The Tobacco Industry's Response to the COMMIT Trial: An Analysis of Legacy Tobacco Documents

SYNOPSIS

We analyzed internal tobacco industry documents that describe the industry's response to the Community Intervention Trial for Smoking Cessation (COMMIT), a multi-center community-based tobacco intervention project funded by the National Cancer Institute from 1988 to 1992. Our analysis of documents from the Legacy Tobacco Documents Library (www.legacy.library.ucsf.edu) suggests that the tobacco industry reacted to COMMIT by (1) closely monitoring trial activities, (2) confronting COMMIT in communities where it was most active, (3) distorting COMMIT findings on underage smoking data reported in the media, and (4) using COMMIT activities as practice to strengthen their attack against the subsequent ASSIST trial, falsely accusing both studies of illegal political lobbying with taxpayers' money.

The tobacco industry closely monitored COMMIT activities and organized local responses to findings and activities perceived as threatening to the industry's public image or interests. Although we could not document a concerted attack by the tobacco industry that impacted the results of the COMMIT trial, data suggest that the industry used COMMIT as a learning opportunity to mount a well orchestrated and potentially damaging response to the larger American Stop Smoking Intervention Study for Cancer Prevention Trial.

The production and dissemination of scientific knowledge on health risks of tobacco use and exposure have been the focus of attacks by the tobacco industry since the early 1960s. The industry has used a number of tactics to undermine the science on health consequences of smoking,¹ including: (1) creating doubt about the credibility of the evidence by paying scientists to disseminate misinformation about the harms of tobacco;² (2) generating confusion, in newspaper articles, letters, and other forums, about the meaning of statistical significance and research methods of studies;^2–4 (3) creating “research” forums to get pro-tobacco information into the scientific literature;^5,6 (4) influencing publications by paying journal editors to write editorials opposing tobacco restrictions;⁷ (5) hiring scientists to provide biased expert testimony in tobacco litigation and regulatory hearings; and (6) countering legitimate scientific findings by funding scientists to conduct biased research and carry out studies designed not to achieve significant results.^6–9

Less is known about the tobacco industry's strategies to undermine science in areas outside the health effects of tobacco exposure, such as community-based randomized trials to test the efficacy of public health measures. A recent paper^10,11 documented a well coordinated attack from the tobacco industry on the American Stop Smoking Intervention Study for Cancer Prevention (ASSIST), the largest and most comprehensive tobacco control intervention trial ever conducted in the United States. Strategies to undermine ASSIST included close scrutiny of and interference with local ASSIST activities, utilizing individuals and organizations secretly linked to the tobacco industry to generate false allegations of misusing federal funds, and direct lobbying of state and federal legislators to conduct stringent audits of ASSIST coalitions and increase limitations on use of funds.^10,11

Studies of the tobacco industry's interference with community-based trials of smaller magnitude have not been published to date. The Community Intervention Trial for Smoking Cessation (COMMIT) is one of the pioneer examples of a community-based study in the U.S. and was both the predecessor of ASSIST and the second largest tobacco trial ever conducted in the United States. COMMIT was designed to reach and influence cigarette smokers through community mobilization and change in social norms.¹² The trial used 11 matched pairs of communities (intervention and control), 10 of them in the U.S. and one in Canada. According to a recent publication from the National Cancer Institute about the trial, the impact of COMMIT's community organization approach on smokers' behavior was modest, at least for the first four years of the intervention. Although no change was noted in the target group of heavy smokers, there was a statistically significant difference in the quit rates between intervention and comparison communities among light-to-moderate smokers.^13,14 The mobilization and organization of communities around an externally defined public health objective is considered by many to be the real accomplishment of the COMMIT program.^13,14

We analyzed internal tobacco industry documents that describe the industry's response to the COMMIT program, funded by the National Cancer Institute from 1988 to 1992. These documents were made available through legal settlement agreements in the 1990s and are considered by many to be the most valuable outcome of lawsuits against the tobacco industry in the past decade.¹⁵ As of September 2005, the online Legacy Tobacco Documents Library contained more than 7 million documents (40 million pages) from 10 different collections.¹⁶

METHODS

We searched documents from the Legacy Tobacco Documents Library (www.legacy.library.ucsf.edu) made available as a result of tobacco litigation during the late 1990s. Search terms used included “COMMIT” + “trial”; “COMMIT + confidential”; “COMMIT + lobbying”; “New Anti-Tobacco program”; “Heavy Smokers Trial” (first name used by NCI for the COMMIT trial); and the names of all 20 U.S. towns involved in the COMMIT trial (control and intervention sites). Using a “snowball” searching technique,¹⁷ we also searched for names of tobacco industry staff and lawyers cited in the documents obtained and adjacent Bates numbers (assigned to tobacco industry documents, such that each page has a unique identifier).

We obtained 1,829 documents from this initial search. After reading each article, we found that 1,712 were not related to the COMMIT trial (e.g., invoices, advertisements, letters discussing stocks, etc.); 86 were related to the COMMIT trial but not of relevance to the study (e.g., scientific papers, newspaper clips, duplicates of other documents); and 32 were related to the COMMIT trial, involving the industry's reactions to the study either locally or nationally.

These documents were read by two researchers (BC and VS), classified by themes that emerged from the search, and assembled chronologically into a narrative case study. Any discordance between the researchers regarding classification of the documents, interpretation of their meaning, or significance to the study were discussed and resolved by consensus.

RESULTS

The documents analyzed fell into four thematic categories that describe the tobacco industry's reaction to and subsequent actions regarding COMMIT: (1) close monitoring of trial activities; (2) confrontation of COMMIT in its most active communities; (3) distortion of COMMIT findings in the media; and (4) lodging false allegations of political lobbying against COMMIT.

Monitoring of COMMIT activities

Tobacco companies and their public relations and lobbying organization, the Tobacco Institute, had lawyers, staff, and other informants closely monitor COMMIT activities. Brown and Williamson Tobacco Corporation (B&W) was aware of the trial before it was implemented and expressed concern about the possible impact of COMMIT activities on their advertising plans. A 1998 internal memo among B&W staff states:

This program (COMMIT) will begin this fall in 11 cities. Our current advertising plans for this fall will include billboards for Richland's in Utica and billboards and/or newspaper ads for Capri in Yonkers, Utica, Raleigh and Cedar Rapids.

Considering the publicity surrounding this program, we will need to respond to any inquiries concerning this issue.

I would like to discuss this article with each of you at your convenience.¹⁸

(The cities of Utica, Yonkers, Raleigh and Cedar Rapids were intervention sites for COMMIT.)

Three weeks later, the internal discussion at B&W resulted in another memo:

I discussed this program with Tom Humber. He expressed concern that we may experience a diminishing return of our advertisement dollars in the cities which participate in the COMMIT program. However, he agreed that we should not discontinue advertising in these cities especially since our current campaigns are not controversial and can be easily defended.

If any issues arise, I will notify Mr. Humber and our Corporate Affairs Department. The Law Department will also monitor these cities and follow closely any legislation which may occur because of the COMMIT Campaign.¹⁹

(Tom Humber was the vice president of Burson-Marsteller Inc., a company hired by Philip Morris to handle public relations and public affairs.)

The Tobacco Institute was also aware of COMMIT since the trial's early stages. The following is a memo to senior staff from W. Kloepfer, vice president of public relations for the Tobacco Institute:

For your information, at a meeting of the Board of Scientific Counselors of the Division of Cancer Prevention and Control at NCI the other day, funding of $42.5 million was approved for the Community Intervention Trial for Smoking Cessation (COMMIT). During discussion a question was raised as to whether tobacco companies might attempt to increase smoking in the intervention communities to skew the program. One participant suggested that if this occurs, a counter weapon might be a boycott against food products owned by tobacco companies.²⁰

Although no evidence was found that Philip Morris was keeping track of COMMIT at its inception, the company obtained detailed information on the trial a few years later, presumably using federal Freedom of Information Act (FOIA) requests (which became a central tactic in the tobacco industry's later attacks on the ASSIST trial). A fax sent to Steven Parrish, vice president of Philip Morris USA, from the law offices of Shook, Hardy & Bacon, states:

Dear Steve: In August 1991, we requested from the National Cancer Institute records pertaining to the Community Intervention Trial for Smoking Cessation (COMMIT). I believe it was John Nelson at Philip Morris who asked you for this material. NCI produced about 200 pages of documents and I sent those to you in November. Now, 13 months after our request, NCI has produced another 1,900 pages …²¹

Confrontation at the local level: Utica, New York

Utica, New York, was both a COMMIT intervention site and one of the cities where RJ Reynolds (RJR) had a Smoker's Rights (SR) program, part of their plan to build a coalition to voice pro-smoking messages in the late 1980s. It did not take long for the local groups that represented these forces—Utica-COMMIT and SR-Utica—to be faced with each other's contrasting views on smoking. This confrontation took place at the Sheraton Inn on March 27, 1991. Utica-COMMIT had decided to hold a meeting at the same time, venue, and date that an SR meeting was taking place, as a strategy to promote debate and voice public health messages to the community (Dr. Michael Cummings, personal communication, May 5, 2005).

The situation created by having these two groups with conflicting agendas confined to the same space was described by Steve Caldeira (RJR's contractor for developing SR groups in several cities) in a memo to Rob Meyne, RJR's coordinator of their SR program:

Needless to say it was a zoo…. At around 6:30 pm partisans started to enter the meeting room…. At about that time 10–15 young teenage kids arrived outside the meeting room. They were passing out ANTI-RJR material entitled, ‘Smokers !!—You Have a Right to Know!!’ This material basically stated that “one in every 6 deaths in America was caused by cigarettes. RJ Reynolds, the sponsor of tonight's ‘Smokers’ Rights' meeting, is the second largest manufacturer of cigarettes in the U.S. In 1988, R.J. Reynolds' cigarettes killed 128,950 Americans.”

On top of that, allegedly a 16 year old girl (Julie Lewis) slipped into the meeting and stayed for a few minutes and grabbed a Smokers' Rights kit.²²

This provocative incident generated two articles in the Utica Observer Dispatch, described by Steve Caldeira, in the same memo, as “not overly favorable” to his work. It also precipitated the decision to suspend subsequent SR meetings in the same hotel, changing its venue to “the Holiday Inn in New Hartford, NY, which borders Utica.”²² This initial confrontation prompted RJR to recognize that COMMIT posed a local challenge to their plans, as stated in another one of Caldeira's memos:

It is quite evident that the Utica COMMIT Program has taken its effect on this city of approximately 50,000 people…. They have been flooding the airwaves (TV & Radio) with anti-smoking spots off and on for the last 3 years. The Utica Observer-Dispatch provides carte blanche access for any and all articles that they choose to submit.²³

This recognition was followed by measures to protect RJR from incidents like the one described above. In the same report, Steve Caldeira wrote:

Due to the overall lack of interest in this area I am abandoning my usual format of monthly meetings and electing officers. The group will try to meet again in September. However, I do have 5–7 people who are committed as heck…. The group members will continue to write and collect petition signatures…. Most important, however, they will serve as key contacts for me in the area, as well as monitor COMMIT activities…. In addition, Ivan Yost in particular, as well as Barbara Ann Zeizio have been successful in getting ‘Letters to the Editor’ published.²³

The letters Caldeira is referring to were planted as criticisms from Utica residents of COMMIT activities. For example, Ivan Yost sent a letter to the City Council President of Utica on May 16, 1991:

I wish to protest, emphatically, over City Youth Bureau participation in COMMIT-sponsored ‘media events.’ This is the second time that Utica-COMMIT has either stepped over the law or come close to it, using young people on each occasion, to further their crusade's purpose as they view it…. But let me remind you, we are dealing with a legal substance, and various legislatures have seen fit to place certain restrictions on the conditions of its sale and use. Nothing in any of these laws calls for the sort of vigilant enforcement of the kind that COMMIT seems to be practicing. If and when our society decided to take the extreme action that COMMIT and similar crusaders seek, we shall have arrived at the Promised Land, when the profession of bootlegging can once again made to prosper.²⁴

The clash between COMMIT and the SR group also resulted in measures involving tracking of documents, activities, and names of COMMIT activists in Utica. As the following memorandum illustrates, these investigations were aimed at exploring the possibility of legal action against COMMIT:

Attached is a copy of information from ‘Utica Commit,’ the anti-smoking program whose members distributed information on RJR at our smokers' rights meeting at March 27. Note this is not the information that was handed out at the SRM. This is the packet they send to people who phone in and express interest in the organization. It seems to be well done….

Janine Sadlik is listed as ‘community coordinator.’ She is the person who got on our database in May of 1989, said she was a smoker, was opposed to smoking restrictions, was interested in attending a smoker's rights meeting, and even requested a copy of our youth smoking booklet. So, she not only was savvy enough to get on our list in the first place, but also communicated to us information she knew would be likely to cause us to keep her on the list as a hot prospect.

It was her address, as listed on the database, that was given by Julie Lewis, the sixteen year-old who attended part of the SRM and subsequently appeared in television and radio news reports.

Also attached is a copy of the piece attacking RJR that Commit handed-out to people at the hotel. Mike Johnson has a copy of this and has promised to give it full consideration from a legal standpoint. He is not, however, encouraging in this regard.²⁵

Finally, as the information disseminated to higher levels within RJR, the degree of concern triggered by the Utica incident led to discussions at RJR about developing an “anti-smokers database,”²⁶ as indicated in an internal letter:

… we don't want to be in a position of developing a Richard Nixonesque ‘enemies’ list…. The existence of such list could, in and of itself, be a negative P.R. story…. The reasons for putting together such a database are simple: one, to insure our communications don't go directly to antis, thus making it harder for them to pull off stunts such as they did in Utica (though hardly making it impossible), and, two, to keep our overall themes and strategies held somewhat close to the extent possible…. I guess my bottom line recommendation, after further thought, would be that we encourage the Field Coordinators to collect names of known antis as they come across them, and delete from our database those who prove to be a problem for one reason or another (e.g., Janine Sadlik in Utica).²⁶

Distortion of COMMIT findings on underage smoking

When COMMIT released a study showing that minors had a strong preference for Camel cigarettes, RJR was already prepared to defend its leading brand, having obtained privileged information before the official results of the survey were released to the public. In January 1992 a letter from KL Verner (Marketing) to MT Payne (Public Relations) of RJR says this:

… We have been advised that interim results from the subject study will be released shortly and will indicate that Camel's share of underage smokers is 29–30%…. We should plan on discussing this information in the near future to ensure we are prepared to respond if necessary.²⁷

Verner's concerns were based on the fact that the 30% underage preference for the Camel brand found by COMMIT in 1991 was very different from the 8% found in the Center for Disease Control and Prevention's (CDC's) 1989 Teenage Attitudes and Practices Survey (TAPS). Given the fact that the Camel campaign had taken effect in 1990, and its cartoon camel seemed to be designed for and particularly appealing to minors, it is not surprising that RJR wanted to be prepared to respond to the COMMIT findings. The response articulated by RJR is outlined in a document wherein COMMIT findings are re-calculated:

• COMMIT Analysis

  • 4,129 students
  • 424 purchased their own cigarettes
  • 30% of 424 students = 127
  • 127 out of 4,129 students have bought Camel in the last 30 days
  • 127 / 4,129 = 3%

• Conclusion: 3% of ninth graders in COMMIT study have bought a pack of Camel's in last 30 days.²⁸

This recalculation formed the basis of RJR's central argument to counter the media dissemination of COMMIT results. Peggy Carter, RJR's manager of media relations, wrote a letter to Rod Little, illustrator for USA Today, in April 1992:

On March 25, you charted statistics from a recent article on the Community Intervention Trial for Smoking Cessation (COMMIT) smoking study…. The heading indicates the figures represent ‘Percentage of brand use self-reported by 9th graders in 10 U.S. communities in 1990,’ which gives the impression that a large number of kids are smoking, and smoking the brands indicated. In fact, those figures represent only the students in the survey who purchase their own cigarettes, or about 10% of the total sample….

There were 4,129 9th graders respondents from 13–16 years of age in the 10-city COMMIT survey. The survey considered a student a smoker if he or she had smoked one cigarette in the last 30 days, an obviously overly inclusive standard. Out of that sample, 424 indicated they usually purchase their own cigarettes; 30% of those students said they bought Camel. Therefore, 3% of all the ninth graders responding to this survey said they bought Camels in the last 30 days, a fact that puts youth smoking statistics related to Camel into a much more accurate and clearer perspective.²⁹

Learn from COMMIT and target ASSIST

ASSIST was the biggest tobacco control trial ever conducted in the U.S. (1991–1998) and was thus considered a major threat to tobacco industry interests and attacked accordingly.^10,11 In the eyes of NCI, ASSIST was considered to be the successor to COMMIT:

… COMMIT's strategy, which ASSIST will extend throughout entire states, recognizes that there are powerful social forces affecting people's use of tobacco in a community and that the community itself must be mobilized to make smoking socially unacceptable…. The evaluation of COMMIT is ongoing. Although the final evaluation will not be complete until mid-1993, some evaluation is now available and will be used to shape the intervention plans for ASSIST…. ASSIST's 2-year planning phase allows emerging information from COMMIT to be fully integrated into ASSIST activities. The intervention phase of ASSIST will not begin until the COMMIT intervention is complete.³⁰

This connection between the two trials was proactively explored by the tobacco industry as the Tobacco Institute began compiling information on COMMIT that could be used against ASSIST a year before its 1991 start date. A Tobacco Institute memorandum in August 1990, sent internally and copied to Tobacco Institute headquarters, expresses the intention of exploring this issue at an upcoming lobbyists' meeting session:

As we discussed yesterday, the National Cancer Institute is in the middle of a ten-year experimental program to determine the effectiveness of various community-based programs in smoking cessation and reduced smoking among young people. Carol Hrycaj has been watching this program and provided the attached publication. The COMMIT program, and its potential for local legislative action is probably worth mentioning at the lobbyist meeting's session on youth issues, along with information on the new ASSIST program being started by NIH.³¹

After briefly describing COMMIT's methodology, the memorandum continues:

Only someone locally could tell us whether recent anti-smoking legislation in Leominster, Santa Fe, and Raleigh was prompted by COMMIT program activities. I think our field should be made aware of the potential for smoking, advertising, sampling and sales restriction legislation in these particular localities with an eye toward such proposals developing momentum in nearby communities or the state legislature. The attached publication includes more information on COMMIT and a little bit about ASSIST.³¹

It was more than one year later that this link between COMMIT and ASSIST was revived, in the wake of the Tobacco Institute's efforts to pass an amendment to the National Institute of Health's authorization bill that explicitly prohibited use of grant funds to influence state or local legislation. A TI memorandum of October 7, 1991 says:

As we have discussed, the following actions may be undertaken to counter potential state and local legislative action resulting from grants to 17 states for the American Stop Smoking Intervention Study (ASSIST).

One of the 13 measures described in the memo related to COMMIT:

Activities in ten cities under NCI's predecessor study, Community Intervention Trial for Smoking Cessation (COMMIT), are being reviewed by State Activities. Particular emphasis is on Raleigh, NC; Santa Fe, NM; and Leominster, MA, where anti-tobacco legislation has been considered in the last two years. Evidence of pursuit of such legislation by grantees will be useful in push for above amendment.³²

A few days later, progress was made by the Tobacco Institute in the direction described above. The use of anti-COMMIT strategies to pro-actively counter ASSIST activities is described clearly in an internal memo:

If we can illustrate to members of congress that those receiving funds from the ASSIST program may well plan to use them for inappropriate lobbying at the state/local level, Congress will be more likely to tighten the rules about how these funds are disseminated and used in FY93 and beyond.

The point was brought home yesterday when our federal people shared with Democratic congressman newspaper clips on the COMMIT program in Raleigh. The articles and other documents clearly showed local lobbying with federal government funds. The reaction from the member was that this was a poor use of COMMIT money, at least in his eyes, and he would work to ensure that ASSIST funds could not be used in this manner.

In a nutshell, we would like to be in a position to tell key members of congress that unless the rules are tightened ASSIST funds will be used in the same questionable manner as COMMIT funds have been used—to lobby for and against specific legislation at the state/local level.³³

This document had a report attached called “Commit Activities in Legislative Arena,” which included four pages describing COMMIT's objectives related to local legislation change, total U.S. dollar amounts granted to each community, and evidence of successful restrictive ordinances.³⁴

The tactic of scrutinizing COMMIT activities to demoralize ASSIST persisted over the years, but took a different twist after the underwhelming results of COMMIT's evaluation became public and ASSIST activities were underway. In a document authored by the Tobacco Institute and retrievable in Philip Morris, Lorillard, B&W, and RJ Reynolds collections, the National Cancer Institute is accused of giving federal tax money “to anti-tobacco activists who lobby for ‘tobacco control’ legislation.” The last paragraph of the three-page report, entitled “Federal Tax Funding to ‘ASSIST’ State Tobacco Control Lobbying” attacks COMMIT in a bold new way:

Finally, the entire premise of ASSIST is based on an earlier NCI program, known as COMMIT, which included local lobbying for anti-smoking laws. The final evaluation of COMMIT was published this year. It reported that $45 million was spent in a four-year program that made no change in smoking rates for ‘heavy smokers’ and produced only a 3-percent higher quit rate for ‘moderate smokers’ compared to the national average. ASSIST is throwing good money after bad.³⁵

DISCUSSION

Our analysis documented that the tobacco industry was not indifferent to the COMMIT trial, but responded to it in a fragmented way. COMMIT was the first national trial based on community organization as a means to promote policy and environmental change to decrease tobacco use and initiation. As COMMIT represented a novel approach, it is possible that the industry was caught somewhat unprepared to respond to COMMIT in a more organized fashion, and as a result, responded with some local actions in an almost improvised way.

It should also be noted that as much as COMMIT was a learning opportunity for researchers to test community-based intervention trials, it was also an opportunity for the tobacco industry to “rehearse” ways to counter initiatives such as these. The tactics used by the industry against COMMIT, as described here, were used in a coordinated and amplified way starting in the early phases of ASSIST, and succeeded in disrupting, if not actually undermining, the trial.^10,11 Moreover, as demonstrated in this article, the interest that the tobacco industry displayed in COMMIT was revived when it became a source for strengthening their arguments to lobby against ASSIST in the legislative arena.

The COMMIT trial also became a focus of the industry because of data collected on underage smoking. RJ Reynolds' attack on the COMMIT findings regarding teenage preference for Camel cigarettes should be understood in a broader context. RJ Reynolds was already concerned with studies carried out by DiFranza and colleagues on the change in Camel's share among different age-populations.³⁶ Before 1990, Camels were most popular with smokers over the age of 65 years, and chosen by less than 1% of underage smokers; three years later Camel became one of the top two brands preferred by adolescents, capturing nearly a third of the market share in this age group. This phenomenon was coincident with RJR Nabisco's “smooth character” advertising campaign, which began in 1988 and featured Old Joe, a cartoon modeled after James Bond, and Don Johnson of televison show Miami Vice. In their study, published in JAMA in 1991, DiFranza et al. reported a finding very similar to the one described in COMMIT: 32.8% of teenagers preferred Camel. They also found a higher rate of recognition, recall, and appeal of the Camel brand in adolescents, compared to adults, and concluded, “in just 3 years, Camel's Old Joe cartoon character had an astounding influence on children's smoking behavior.” A preference for Camel among teenagers found in the COMMIT study was additional bad news for RJR, and as expected, the company devised an alternative explanation for the finding.

Finally, it is noteworthy that a local incident like the one described in Utica, NY, became the center of a discussion on “antis,” as described by RJR. The fact that tobacco activists turned the tables and used the same tactics that the tobacco industry has been using for so many years³⁷ was upsetting for RJR. There are indications that other tobacco companies may have been similarly affected by COMMIT at other sites; a document from Philip Morris for its 1991 stockholders meeting identifies 20 pages of “antis,” both individuals and organizations. The name of Meg Houseworth, a member of the coalition board of COMMIT-Raleigh (North Carolina) appears on this list.³⁸ While no further tobacco documents were found to explain this special attention from Philip Morris, it is reasonable to consider that something akin to the activities documented in Utica may also have occurred in Raleigh. Other sources could be used in the future to further explore this possibility.

Some limitations of this study should be acknowledged. First, despite the vast number of tobacco industry documents compiled by the Legacy library, there is no way to ensure that all key documents related to COMMIT were retrieved, given the inconsistent indexing of the material available.¹⁰ It is possible that other researchers may have found or utilized different materials than we did, and thus may have come up with different findings and conclusions. Another limitation is the use of just one source, the Legacy Tobacco Documents Library. Analysis of local newspaper archives at COMMIT sites, interviews with professionals and community activists, and other industry archives are resources that could have offered additional information and insights into our analysis. Nonetheless, despite these limitations, the Legacy Library has been the sole resource used in several other papers that have advanced our understanding of tobacco industry tactics to undermine tobacco control efforts and scientific knowledge about the harms of tobacco.

In conclusion, our study did not find sufficient evidence to support an orchestrated attack by the industry against COMMIT on the scale of that found later against ASSIST. However, the documents we analyzed make it clear that the tobacco industry was interested in the trial, and committed considerable resources and attention to investigating this project. As is true with the ASSIST case, the impact of the tobacco industry's attacks on COMMIT outcomes remains unknown and is beyond the scope of our analysis.

Considering the findings of this study and others that have taken this approach, it is apparent that any significant research conducted in the tobacco control field is likely to face the added challenge of overcoming interference from the tobacco industry. It is hoped that studies like this one may help researchers and policy makers effectively counter the industry's ability to derail public health efforts to reduce the tremendous toll of tobacco on our society.