Though normally we applaud good preparation in the pursuit of science, the activities described in this quote are suggestive of an attempt to influence public debate about the effect of smoking on health. Researchers suggests that the tobacco industry conducts research designed to facilitate its business interests, as well as support national and international conference “road shows” to obtain favorable exposure in the popular media and extensive political lobbying. Well known in the West, such efforts now target Asian countries; however, little is known about how the tobacco industry influences the science of secondhand smoke in Asia. This week, ASHES reviews a recent examination of the tobacco industry documents pertaining to the Asian Environmental Tobacco Smoke (ETS) Consultants Program.
Assunta et al. (2004) reviewed the tobacco industry’s Asian Environmental Tobacco Smoke (ETS) consultants program. This group based their review on the relevant tobacco industry documents made public through the Master Settlement Agreement in 1998. Initial document searches focused on several key regions concerning the consultant program: mainland China, Malaysia and Hong Kong (PRC). Interested readers can access the same data used in this research at http://tobacco.health.usyd.edu.au/site/gateway/docs/research.htm#search.
Assunta et al report that the tobacco industry documents show an attempt to steer science in a direction favorable to the industry. They provide a number of quotes to support this argument. For example, one tobacco industry document noted concisely, “Our objective is to limit the introduction and spread of smoking restrictions and maintain the widespread social acceptability of smoking in Asia” (PM 1989, p. 7). To further support their argument, Assunta et al (2004) described a series of activities employed by the transnational tobacco companies to achieve this objective, including: (1) report dissemination, (2) media exposure, and (3) lobbying government policy makers. They suggested that the recruitment, training and coordination of local science consultants was the most important activity.
The tobacco industry sought local consultants who would provide the authority for the industry position that it was not possible to generalize to Asian cultures the extant scientific evidence of the harmful effects of smoking. A good example of this strategy was the tobacco industry’s assertion that certain characteristics of the region preclude a confident conclusion that environmental tobacco smoke (ETS), otherwise known as second hand smoke, is harmful to health. The tobacco industry proposed that several Asian cultural and dietary features were confounding variables1 in the central question of environmental tobacco smoke affecting health. Some of the Asian characteristics in question include extreme heat and high humidity, incense smoke from Buddhist shrine, and highly spiced food often prepared on smoky coal burners (O'Sullivan et al. 2000). Assunta et al. further conclude that the ETS consultant program was successful in delaying and diluting smoke-free legislation in Asian but the ETS consultants program was in disarray by the mid 1990s.
The report by Assunta et al. (2004) reviews tobacco industry documents on the Asian ETS program. It revealed a major issue among tobacco industry sponsored research, which poses a threat to scientific integrity. We consider as a professional ethic that scientists should start with a null hypothesis. This strategy reduces the likelihood that subjective bias will influence the process of scientific investigation. However, Assunta et al observed the opposite from the ETS Asian program sponsored by the tobacco industry. The objective of promoting social acceptability guided the entire ETS consultants program. The subsequent strategies including recruiting researchers to fulfill tobacco industry’s self-serving agenda is simply non-scientific.
Although Assunta et al. provide ample examples of internal documents and quotes that can serve as qualitative evidence, their process of evidence collection and presentation is often subjective. An improved method would be to gather documents and create a coding system to use for systematic assessment. This system also allows for quantitative analysis of originally qualitative data. As always we encourage a more transparent and quantitative coding mechanism to compile and analyze the tobacco documents. This is true for research by the tobacco industry, as well as research that is critical of the tobacco industry.
Unfortunately, though critical of the science, Assunta et al.’s review does not contain empirical data that is necessary to evaluate the validity of ETS research sponsored by the tobacco industry. To assess, for example, whether the risk of lung cancer can be explained by other non-tobacco risk factors in the Asian context, scientists must determine whether the potential confounding variables suggested by the tobacco industry-sponsored researchers actually bias the relationship between ETS and the risk of lung cancer. Assunta et al. (2004) did not discuss this critical information. Without such evidence, Assunta et al. simply have offered their opinion, which might be accurate, about the quality of the tobacco industry-sponsored research. Regardless, we encourage unbiased scientific methodology and rigorous peer review of all research before dissemination.
What do you think? Comments can be addressed to Ziming Xuan.
1. In epidemiology studies, without controlling for a confounding variable, the relationship between an exposure variable and an outcome variable can be biased. For example, let us suppose that ETS victims have a higher risk of lung cancer than non-ETS victims. If we also find that ETS victims are more likely to be associated with eating smoked food and it is known that people eating smoked food have higher risk of lung cancer independent of ETS, eating smoked food can be viewed as a confounding variable. However if it is not true that ETS victims are related to eating smoking food, eating smoking food should not be considered as a confounder in the relationship between ETS and risk of lung cancer.
Assunta M, Fields N, Knight J and Chapman S. “Care and feeding”: the Asian environmental tobacco smoke consultants program. Tobacco Control 2004; 13 (suppl II): ii4-ii12.
O'Sullivan B, Chapman S. Eyes on the prize: transnational tobacco companies in China 1976-1997, Tobacco Control 2000; 9:292-302.
Philip Morris Asia Corporate Affairs Plan 1990-1992 December 1989 PM2500084000.
Rupp JP, Billings DM. Asia ETS Consultant Status Report. 14 Feb 1990. PM2500048976/8998