Legacy Foundation Public Comment

February 13, 2012
Food and Drug Administration
Department of Health and Human Services
5630 Fishers Lane, Room 1061
Rockville, MD 20852
Re: Docket No. FDA-2011-N-0867
The undersigned organizations hereby submit their comments on the notice of proposed agency
information collection in the above-designated docket.
Section 904(a) of the Family Smoking Prevention and Tobacco Control Act (“FSPTCA”) requires
all tobacco product manufacturers to submit, by brand, sub-brand, and quantity, a list of all constituents,
including smoke constituents, identified by FDA as harmful or potentially harmful to health. Section
904(e) requires FDA to establish and periodically revise, as appropriate, in a format that is understandable
and not misleading, a list of harmful and potentially harmful constituents, including smoke constituents in
each tobacco product by brand and by quantity in each brand and sub-brand. On August 11, 2011, in
Docket No. FDA-2011-N-0271, FDA published a proposed list of harmful and potentially harmful
constituents based on a list compiled and endorsed by the Tobacco Products Scientific Advisory
Committee. Significantly, the explanatory materials accompanying the notice stated that the list was
incomplete because FDA “has only focused on []five disease outcomes” rather than all potential harms
and because the criterion for listing dependent upon a substance “being both studied and listed by another
entity.” 76 F.R. at 50228. As FDA noted, there could well be other constituents that would warrant
listing had they been studied more extensively. The undersigned groups submitted comments on that list
on November 11, 2011, a copy of which is attached hereto at Tab A.1
Section 904(d) requires the Secretary to “publish in a format that is understandable and not
misleading to a lay person, and place on public display (in a manner determined by the Secretary) the list
established under Subsection (e).” Section 904(d)(2) directs the Secretary to conduct periodic consumer
research to ensure that the list published is “not misleading to lay persons.” In enacting these
requirements, the Congress was attempting to reconcile two important objectives: (1) to ensure that
detailed information concerning constituents in tobacco products was systematically gathered and made
available to the public; and (2) that the information so gathered was made available in a way that
facilitated understanding by lay persons of substances in tobacco products and in smoke that are
hazardous but did not lead to misperceptions about the relative risk of different brands and products or
misperceptions about the health impact of presence or absence of different constituents or different
quantities of different constituents.

1

Those comments called into question, inter alia, discrepancies between the list endorsed by the TPSAC and
the list published by FDA.

1

In this notice, FDA seeks public comment on a proposed collection of information designed to
determine how FDA can publish the List of Harmful and Potential Harmful Tobacco Constituents
(“Harmful Constituents List”) in a format that is understandable and not misleading to a lay person. In
developing such a comment, it is useful to review both the reasons for developing such a list and making
it publicly available and the reasons why there are legitimate concerns that publication of such a list, in
the absence of appropriate safeguards, could mislead lay readers and lead to a result that would not be
beneficial to the protection of the public health.
Reasons for developing and publishing a list of harmful and potentially harmful constituents.
Prior to the enactment of the FSPTCA, tobacco companies were not required to report to a federal
agency the constituents2 in their products and were not required to disclose these constituents to the
public. There was a legitimate concern that neither agencies charged with protecting the public health nor
consumers could possibly know what they were being exposed to when they smoked or otherwise
consumed a tobacco product. Disclosure of this information to FDA serves several purposes. First, it
serves a regulatory purpose by helping to establish a baseline for product regulation. Once such
information has been provided to FDA, it should be clear that any increase in any harmful constituent
would render the product a “new product” within the meaning of the statute and require the manufacturer
to obtain pre-market approval before any such product could be marketed. This standard will therefore
require a degree of consistency and standardization that is highly desirable. Second, submission of the
information will facilitate research on tobacco products. Prior to the legislation, only the companies
themselves knew precisely what the constituents in their products were. Researchers could “reverse
engineer” the product, but submission of test results by the companies will greatly facilitate useful
research. Furthermore, neither government nor researchers were able to evaluate the impact of specific
changes, including seemingly benign changes in substances like sugar, that could impact other ingredients
and constituents. Third, consumers are entitled to information that is not misleading about the contents of
tobacco products to help them more accurately evaluate the health or potential health impact of using
those substances. The need for such information seems particularly compelling when the substances at
issue are toxic, carcinogenic or addictive and have profound effects on the consumer.

The form and content of disclosure should take into account how consumers will perceive and
interpret the information disclosed.
It is normally the case that any increased disclosure of information to consumers will enhance the quality
of decision-making. However, Congress recognized that factual disclosures that the public misinterprets
can do more harm than good and, therefore, required FDA to disclose information about harmful and
potentially harmful constituents in tobacco products in a way that is not ‘misleading.” Congress
recognized that there are legitimate reasons to be concerned that disclosure of these lists to the public,
without adequate safeguards, could mislead consumers and be detrimental to the public. This is the case
both because of the limitations of the lists themselves and because of the possibility that the information
2

Unless otherwise noted, the term “constituents” as used in this comment refers to constituents in the
product and smoke constituents as well.

2

about these constituents—accurate in itself—could very well be misunderstood either if they were not put
in context or if consumers misinterpreted the significance of the information in making individual
decisions about the use of tobacco products. Such misperceptions could influence consumer behavior in
ways that are detrimental to the public health. If current scientific knowledge is not sufficient in all cases
to describe accurately the health consequences attributable to the presence of specific constituents or
specific combinations or quantities of specific constituents, but consumers interpret the disclosures as
implying that products with fewer harmful constituents or products with lower levels of certain
constituents are safer, then the disclosure or disclosure format would be inconsistent with the goals of the
statute and intent of Section 904. The potential for misunderstanding is substantial. In attempting to
develop a format for disclosure that would minimize the risks of misunderstanding, it is helpful to list the
major potential sources of misunderstanding.
First, the lists themselves are incomplete. They include only some of many disease outcomes.
Moreover, the list includes only substances that have been found by other agencies to be carcinogenic,
toxic, or addictive. Many other substances that have not been the subject of comparable study may fit one
or more of the categories. FDA must study the impact of such a partial disclosure, including the adequacy
of a disclaimer that states that that there may well be other harmful constituents that cause cancer or other
diseases that do not appear on the list. If such a disclaimer is not sufficient to undo the misinterpretation
that results, FDA must develop a disclosure plan that does not suffer from this problem.
Second, the quantitative test data currently available do not accurately measure the actual
exposure of consumers. Although the FDA’s proposed rule for testing requires submission of data from
the FTC testing method and the Canadian government testing method, it is conceded that neither
method—and no existing machine testing method—accurately measures constituents as smokers
experience them in practice. Again, FDA must evaluate the efficacy of such a disclaimer to prevent
consumers’ misperceptions; if such a disclaimer is inadequate to do so, FDA should develop an
alternative disclosure plan that does not mislead consumers about the health impact of these numbers.
Third, there is a high likelihood consumers will conclude that lower numbers or fewer
constituents means a product is less risky even though the scientific evidence is not sufficient to reach
such a conclusion. In most instances, current scientific information does not exist to evaluate or quantify
the impact of individual harmful constituents or, indeed, of such constituents in combination or of
different levels of different constituents. Tobacco products are also unique in the number of harmful
constituents each product contains. The multiplicity of harmful constituents makes it more difficult to
draw conclusions about the relative risk posed by various tobacco products and makes it more likely that
consumers will misinterpret the significance of fluctuations in the level of one or more such constituents.
Even if Product B had only half as much of Constituent X as Product A, this difference might be
immaterial to the health consequences of using one rather than another. With thousands of brands and
subbrands and hundreds of harmful constituents, even the most sophisticated consumer would not be able
to draw reliable conclusions about the relative safety of competing brands on the basis of this information.
In developing its system for numerical disclosures of harmful and potentially harmful
constituents, FDA must also carefully take into account the mandate of Section 911 of the FSPTCA.
Section 911 recognizes that an assertion by a manufacturer that a product exposes consumers to a lower

3

level of a constituent constitutes a “Modified Risk Claim” that triggers the protections of Section 911.
FDA must be careful not to allow its own disclosures to cause the same harm.
The danger of such misunderstandings has been underscored by the regulatory experience in the
listing of tar and nicotine content in cigarettes. The attached statement submitted in this docket by
Professor Joel Cohen, an acclaimed consumer behavior researcher with more than thirty years of
experience in evaluating government and industry communication of health-related information, describes
this experience in more detail and draws important conclusions from it that are highly relevant in
considering what must be done to avoid unintended harm to the public health. Drawing on historical
experience regarding “light” and “low tar,” Professor Cohen describes the potential that disclosure of the
information at issue here could have unintended consequences for the public health and suggests
approaches for minimizing this potential.
In the current context, where there is legitimate concern that public disclosure of this information
could have negative public health consequences, the use of surveys to test the actual results of
information disclosure is particularly important. No plan of disclosure should be implemented unless and
until FDA has conducted surveys that demonstrate how consumers perceive and digest information, that
consumers find understandable and not misleading, and that produce results where the benefits of
disclosure outweigh the costs.
Accordingly, in developing a research protocol to design a survey to determine how information
can be conveyed to consumers in a way that would minimize the prospect that such information would
lead consumers to draw misleading conclusions from the data, the protocol should be designed to identify
consumer perceptions from the disclosures, examine alternate formats and methods of disclosure and
examine whether clarifying statements that accompany the published lists are adequate to prevent
consumers from being misled. If clarifying statements can accomplish this goal, the FDA should test
different statements and different forms of disclaimers to determine which are most effective.
Furthermore, FDA should not conclude that inclusion of such statements is adequate to prevent consumer
misperceptions unless and until survey results demonstrate that such statements actually accomplish this
goal.
Moreover, the manner in which the information would be disseminated is extremely important.
While it might be appropriate to publish the lists on the FDA website where they could be accessed by
interested members of the public, it might be highly misleading to publish the same information on the
pack. Moreover, publication of a part of the list while leaving certain constituents off the list could well
be misleading. Thus, any study should seek to determine the effect of disclosure in various modes of
dissemination.
Finally, the study should take into account the consequences of disclosing such a list on the
ability of manufacturers to use the information contained on the list. Section 911 of the Act significantly
constrains this ability. In the absence of an order issued by the under Section 911(g), a manufacturer may
not represent, in its labeling or advertising, that
(1) the product presents a lower risk of tobacco-related disease or is less harmful than one or
more other commercially marketed products;

4

(2) the product contains a reduced level of a substance or presents a reduced exposure to a
substance; or
(3) the product or its smoke does not contain or is free of a substance.3
In addition, the Act prohibits a manufacturer from taking any other action directed to consumers through
the media or otherwise “that would reasonably be expected to result in consumers believing that the
tobacco product or its smoke may present a lower risk of disease or is less harmful than one or more
commercially marketed tobacco products, or presents a reduced exposure to, or does not contain or is free
of, a substance or substances.”
In order to ensure that publication of the list does not result in misleading characterization of the
information contained in it by tobacco product manufacturers, FDA should make it clear to tobacco
product manufacturers that any characterization of or reference to any such list is subject to the limitations
contained in section 911. Moreover, FDA should make it clear that any quotations from the list used by
manufacturers in labeling or in advertising must also fully comply with Section 911.
Format of Publication
FDA should give careful consideration to the manner in which the list should be made available
to the public. If it is made available on the agency’s website, it should be presented with sufficient
explanatory materials to minimize the possibility that the results would be misleading to lay persons.
Moreover, the format of presentation should be considered so that the results are grouped in a manner
designed to avoid misleading presentation.
If FDA uses any portion of the list or any characterization of it in any media campaign, it must
ensure that any such publication or characterization should include sufficient explanatory material to
ensure that such publication is not misleading.
Design of the Survey Sample to Meet These Criteria
It is appropriate for FDA to conduct surveys to minimize the possibility that publication of the list
could mislead the public. In devising the surveys, FDA should bear in mind that there are many different
segments of the public and that presentations that might not be misleading to one group would be
misleading to others. FDA’s goal should be to identify presentations that avoid misleading any segment
of the public. In order to accomplish this goal, FDA will need to conduct surveys in multiple demographic
groups. Given this requirement, FDA should consider whether the intended sample size is sufficient and
whether it will be sufficiently representative of the many different segments of the public to yield the
information necessary to provide appropriate protection to all segments of the public.
The survey should be specifically designed to determine what the effect of disclosure would be
on actual consumer behavior. Insofar as possible, the survey should seek to determine what consumers
would likely have done in the absence of the information and what they are likely to do once the
information has been provided. FDA should proceed with disclosure only if survey results demonstrate
that disclosure would more likely than not result in changes in consumer behavior that, on balance, have a
positive impact on the public health.
3

FSPTCA, Sec. 911(b)(2)(A)(i)

5

The survey should seek to provide persuasive answers at least the following essential questions:
Do consumers understand the information that is being provided?
What conclusions regarding the relative health risks of various products will consumers take from
looking at the data?
Will consumers understand that even products that have lower levels of some harmful substances
still contain many substances that cause death and disease and that products that contain lower levels of
such substances may be just as dangerous?
Can the information be conveyed to consumers in a way that that the dominant message is that
every tobacco product exposes users to many potentially fatal substances? If so, how should format and
context of disclosure be formulated to achieve this result?
How likely are consumers to focus on the differences in content of hazardous constituents
between combustible and smokeless tobacco products as opposed to differences in content of hazardous
constituents as between different brands?
In evaluating the survey results and considering what disclosure can best contribute to the
protection of the public health, the FDA should keep in mind the Congressional directive to ensure that
disclosure will not be misleading. If a more extensive disclosure cannot be accomplished without
misleading the lay public, FDA should consider limiting publication to a listing on its website. In any
event, FDA should make it clear to manufacturers that Section 911 applies to any characterization of the
listing.
Sincerely,
American Cancer Society Cancer Action Network
American Heart Association
American Lung Association
Legacy
Campaign for Tobacco-Free Kids

6