|Home | About | Journals | Submit | Contact Us | Français|
Many people may search for information about tobacco use, the largest cause of preventable mortality in the United States, on the Internet. In 1999, Philip Morris U.S.A. (PM), the country’s biggest cigarette manufacturer, posted a Web site and launched a campaign to encourage people to obtain information about tobacco and health issues there. The company asserted that its goal was to deliver the messages of the public health community about tobacco. However, internal tobacco company documents reveal that the site was a public relations effort intended to help the company avoid punishment and regulation. Examination of the language on the Web site reveals many contradictions and omissions that may undermine public health messages. Among these are vague and confusing information about addiction, tar, and nicotine, a lack of motivators to quit smoking, and silence about tobacco-related mortality. By appearing to join with public health organizations in disseminating “responsible” messages about tobacco, PM may improve its image, thus facilitating its ability to continue to sell its lethal products. Public health nurses should be prepared to examine health information on the Internet for subtle biases, to counter PM’s specific language about smoking to patients, and to challenge PM’s larger corporate goals.
Tobacco is a priority issue for public health nurses. Approximately 450,000 people in the United States and more than 4.8 million worldwide die annually from tobacco-related diseases (Health Consequences of Smoking, 2004; MacKay, Eriksen, & Shafey, 2006). Nurses working in tobacco use cessation or prevention may find that, like millions of Americans (Fox, 2006), clients are accessing information about health topics via online sources (Khechine, Pascot, & Premont, 2008; McCaw, McGlade, & McElnay, 2007; van de Poll-Franse & van Eenbergen, 2008). However, information online is not always accurate or useful (Eysenbach, Powell, Kuss, & Sa, 2002). Several studies have suggested adopting standardized criteria to evaluate health-oriented Web sites (Cheh, Ribisl, & Wildemuth, 2003; Khazaal et al., 2008; Makar et al., 2008; Monahan & Colthurst, 2001), including currency of information, references cited, and conflict of interest disclosure (Bernstam et al., 2008; Eysenbach, 2002); however, there is little agreement about what criteria reliably predict high quality (Bernstam et al., 2008; Eysenbach et al., 2002).
Philip Morris U.S.A. (PM) sells just over half of all cigarettes in the United States. In 1999, PM developed a Web site and began an aggressive campaign to refer people there for health information about tobacco products (Cigarette Makers Frankly Spell Out Smoking Danger, 2000). PM publicized the Web site through television, magazine and newspaper ads, and cigarette pack inserts (Szymanczyk, 2005, pp. 90–91). PM had denied for decades the harms associated with its products (Brandt, 2007; Carter & Chapman, 2003; Hiilamo, 2003), an increasingly unbelievable position, but in 1997, amidst legal negotiations with public health groups and numerous states’ attorneys general, PM agreed to no longer publicly argue about whether smoking was addictive or caused disease (Philip Morris’ Statement of Position, 1997). This agreement, called the Hatch Statement, was supposed to ensure that there was one consistent public health message about tobacco (Philip Morris’ Statement of Position, 1997). The Web site claims to support that public health message (PM, 2008j).
PM’s Web site has been identified by public health researchers as a public relations vehicle for the company, which continues to deny in court that its products cause disease (Friedman, 2007). However, PM’s extensive promotional efforts mean that people are likely to visit it for information. This study examined PM’s Web site to (a) evaluate whether its messages live up to the company’s promises, (b) determine whether it reflects the scientific evidence about tobacco products and health, and (c) analyze the implications for nurses of PM positioning itself as a source of health information.
We reviewed 78 separate Web pages (URLs), collected from the PM Web site (http://www.philipmorrisusa.com) in February 2008, for statements relating to health of smokers and those exposed to cigarette smoke. Pages not examined encompassed financial information, business issues, philanthropy, and the Master Settlement Agreement (a legal agreement reached between the tobacco companies and the attorneys general of 46 states as a result of litigation over Medicaid costs for tobacco-related illness; National Association of Attorneys General, 1998). To determine whether and how pages had changed over time, we examined previous versions of the Web site from the Internet Wayback Machine (http://www.archive.org/index.php) that archives portions of the World Wide Web. We also searched the Legacy Tobacco Documents Library (http://legacy.library.ucsf.edu) (an online repository of more than 8 million internal tobacco company documents released as a result of litigation and associated trial transcripts and depositions), using a snowball sampling strategy, beginning with the search term “website” (Carter, 2005; Malone & Balbach, 2000). We examined approximately 640 PM documents pertaining to the development of the Web site and testimony from selected PM executives. For comparative purposes, we examined materials from public health organizations that were referenced by PM’s Web site.
Using archival strategies (Hill, 1993), we sought to interpret and contextualize our findings by triangulating data sources. We analyzed Web page written content for comprehensiveness, internal consistency, and consistency with information available in the scientific literature and with other company statements and documents.
This study is based on a series of snapshots of Web pages, beginning in November 1999 and ending in February 2008. Web pages can be altered and changes may not be apparent; there could have been site versions unknown to us. Material discussed here may not still be extant. The study does not encompass the Web site for Philip Morris International, which is responsive to multinational concerns and pressures. Between 1999 and 2002, the Wayback Machine has only the PM homepage, and so a detailed comparison with pre-2002 versions was not possible. It is possible that there are other documents related to the Web site that we were unable to locate in the Legacy library, or that other relevant documents were concealed or destroyed by the tobacco industry (LeGresley, Muggli, & Hurt, 2005). The database of trial transcripts we examined is not a complete set of all tobacco-related litigation; however, the statements of executives in these cases were made under oath.
In 2004, PM said that the Web site was part of its “efforts to inform people about the risks of smoking” (PM, 2004). However, internal company documents show that the main purpose of the Web site was to support the company’s PM21 (“Philip Morris in the 21st century”) image enhancement campaign “to project a positive impression and sustainable image for our company” (Dangoor, 1999). Corporate ads that began airing on television in late 1999 were, according to Denise Keane, Senior Vice President at Philip Morris Management Corporation, to be supported by “a Web Site that presents information on the controversial tobacco issues” (Keane, 1999). This would help ensure that the company was “viewed as forthright” (Keane, 1999). Reference to tobacco issues as “controversial” is a long-standing rhetorical ploy of the tobacco industry, which deliberately manufactured debate over the well-established connection between tobacco use and disease (Brandt, 2007; Kennedy & Bero, 1999).
This Web site area has undergone extensive revision in tone and substance. The first versions of the site, 1999–2002, had no separate health issues section (PM, 1999,PM, 2001PM, 2002a). Pages added later include “Addiction” (PM, 2008i), “Low Tar Cigarettes” (PM, 2008k), “Surgeon General (SG) Reports” (PM, 2008o), and “Smoking and Pregnancy” (PM, 2008n).
Overall, information in this section was and remains factually accurate. The “Disease in Smokers” page (PM, 2008j), while less than comprehensive, stated that PM agreed with “the overwhelming medical and scientific consensus that cigarette smoking causes lung cancer, heart disease,” and other diseases (PM, 2008j). (Earlier versions of the page had merely acknowledged the existence of the scientific consensus; PM’s concurrence was added in 2000, as a result of public pressure; Friedman, 2007.) It also linked to public health sources including the American Cancer Society, Centers for Disease Control and Prevention pages about the health consequences of smoking and U.S. Surgeons General reports, and the World Health Organization (WHO) (PM, 2008j). In trial testimony, Philip Morris Incorporated Chairman Michael E. Szymanczyk said that the site emphasized that smoking causes disease to affirm that “the public health community” should be the “one voice” communicating with the public (Szymanczyk, 2005, p. 59).
However, the language on the site, while not wrong, does not fully deliver “consistent public health messages.” One key public health message about tobacco use is its related suffering and mortality; tobacco kills half of its long-term users (Doll, Peto, Boreham, & Sutherland, 2005). “Death,” except in the context of sudden infant death syndrome (PM, 2008m, 2008n) and twice in brochures about teens (PM, 2008d, 2008e), was absent from the Web site. The number of deaths caused and the percent of smokers who die from tobacco-related diseases were not mentioned. (The brochures about youth smoking both urge parents to talk to their kids because “it has been estimated that more Americans die from tobacco-related illnesses than from alcohol, car accidents, HIV/AIDS, firearms and illegal drugs combined”; PM, 2008d, 2008e.) An internal Q&A document, apparently written to help employees answer questions about the Web site, asked “Do you now accept that smoking kills 400,000 people per year in the U.S., and millions more around the world?” The answer: “We don’t know,” although the company admitted that it did “accept that, over the years, smokers have died at least in part because of their smoking” (Smoking and Disease in Smokers, 2000).
“Tar” refers to smoke particulate matter, the source of most cigarette carcinogens; nicotine is the addictive substance in tobacco. The Federal Trade Commission (FTC) machine-tests cigarettes for their T&N yields and cigarette manufacturers usually place the FTC figures on packages and advertisements, although these figures do not correspond to the levels smokers actually receive, and no level of tar or nicotine has been determined to be safer or less addictive (Gray, 2000). However, smokers frequently believe that low-tar cigarettes pose less risk (Borland et al., 2004; Shiffman, Pillitteri, Burton, Rohay, & Gitchell, 2001).
T&N were not defined under “Understanding tar & nicotine numbers” or anywhere on the site (PM, 2008f). The site indicated that because smokers can “compensate”—alter their smoking practices and thereby increase the T&N received from an ostensibly low T&N cigarette—the T&N numbers do not indicate either the “actual amount” of T&N a smoker might inhale or the “relative amount,” compared with what the smoker might get from another brand (PM, 2008f). What information the smoker should get from the numbers was not explained, except in the context of product descriptors. This page did not point out that T&N numbers vary depending on the state in which the cigarette is purchased or the packaging (e.g., box or soft pack): a separate page offering numbers for specific brands revealed this (PM, 2008g). However, T&N numbers for about half of the types listed (75/155) were unavailable (PM, 2008g).
“Descriptors” such as “light” and “mild” are used by tobacco companies for marketing purposes. Public health authorities strongly oppose descriptors, because consumers wrongly assume they indicate that a product may be less harmful (Borland et al., 2004). Although a Web site page was devoted to explaining PM’s “use of brand descriptors” (PM, 2008k), what they represented was unclear. Szymanczyk testified that they represented “a range of [T&N] yields” as measured by the FTC machine method (Szymanczyk, 2005, p. 74). (The FTC does not define or approve use of these terms.) This suggests that the terms did indicate absolute values of machine-method yields, but if so, the specific values were not revealed. The Web site claimed that descriptors helped smokers distinguish “strength of taste” and “reported tar yields,” but it was not explained how these characteristics related to one another (PM, 2008f). Nor was any reason given for smokers to choose a brand based on reported yields if they did not reflect what the smoker might inhale. PM claimed that descriptors used “tar numbers” as a reference point, but elsewhere said that that they also reflected nicotine deliveries (PM, 2008f).
PM’s vague and contradictory language is exemplified in Table 1. Language in the first row of the table cautions that neither T&N numbers nor descriptors mean that the product delivers a specific, or relative, amount of T&N. A smoker could get more tar and/or nicotine from a “light” cigarette than from a “full flavor” cigarette. In the second row, the language suggests that T&N numbers and descriptors do offer a meaningful basis for comparisons. It was not explained how, if the numbers and descriptors do not correspond to actual or relative values, they help with such comparisons.
The “Addiction” page stated that “Philip Morris USA agrees with the overwhelming medical and scientific consensus that cigarette smoking is addictive” (PM, 2008i). The page continued by describing quitting smoking as “very difficult” (PM, 2008i). This was the only “definition” of addiction offered. There was no mention of either the physiological or the behavioral aspects of addiction; nicotine was not mentioned, nor was it identified anywhere on the site as the addictive substance in tobacco—a rhetorical strategy that PM defense attorney Dan Webb called a matter of “policy” but not of “fraud” (Defendants’ Closing Statement, 2005). Despite these omissions, the page also referred to the company’s support of “a single, consistent public health message” on the subject (PM, 2008i).
SHS has been shown to be a cause of cardiovascular and respiratory diseases, as well as lung and breast cancer (The Health Consequences of Involuntary Exposure to Tobacco Smoke, 2006). No safe level of exposure is known (The Health Consequences of Involuntary Exposure to Tobacco Smoke, 2006). Health risks posed by SHS were not covered by the Hatch Statement, which only addressed primary smoke disease causation and addiction. PM’s divergence from public health messages on this issue (Drope, 2004) was subtly presented.
The “Secondhand Smoke” page repeated that “public health officials” concluded that SHS was hazardous, but not that PM concurred (PM, 2008m). PM executive Ellen Merlo testified in 2001 that the company had altered the page about addiction to declare PM’s agreement that smoking was addictive (Merlo, 2001). PM did not likewise clarify the page about SHS. The “Smoking Restrictions” page asserted that the conclusions of public health officials “warrant certain measures that regulate smoking” so that people can “avoid being around secondhand smoke.” The onus was placed on individuals, who must decide “whether to be in places where secondhand smoke is present” (PM, 2008a). The smoke’s presence was assumed. Children were an exception. The Web site says: “Particular care should be exercised where children are concerned. Adults should avoid smoking around them” (PM, 2005b).
The Web site asserted that “elevators, places where a specific fire hazard already exists, [and] areas occupied primarily by children” were suitable for smoking bans (PM, 2008a). “Avoidance” of smoke was desirable in places where people “must go, such as public buildings, public transportation, and many areas in the workplace [emphasis added]” (PM, 2008a). The Web site did not explain why people should be exposed to smoke in some areas of the workplace or which areas these were. The practical consequence of adopting PM’s recommendations would be minimal: the places described as suitable for bans have mostly long been smokefree. PM’s Web site still maintained that complete bans “go too far” (PM, 2008a).
Selecting “reasonable ways” to respect “the comfort and choices of both nonsmoking and smoking adults” lay with businesspeople—”particularly owners of restaurants and bars” (PM, 2008a). This is “accommodation,” a policy PM has espoused since 1988 (Dearlove, Bialous, & Glantz, 2002; Smith & Malone, 2004), which proposes “separation, separate rooms and/or high quality ventilation” (PM, 2008a) instead of smoke-free air. Although other pages linked to relevant scientific literature, the “Smoking Restrictions” page linked only to PM’s own “Secondhand Smoke” page (PM, 2008m), not to research showing that separation and ventilation fail to protect non-smokers (The Health Consequences of Involuntary Exposure to Tobacco Smoke, 2006; Repace, 2004; Repace & Lowrey, 1980). Early drafts of this page suggested that hospitality workers (e.g., wait staff, bartenders) who were likeliest to be exposed in this scenario should “make their own decisions” about working “in establishments where they are exposed to smoke” (Web Site Backgrounder, 1999). The current page does not address this issue.
Quitting—even after many years of smoking—benefits long- and short-term health (U.S. Department of Health and Human Services, 1990). Quitting has appeared on PM’s Web site since its inception. The language on the current page appeared in 2002: “To reduce the health effects of smoking, the best thing to do is quit” (PM, 2008l). This brief page did not name the “health effects” of smoking or mention addiction; it did acknowledge that quitting smoking “can be difficult” (PM, 2008l). The page linked to the Web sites of major U.S. voluntary health organizations and government health offices, and gave the national quitline number. Finally, it linked to PM’s “information resource,” QuitAssist (PM, 2008h).
The 50-page QuitAssist booklet was primarily comprised of materials from other quitting manuals, including those produced by the American Cancer Society (2007), the American Heart Association (2007) and the American Lung Association (2003), and the National Cancer Institute (2007). However, these materials begin by talking about either addiction (American Cancer Society, 2007; National Cancer Institute, 2003; U.S. Department of Health and Human Services, 2000) or death (American Lung Association, 2003; National Cancer Institute, 2007; Pathways to freedom, 2002). QuitAssist made no reference to death or nicotine and only obliquely referred to addiction, saying that “Rewards of quitting” included being “free from the …dependence of cigarette smoking” as well as having “more control” over one’s life (PM, 2008h).
Diseases caused by smoking include cancers of the lung, mouth, throat, kidney, bladder, and cervix, coronary heart disease, chronic obstructive pulmonary disease, and emphysema, among many others (Health Consequences of Smoking, 2004). QuitAssist mentioned a much shorter list: heart attack, stroke, lung cancer, emphysema, and other lung diseases were identified as “reasons to quit smoking” (PM, 2008h). QuitAssist provided no specific information about how likely it is that a smoker might contract or die from these conditions. In discussing SHS, QuitAssist also restricted its consequences to giving children “coughs, colds, and earaches” (PM, 2008h); public health sources mention cancer and heart disease (American Lung Association, 2003), SIDS (National Cancer Institute, 2007), and death (American Cancer Society, 2007; Pathways to Freedom, 2002).
The Web site used the conditional tense to discuss cessation. Thus, QuitAssist was described as a resource “If you decide to quit smoking” (emphasis added) (PM, 2008h). The T&N page recommended quitting, “If smokers are concerned” (PM, 2008f). Parents were told that “If you are considering quitting,” they should remember that smoking kids were likelier to quit if their parents quit (PM, 2008d). However, the page did not advise either parents or anyone else to quit, although the Quitting Smoking page observed that “To reduce the health effects of smoking, the best thing to do is to quit” (PM, 2008l).
The audience for PM’s cessation materials was consistently described as “smokers who have decided to quit” (PM, 2008b, 2008l) in contrast to the public health materials, which frequently started with explicitly labeled arguments to persuade smokers to make that decision (American Cancer Society, 2007; American Lung Association, 2003; National Cancer Institute, 2007; Pathways to Freedom, 2002). The PM site lacks encouragement of or reference to the crucial process of making the decision to quit, explicitly addressing only those who have already made that decision. Thus, no motivation is given for this key element in the process of cessation.
PM’s Web site noted that “people may question” the company’s support of cessation, since “it may not seem to make sense” (PM, 2008b). However, the only reasons given for this support were that “smoking causes diseases and is addictive” (PM, 2008b) and so the contradiction remained: it does not make sense to both promote cigarettes and encourage smoking cessation. This fundamental contradiction was perhaps best illustrated by the page entitled “Providing smoking pleasure and reducing harm” (PM, 2008c). PM described its efforts to “exceed the expectations of consumers” by providing smokers with “high-quality products” that they “enjoy,” in spite of the fact that according to the same page, those products “cause serious diseases” (PM, 2008c). Although the company says it is attempting to develop “consumer-acceptable products that reduce smokers’ exposure to potentially harmful compounds” (PM, 2008c), the page assumes that it is acceptable to continue to make and promote hazardous products.
Public health is increasingly concerned with corporate disease vectors: companies whose products or practices damage public health (Freudenberg, 2005; Guardino & Daynard, 2007; Markowitz & Rosner, 2002). Philip Morris’s Web site should be a concern for public health nurses because of its omissions, inconsistencies with public health messages, and ambiguities, but also because the company is aggressively positioning itself as a provider of health information. Despite the company’s claims that it is being more “responsible” in offering information about tobacco and health, some positions on the Web site directly conflicted with public health consensus. PM Chairman Szymanczyk claimed PM’s language (2005a) about T&N “accurately reflect[ed] the position of the public health community” (Szymanczyk, 2005, p. 96). But public health practitioners oppose descriptors (Borland et al., 2004; Pollay & Dewhirst, 2002). The Framework Convention on Tobacco Control recommends banning them, and 65 countries around the world, including most of Europe, have done so (WHO Report on the Global Tobacco Epidemic, 2008). In the United States, the decision in the recent suit brought by the Department of Justice against the tobacco industry prohibits them (although that decision is under appeal) (United States v. Philip Morris USA Inc., et al., 2006). Similarly, PM’s opposition to “smoking bans” and promotion of “ventilation” (PM, 2008a, 2008m) contradict public health policy recommendations (Drope, 2004; Framework Convention on Tobacco Control, 2004).
Quitting is discussed conditionally, with only weak mentions of the impact of smoking on health. Thus, unlike materials produced by public health organizations, PM’s information about quitting smoking omits compelling motivators to do so. Although the site describes cigarette smoking as addictive, it does not give any definition of the term. This is consistent with the company’s history of arguing that “addiction” means only “habit forming,” comparing tobacco use with, for example, a penchant for gummy bears (Henningfield, Rose, & Zeller, 2006). Other important omissions include definitions of “tar” and nicotine, mention of “addiction” in cessation materials, information about many serious diseases caused by tobacco, and, most significantly, specific risks for smoking-related morbidity and mortality. Smokers are likely to underestimate their personal risks from smoking (Borland et al., 2004; Dillard, McCaul, & Klein, 2006). Omission of facts about the high likelihood of smoking-related mortality may support this tendency.
The positioning of a tobacco company as a source of health information is disturbing from several perspectives. First, by acting as an information distributor, PM tailors the messages to benefit the company rather than the consumer. PM does not need to give false information to serve its purpose; it need only frame the information correctly (Entman, 1993). For example, most readers will not notice the conditional language that frames PM’s discussions of quitting, but it implicitly excludes smokers who have not yet decided to quit, and may thus be subtly discouraging.
Second, PM’s Web site points to additional issues for those who wish to evaluate the quality of health-related Web sites. Under many of the guidelines suggested, the PM site might score very well: it is transparent as to ownership, does not misinform, and provides current citations (Eysenbach, 2002). The standards proposed—even if consistently applied—may be inadequate to alert users to more subtle biases such as those PM’s site displays. Most nurses do not have the time to perform the in-depth analysis necessary to detect such biases; clearly, it is preferable to recommend Web sites of health authorities such as the U.S. Surgeon General and to point out to patients the tobacco industry’s clear conflict of interest in health matters.
Third, PM’s positioning itself as a public health resource provider blurs distinctions between authentic health-related organizations and corporate public relations. Its “partnership” with public health and “concurrence” with public health conclusions allow PM to argue that it is now “responsible” and should not be subject to punitive action or strong regulation (Balbach, Smith, & Malone, 2006; McDaniel, Smith, & Malone, 2006). In addition, PM has begun to insinuate itself into the “public health community,” regardless of whether it is welcomed (Levin, 2004). Company representatives presented a poster about the QuitAssist program at the American Public Health Association (APHA) meeting in 2005 (Largo & Schendel, 2005). Linking PM’s Web site to those of health authorities such as the WHO tacitly associates PM with them. In reality, the tobacco industry has repeatedly undermined effective tobacco control initiatives from the WHO and other organizations (Committee of Experts on Tobacco Industry Documents, 2000; McDaniel, Intinarelli, & Malone, 2008).
Fourth, PM’s Web site obscures its fundamental conflict of interest: the company cannot promote both cigarettes and public health. The health information on PM’s Web site never addresses that conflict. PM and public health practitioners do not share the most basic principles. For instance, for public health, any high-quality product by definition does not harm its user; PM calls its cigarettes “high-quality” even as it acknowledges they cause disease. PM maintains that it can neutrally deliver health information, without attempting to persuade (Balbach et al., 2006). A public health message about tobacco that does not make a case for prevention or cessation is wholly inadequate. PM does not eschew persuasion in the messages it cares most about: the cigarette advertising and promotion it spends billions to disseminate.
Nurses are trusted by the public and occupy a privileged role in providing and interpreting health information for consumers and communities. As online health resources become more ubiquitous and accessible, public health nurses should explore clients’ use of Web sites for finding information about tobacco and disease and be alert to those Web sites’ omissions and framing of issues. Public health nurses should discourage use of the PM Web site as a source of health information by the public, professionals, and students; they should point out its omissions and inconsistencies and educate communities about tobacco industry duplicity. Public health professionals should also take care not to inadvertently provide a legitimating forum for tobacco industry perspectives on health, as happened at the 2005 APHA meeting. PM Chairman Szymanczyk testified that the purpose of the Web site was to communicate “openly, honestly and effectively regarding the health effects” of tobacco use (Szymanczyk, 2005). Instead, it contributes to cognitive dissonance about the true goal of all tobacco companies, which is to maximize profits through selling deadly products.
We appreciate comments and assistance from Patricia McDaniel, Vera Harrell, and three anonymous reviewers.
Conflict of interest statement: This study was supported by the National Cancer Institute (Grants #CA095989 and CA 120138) and the California Tobacco-Related Diseases Research Program (Grant #13KT-0081). Neither author has ever received support from the tobacco industry. Ruth E. Malone owns one share each of Altria/Philip Morris-USA, Philip Morris International, and Reynolds American stock to enable her to attend shareholder meetings for research and advocacy purposes. Both authors also disclose that they served as tobacco industry documents consultants for the U.S. Department of Justice in United States v. Philip Morris et al.
Elizabeth A. Smith, Assistant Adjunct Professor, Department of Social and Behavioral Sciences, University of California, San Francisco, San Francisco, California.
Ruth E. Malone, Professor, Department of Social and Behavioral Sciences, University of California, San Francisco, San Francisco, California.