Tilted Forum Project Discussion Community - View Single Post

raveneye · 11-25-2005, 05:49 AM

I don't want to get into another big honking debate about SHS (secondhand smoke) but for longbough I'd like to point out a few things:

--the EPA did appeal the court decision against it, and won the appeal (in 2002). That judge's ruling was completely vacated. Isn't it interesting the Crichton didn't point this out?

--the EPA report did pass peer review. There were over a dozen independent scientists that approved the study. It was and is approved by the American Cancer Society, the Surgeon General, and a slew of other scientific agencies.

--it is not at all correct to say that the CRS panned the study. The CRS report simply was a neutral review of the information at the time, and it neutrally reported the criticisms of the EPA study. It did not perform any statistical analysis that resulted in any negative evaluation of the study.

--the CRS report was not a peer-reviewed published study. It was just a preliminary government report.

--there are several other studies besides the EPA study that demonstrate an increased risk of lung cancer due to SHS. Even if you completely disregard the EPA study, the Fontham et al. study, the largest case control study performed at the time, showed essentially the same results as the EPA study. Their reported median value of lung cancer deaths per year is about 2700 (if I remember right), which is not significantly different from the EPA value. There are several others in addition that corroborate the EPA conclusions.

--there is nothing unusual about a one-tailed statistical test for the effects of environmental toxins. The EPA uses these routinely and nobody questioned them in studies of other toxins. But their other studies did not involve the tobacco industry.

--anybody interested in the politicization of science in this context should consider that, if you want to predict whether any particular study is going to conclude that smoke is harmless, your best predictor is the funding source. If the funding source is the tobacco industry, then the study nearly always predicts no harmful effect. Nearly 3/4 of all studies concluding no effect of SHS were funded by the tobacco industry.

--longbough: you say you have reviewed the methods of the EPA study and found them to be completely invalid. I'm a statistician myself, and have also looked at the methods. It's a 600-page report, and I certainly haven't read every word. But I've looked at the tables and the associated methods, and these indicate to me, clearly, a very highly significant overall effect. They are completely convincing. Even if you use a 95% confidence interval, the effect is still very highly significant. If you'd like to point to me the specific passages in the report that you disagree with, I'd like to see them. To my knowledge none of the critics of this report have done this, including Judge Osteen. So I would be very interested in seeing exactly where in the report you have found any fatal statistical flaw.

Bottom line: the EPA study is just one of many, and it has been completely vindicated by subsequent research. Even if you disagree with it, the overall picture is a clear increase in lung cancer due to SHS exposure. Of course the effect is not as great as with smoking, but it is real, and it increases with exposure. This stands to reason: SHS contains the same carcinogens as cigarette smoke.

11-25-2005, 05:49 AM	#44 (permalink)
raveneye Born Against	I don't want to get into another big honking debate about SHS (secondhand smoke) but for longbough I'd like to point out a few things: --the EPA did appeal the court decision against it, and won the appeal (in 2002). That judge's ruling was completely vacated. Isn't it interesting the Crichton didn't point this out? --the EPA report did pass peer review. There were over a dozen independent scientists that approved the study. It was and is approved by the American Cancer Society, the Surgeon General, and a slew of other scientific agencies. --it is not at all correct to say that the CRS panned the study. The CRS report simply was a neutral review of the information at the time, and it neutrally reported the criticisms of the EPA study. It did not perform any statistical analysis that resulted in any negative evaluation of the study. --the CRS report was not a peer-reviewed published study. It was just a preliminary government report. --there are several other studies besides the EPA study that demonstrate an increased risk of lung cancer due to SHS. Even if you completely disregard the EPA study, the Fontham et al. study, the largest case control study performed at the time, showed essentially the same results as the EPA study. Their reported median value of lung cancer deaths per year is about 2700 (if I remember right), which is not significantly different from the EPA value. There are several others in addition that corroborate the EPA conclusions. --there is nothing unusual about a one-tailed statistical test for the effects of environmental toxins. The EPA uses these routinely and nobody questioned them in studies of other toxins. But their other studies did not involve the tobacco industry. --anybody interested in the politicization of science in this context should consider that, if you want to predict whether any particular study is going to conclude that smoke is harmless, your best predictor is the funding source. If the funding source is the tobacco industry, then the study nearly always predicts no harmful effect. Nearly 3/4 of all studies concluding no effect of SHS were funded by the tobacco industry. --longbough: you say you have reviewed the methods of the EPA study and found them to be completely invalid. I'm a statistician myself, and have also looked at the methods. It's a 600-page report, and I certainly haven't read every word. But I've looked at the tables and the associated methods, and these indicate to me, clearly, a very highly significant overall effect. They are completely convincing. Even if you use a 95% confidence interval, the effect is still very highly significant. If you'd like to point to me the specific passages in the report that you disagree with, I'd like to see them. To my knowledge none of the critics of this report have done this, including Judge Osteen. So I would be very interested in seeing exactly where in the report you have found any fatal statistical flaw. Bottom line: the EPA study is just one of many, and it has been completely vindicated by subsequent research. Even if you disagree with it, the overall picture is a clear increase in lung cancer due to SHS exposure. Of course the effect is not as great as with smoking, but it is real, and it increases with exposure. This stands to reason: SHS contains the same carcinogens as cigarette smoke.