Friday, February 23, 2007

Harvard Report Continues to Deceive: FDA Bill Would Preclude Single Most Effective Regulatory Action to Protect Health that is Politically Feasible

The report from Harvard researchers which claims that all cigarette manufacturers (including Philip Morris) have steadily increased the nicotine yields of their cigarettes over the past 9 years continues to be used to deceive the public, and now it is being used to support legislation that would actually preclude the one regulatory intervention that could effectively utilize control over nicotine yields to improve the public's health.

An article in today's The Harvard Crimson, entitled "Prof to Fight Big Tobacco in Senate," reports that one of the authors of the nicotine report is preparing to "fight Big Tobacco" by testifying before a Senate committee in favor of legislation that would give the FDA some regulatory authority over tobacco products. And according to the article, the ammunition that is being used primarily is the fact that cigarette companies have allegedly been increasing the nicotine yields of their cigarettes. The nicotine study author implies in the article that the ideal solution to this problem of addiction of our nation's children would be to force the companies to lower the nicotine yields:

"Harvard School of Public Health (HSPH) Professor Gregory N. Connolly, whose findings on the rising nicotine content in cigarettes drew ire from tobacco giant Philip Morris USA last month, will bring his fight against Big Tobacco to the U.S. Senate on Tuesday. Connolly will appear before the Senate’s Health, Education, Labor, and Pensions Committee to argue in support of legislation recently introduced by Sen. Edward M. Kennedy ’54-’56 (D-Mass.) and Representative Henry Waxman (D-Calif.) calling upon the Food and Drug Administration (FDA) to regulate the cigarette and smokeless tobacco industries."

"Connolly said he hopes that FDA regulation of the tobacco industry will lead to greater reductions in the level of nicotine found in cigarettes. 'We’ve spent decades relying on the tobacco industry and on their science and it’s been a disaster,' Connolly said. 'We can’t trust this industry for self-regulation.' ... In a letter published in several news sources, John R. Nelson, president of operations and technology for Philip Morris USA, said that 'contrary to the implications of the [HSPH] report, we have not changed the design of our cigarettes with the intention of increasing nicotine yields in order to make the product more addictive.' However, Connolly said that FDA regulation of the tobacco industry is essential to determine whether Philip Morris’ claims are correct. 'If we knew that they’re actually lowering nicotine, I’d be the happiest man in medicine,' Connolly said."

The article notes that the Harvard research team maintains its contention that despite no difference in nicotine levels in Marlboro cigarettes between 1997 and 2006, the data do indicate that these levels have steadily increased; the analysis by Philip Morris (and by myself, by inference) are invalid, the team suggests, because no fancy statistical procedures were used: "Connolly said none of the analysis produced by Philip Morris has included any of the basic statistical measures—such as p-values or R-squared values—necessary to make the analysis valid."

The Rest of the Story

It honestly seems disingenuous to me to condemn the cigarette companies for addicting our nation's youths with the nicotine in their cigarettes and then to support legislation that would institutionalize the addiction of our nation's youths by precluding the FDA from removing the nicotine. It seems disingenuous to me to suggest to the public that we need FDA legislation to address the problem of the nicotine in cigarettes, but then support legislation that precludes FDA from doing anything other than reducing the nicotine levels.

The reason? Reducing nicotine levels will not make cigarettes non-addictive. Reducing nicotine levels will not stop kids from smoking. Reducing nicotine levels will not end the problem of the addiction of our nation's youths.

What will reducing nicotine levels do? It will create a public health disaster by deceiving the public into thinking cigarettes are safer. It will lead to compensation by smokers, who will smoke more to maintain their dosage of nicotine. These smokers will therefore be exposed to higher levels of tar, which will lead to more cancer and emphysema. In short, reducing nicotine levels, without eliminating the nicotine, will kill people.

I think it is most reasonable to argue that removing the nicotine from cigarettes is not a feasible solution to the problem of addiction. I would never criticize someone for suggesting that the FDA should not be given the authority to require the elimination of nicotine from cigarettes. However, I find it inappropriate to mislead people by suggesting that the increased nicotine yields in cigarettes demands enactment of the FDA legislation that has been introduced in Congress. If the fact that cigarette companies are using nicotine to addict youths is a travesty and it needs to be stopped, then the only way to do that is to get rid of the nicotine. You can't bemoan the presence of nicotine in cigarettes, suggest that we need legislation to take care of the problem, and then deceive the American people by supporting legislation whose fine print actually precludes the FDA from taking care of the problem.

I can't over-emphasize this fact: reducing the nicotine yields of cigarettes will not take care of the problem.

What is so disturbing about this story is that we, as tobacco control advocates, have condemned the tobacco industry for doing precisely this: reducing the nicotine yields of their cigarettes.

In fact, we have taken the tobacco companies to court and helped to convict them of racketeering and fraud by virtue of the fact that they chose to decrease the nicotine yields of their cigarettes and market the cigarettes as having lower nicotine yields and therefore being "lighter." We have argued, apparently successfully, that marketing low-nicotine-yield cigarettes is fraudulent, because it deceives the American people into thinking that the product is somehow safer when the truth is that it is not any safer.

So why in the world would we propose a regulatory scheme in which we will do to the public exactly what the tobacco companies have done and been convicted of a crime for doing?

The rest of the story is that tobacco control groups and advocates who are supporting this FDA legislation are essentially calling on the government to do exactly what the tobacco companies tried to do, but for which they were accused and convicted of racketeering and fraud: to reduce nicotine yields of cigarettes.

Such an action by the FDA would certainly mislead smokers into thinking that the product is safer. The truth, however, is that the product would not be any safer. And it might actually be more dangerous.

The truth is that, short of removing the nicotine, the only effective regulatory action that could actually protect the public's health would be to require increases in the nicotine yields of cigarettes. Greatly increasing the nicotine/tar ratio of cigarettes would allow smokers to obtain the same amount of nicotine dosage while inhaling substantially lower amounts of tar. This could potentially reduce cancer and chronic lung disease rates.

However, the proposed legislation precludes the FDA not only from eliminating the nicotine, but also from requiring such increases in nicotine.

According to section 907(a)(4)(A)(i) of the proposed legislation: "A tobacco product standard established under this section for a tobacco product— (A) shall include provisions that are appropriate for the protection of the public health, including provisions, where appropriate— (i) for the reduction of nicotine yields of the product."

So the FDA cannot do the one thing that might actually produce some health benefits from regulation of the nicotine content of cigarettes.

I don't quite understand the quote that suggests that if cigarette companies reduced the nicotine levels, that would be the greatest thing. In my view, that would be a public health disaster.

It really seems to me that we have a double standard in tobacco control. When the cigarettes companies reduce the nicotine yields of their cigarettes, we condemn them. Then, when they increase the nicotine yields, we condemn them.

When the cigarette companies reduce the nicotine yields of their cigarettes, we take them to court, accusing them of fraud. Then we turn around and tell the American people that we need legislation to allow the government to reduce the nicotine yields of cigarettes.

The FDA legislation essentially institutionalizes into the government's hands the fraud that the tobacco companies have just been convicted of committing. It continues to baffle me how any public health advocate or group can support such an approach.

But while I can recognize that there are differences of opinion in the public health merits of the proposed legislation, what I don't find excusable is the deceptive techniques being used to promote this legislation. Supporters of the bill are casting it as a battle against Big Tobacco. But the truth is that the largest company within Big Tobacco - Philip Morris - is supporting the legislation.

These deception is apparently paying off. Newspapers are portraying this as a battle of public health against Big Tobacco. Just look at the headline of the Harvard Crimson article: "Professor to Fight Big Tobacco in Senate." The truth of the matter is that the professor is going to support Big Tobacco in the Senate by helping Philip Morris to promote its prized legislation that it desperately wants enacted. Or at least, that the professor is going to support the desires of the largest company within Big Tobacco. Clearly, it is deceptive to cast the story as a professor fighting Big Tobacco. It is more accurate, I think, to describe it as a professor supporting the chief legislative priority of the nation's leading tobacco company.

I cannot conclude this post without commenting on the argument that my analysis, demonstrating that average nicotine yields of Marlboro cigarettes were no different in 1997 than in 2006, is invalid because it doesn't employ any fancy statistics.

You don't need any fancy statistics to conclude that there has not been a steady increase in nicotine yields if the nicotine yield is the same at the end of a nine-year period as it was at the beginning of the period. Especially if there is a clear trend of decreasing yields over the most recent three years. It's not that the data show an increase from 1997 through 2005, and that the 2006 data point is simply an outlier. The data clearly show an increase from 1997 to 2003, and then a consistently decreasing yield from 2003 to 2006.

If anything, the use of the fancy statistics obscures the true picture. If you force a simple linear pattern to the data, you're going to find an overall increase in the nicotine yields. The point is that the pattern is not a linear one. It really is an upside down "V" pattern - with an increasing yield from 1997 to 2003 and a decreasing yield from 2003 to 2006, with no overall change from 1997 to 2006. Fitting a single line to this pattern obscures the truth.

Let's take the following example to illustrate this. Suppose that we want to determine whether there has been a steady increase in the number of games that the Celtics have won per month during the NBA season. Our data show the following number of wins per month:

November - 1
December - 2
January - 3
February - 4
March - 5
April - 3
May - 1

Anyone looking at these data can see clearly that what the pattern shows is that there has not been a steady increase in the number of wins per month for the Celtics. What the pattern actually shows is an increase in wins from November through March, and then a decrease in wins from March through May, such that the team won the same number of games in May as in November. Clearly, there has not been a steady increase in the number of wins.

What would happen, however, if you tried to fit a simple line in modeling the number of wins per month?

What happens is that you obtain a line with increasing slope. Such an regression indicates that there has been a steady and consistent increase of 0.14 wins per month. In other words, each month, the Celtics win 0.14 more games than the previous month.

Fitting the data with a simple linear pattern forces a conclusion that there has been a steady increase in wins. But the truth, clearly, is that there has been no steady increase in wins. If you were a betting person (and I am), you would certainly not want to put too much money on the Celtics winning many games in June.

You can take your fancy statistics all the way to the bank (to take out a loan for the money you are going to owe for your losing bet).

What is a safe bet, however, is that if the FDA legislation is enacted, the public will be massively deceived about the safety of tobacco products. The saddest part is that this time, the source of the fraud will not be the tobacco companies; it will be our own government.

No comments: