung cancer every year.” Is there a particular reason why you have not tried to illuminate me? Does your stat in fact exist?
Of course, Rich and Idlex have had a conversation about Doll’s work on active smoking. It’s got nothing to do with Simon’s post here or my points, of course, but I did find it interesting reading. I was fascinated by Idlex’s belief that the first Doll and Peto questionnaire asked questions in exactly the same way as was reported in the 1954 article. Idlex – you do know the difference between direct quotation and indirect reporting, don’t you? You also didn’t notice that in that first questionnaire current and former smokers were asked follow-up questions. And that the questioning became more sophisticated in follow-up surveys. Or maybe you did know, but knew you couldn’t acknowledge these facts because it would ruin the picture you were trying to create.
And as for Idlex suggesting that Doll might have been leaned on heavily by peers to change his views – please! If you have evidence that this happened, then show it. Otherwise, scurrilous allegations have no place in a supposedly mature debate.
As I said earlier, the irony is that Simon Clark has just applauded someone else (Michael Siegel) for condemning the tobacco control movement for using extreme claims which they cannot justify. But he appears not to have the decency and courage to challenge deceit and distortions from his own side. Let’s see if the Easter break has allowed his conscience enough time.
Rollo, I can't be bothered to your respond to that post, you're still suffering the same problems I pointed out on the Scotsman.
But, the smokers percentage, is 0.3% actually. CRUK says 11 million smokers, 38, 598 lung cancers diagnosed a year. Assuming ALL those are smokers, that's 0.3% of smokers a year. Of course, they're not all smokers, so the figure is even less.
They also admit most lung cancer patients are over 60 (curiously, until recently that said '65', but that was obviously skewing their statistics further).
Oh, and to point out, Law, Hackshaw, IARC and SCOTH can and have been demolished, not only by F2C but various people. Rollo, ignoring or rejecting something doesn't mean it hasnt happened
Thank you for your answers to my first post on here, which although I do not agree with in essence, I applaud the honesty in your answers.
The poster whom I referred to as a 'fictional poster' has apparently just shown up again, and because he shows up, as someone pointed out, on various websites and blogs, you all seem to assume that he is a real person. I Googled his name, and the answer that came up was some obscure popstar, which I somehow do not think, this character is! He also seems to have an equally fictional wife on here who crops up now and then, and according to the little I have read, talks even more rubbish than he does.
A big problem with the internet is that people can call themselves whatever they want, and many people, including myself, do not want our full names bandied around for various reasons. My reason relates to my job and my family. But what I write, are my own personal opinions, which you are all free to ignore if you so wish. The character whom I accuse you all of wasting time with, asks you to believe his 'facts', which it seems most of you most definitely do not!
This is my argument, this is why I say he is coercing you all into wasting your time. He knows you could and should, be putting your energies into fighting the real villains, and if you did, you would have a very strong chance of stating your case and winning your battle, but while you are wasting time with 'fiction', you will achieve nothing!