Tobacco Documents Online

I. Criteria for the Development of Health Standards 5 A. Quality of the data 9 B. Reasonableness of the risk assessment 9 C. Statistical significance 10 D. Type(s) of risk(s) presented 10 E. Significance of the risk 10 II. OSHA's Determination that Occupational Exposure to ETS Poses a Significant Risk of Lung Cancer and Cardiovascular Disease is not Supported by Substantial Evidence 10 A. The quality of the underlying data 11 1. Lung Cancer 12 a. Spousal smoking lung cancer studies 13 b. The Fontham study 15 c. Workplace exposure estimates 16 d. Workplace epidemiologic lung cancer studies 18 2. Cardiovascular disease 19 a. Spousal smoking cardiovascular studies 20 b. The Helsing Study 21 c. Workplace cardiovascular epidemiologic studies 23 B. The reasonableness of the risk assessment 24 1. Lung cancer 26 - 2 -

. ~ --- OVERVIEW.DOC - Page 3 2. Heart disease 27 3. Pharmacokinetic Modeling of ETS Exposure 28 C. The statistical significance of the risk 29 1. Lung Cancer 30 2. Heart disease 30 D. The significance of the risk 31 1. OSHA has not shown current occupational ETS exposure to result in a significant risk 31 2. OSHA has not identified the level at which exposure to ETS poses an insignificant risk 33 III. Indoor Air Quality Can Be Addressed Comprehensively Through Ventilation 35 IV. OSHA Has Grossly Underestimated the Costs Associated With the Proposed Smoking Restrictions 37 Conclusion 40 - 3 -

Reynolds objects to the Proposed Rule on the following grounds, each of which is addressed in these comments: • OSHA's determination that occupational ETS exposure poses a significant risk is not supported by substantial evidence in the record. • OSHA's conclusion that exposure to ETS under existing workplace conditions causes lung cancer and cardiovascular disease in nonsmoking workers is not supported by substantial evidence in the record. • The Fontham [Ex. 4-106] and Helsing [Ex. 9-139] studies do not provide adequate scientific or legal bases for a risk assessment to determine the risk of lung cancer and cardiovascular disease from occupational ETS exposure. • OSHA's ETS Risk Assessment does not follow the Agency's consistent approach to risk assessment in standard settings. • OSHA has not adequately evaluated or explained its risk assessment for ETS to determine whether an occupational exposure presents a significant risk. • OSHA has not determined a level at which the alleged risk from ETS becomes significant, or would be eliminated, and has failed to explain why the Proposed Rule is reasonably necessary or appropriate to substantially reduce a significant workplace risk. • The Proposed Rule would insignificant risks of harm. impermissibly regulate • OSHA dismisses dilution ventilation as an adequate ETS control, yet mandates the use of dilution ventilation to control a host of other substances. • OSHA has grossly underestimated the costs associated with the proposed smoking restrictions. Criteria for the Development of Health Standards In this rulemaking, OSHA has taken the position that no safe - 8 -

OVERVIEW.DOC Page 5 j I Reynolds is a major U.S. manufacturer of tobacco products and employs approximately 9,000 employees in over 370 buildings in North Carolina. Reynolds' Research & Development Department (the "Department") is located in a 650,000-square-foot complex known as the Bowman Gray Technical Center ("BGTC"). One of the most sophisticated technical centers in the world, BGTC contains state-of-the-art analytical capabilities. The center is staffed by approximately 500 scientists, engineers, technologists, quality assurance specialists and support staff. More than 150 of these personnel have advanced academic degrees. Reynolds' researchers have conducted advanced work in aerosol formation and analytical methodologies regarding smoke chemistry and nicotine pharmacokinetics. The Department's expertise is demonstrated by the amount of collaborative research performed with other leading researchers and the number of articles the Department publishes in peer-reviewed publications. From 1987 through 1994, Reynolds' scientists have had over 100 articles published in the areas of environmental tobacco smoke ("ETS") and toxicology alone. An even larger number of contributions have been made to conferences and professional meetings (with subsequent publication in the proceedings of such events) during the same time frame. Reynolds has applied these unique scientific and technical resources to provide OSHA with new data and analyses on the issues raised in OSHA's evaluation of whether occupational ETS exposure warrants - 5 -

For the following reasons, the spousal smoking lung cancer studies provide no rational basis for concluding that exposure to ETS causes lung cancer: • OSHA has engaged in only a cursory review of the spousal studies, and its conclusions must be dismissed as lacking an adequate foundation for determining their relevance or utility in a significant risk evaluation. This superficiality is exemplified by OSHA's classification of the spousal smoking studies as "positive," "equivocal positive trend" or "equivocal," which is simplistic, inaccurate and uninformative. See Comments of Sears/Steichen (Exhibit F), and Roth Associates, Inc., "Comments on OSHA's Analysis of the ETS Lung Cancer and Heart Disease Data," (hereinafter "Roth"). • OSHA did not evaluate the ETS and lung cancer studies based on accepted criteria for evaluating epidemiologic studies such as: relevance of study; study quality, including appropriateness and integrity of disease endpoints; representativeness of study populations; proper matching of cases and controls; consideration of and correction for confounding variables and biases; validity of exposure data; study size; proper application of statistical methodology; and existence of a valid dose-response relationship. See Comments of Sears/Steichen (Exhibit F). • OSHA omitted without explanation a number of the published spousal smoking epidemiologic studies from its analysis. See Comments of Roth. • The exposure estimates contained in the studies are indirect and imprecise and are poorly defined. See Comments of Sears/Steichen (Exhibit F), and Roth. • The majority of the spousal smoking epidemiologic studies have methodologic problems that OSHA failed to acknowledge or explain, including poor design, poorly defined study populations, questionable cancer diagnosis, and questionable exposure information. See Comments of Roth. • Spousal smoking epidemiologic studies adjust poorly, if at all, for confounding variables, including ethnicity, - 18 -

Page 10'1 finding that the employment and places of employment in question are not "safe." "But, 'safe' is not the equivalent of 'risk- free."' Id. Under the OSH Act, a workplace is not unsafe unless it presents a "significant" risk of harm. The risk must be quantified sufficiently to enable OSHA to characterize it as "significant" in "an understandable way." Id. at 646, 663. OSHA must set standards under its own standard-setting authority, not the Delaney Clause. The OSH Act does not permit OSHA to ban a substance because it may be carcinogenic. Here, OSHA never answers the real question this rulemaking is purportedly designed to address: "At what level of exposure does ETS pose a significant health risk?" Indeed, it never even asks that question.3 However, the Supreme Court made clear that OSHA was not authorized to require workplaces to be "risk-free" and therefore prohibited the setting of a standard that would result in the regulation of "insignificant risks." Id. at 649-50. other courts have confirmed OSHA's obligation to demonstrate objectively that it is not regulating insignificant risks. For example, in AFL-CIO v. OSHA, 965 F.2d 962, 976 (11th Cir. 1992), the court held that the determination of whether a new standard 3For all ETS constituents regulated by OSHA as a carcinogen, OSHA has established a permissible exposure limit. See Comments of Green (Exhibit E).

[OVERVIEW.DOC Reasonableness of the risk assessment • Do the dose-response curves demonstrate a good fit of the selected risk assessment model to the measured data, thus increasing the confidence that the model is the appropriate one? • Are the studies "well-suited" for risk assessment, e.g., is there good exposure and response information? • Do other analyses provide relatively close and consistent risk estimates? • Have confounders been adequately accounted for? • Overall, how much confidence can be placed in the risk estimates? Statistical significance • Is the observed excess risk in the study on which the risk assessment is based statistically significant? • Is the dose-response relationship derived by the risk assessment statistically significant? Type(s) of risk(s) presented • What are the nature and seriousness of the health effects being evaluated? Significance of the risk • What is the magnitude of the risk posed at current exposure levels (or if there is an existing standard, at the current exposure limit)? • What is the magnitude of the risk reduction expected at the new level? • What is the residual risk at the proposed new level? Is the remaining risk significant? • What is the level of confidence in these projections? OSHA's Determination that Occupational Exposure to ETS Poses a Significant Risk of Lung Cancer and Cardiovascular Disease is not Supported by Substantial Evidence - 14 -

- OVERVIEW.DOc Page 11 ~ is "reasonably necessary or appropriate," requires "some assessment of the level at which significant risk of harm is eliminated or substantially reduced." OSHA has incorporated these principles into Agency policy and consistently applied them in the promulgation of new standards for occupational health hazards. E.g., 56 Fed. Rce . 64023 (1991) (Bloodborne Pathogens); 52 Fed. Reg. 46168, 46230 (1987) (Formaldehyde). The Agency has acknowledged that its authority is limited to "reduc[ing] risks of average or above- average magnitude when feasible" and has agreed "that risk assessment should be put into quantitative terms to the extent 56 Fed. Re . 64023, 64037 (1991) (Bloodborne In every post-Benzene decision rulemaking undertaken by OSHA to reduce the risk from an allegedly toxic substance, the Agency has engaged in a detailed analysis of the extent to which significant risk will be reduced at the proposed (and often at alternative) significant. (Bloodborne levels and whether the residual risk is no longer See, e.g., 56 Fed. Rec. 64023, 64036 (1991) Pathogens); 57 Fed. Reg. 42102, 42182 (1992) (Cadmium); 52 Fed. Rec. 46224 (1987) (Formaldehyde). OSHA has not issued separate, formal criteria for the performance of quantitative risk assessments and the determination of significant risks in the development of its standards. However, OSHA has adopted and followed "a consistent

OVERVIEW-DOC Pa9ea 20' request by Reynolds to make the data available for further analysis. • The study population has no male representation and is not even representative of the nonsmoking U.S. female population, under-representing rural subjects and overrepresenting minorities (especially Asians). See Comments of Sears/Steichen (Exhibit F). • The absence of dose from the "risk equation" necessitates reliance upon the recall of exposures that • sometimes took place as much as decades ago, often by a surrogate respondent. See Comments of Sears/Steichen (Exhibit F), and Gradient Corp. The results reported by Fontham et al. should not be relied upon because they are inconsistent with the overall data on occupational ETS exposure and lung cancer risk. See Comments of Sears/Steichen (Exhibit F), Gradient Corp., Butler, and Roth. • The percentage of adenocarcinoma cases in the Fontham study is unusually high, possibly a reflection of abnormal demographics in the study population. See Comments of Sears/Steichen (Exhibit F). • The technique used by Fontham to detect misclassification of smoking status is insufficient, was inadequately performed and leads to a bias that elevates the observed risk estimate. See Comments of Sears/Steichen (Exhibit F), and Gradient Corp. • The Fontham study insufficiently adjusts for recall bias by failing to promulgate the use of colon cancer controls despite the evidence published in the interim report that such bias was occurring. See Comments of Sears/Steichen (Exhibit F). • The use of frequency-only matching within age categories, combined with the high sensitivity of cancer incidence to age differences, likely introduces a bias resulting in inflated estimates of risk. See Comments of Sears/Steichen (Exhibit F). • The categorization of individuals by broad race groupings fails to account for important lifestyle differences, especially among the large Asian subset of this study. See Comments of Sears/Steichen (Exhibit -20-

of ETS must be eliminated.2 OSHA has also determined that the proposed smoking restrictions are economically feasible based on estimated annual costs of $68 million, but requests comment on situations in which the restrictions may be "technologically challenging." 2In the NPR, OSHA also discusses ETS and irritation, non- malignant pulmonary disease and reproductive effects. OSHA concludes that ETS exposure may result in an increase in eye irritation and in the risk of certain pulmonary and reproductive problems. However, OSHA does not include these health effects in its "Preliminary Quantitative Risk Assessment" and does not otherwise attempt to quantitate the risk from these effects. Apparently, OSHA has determined that the available data and information with respect to these alleged health effects do not provide a sufficient basis for a quantification of the risk to workers or for a determination that there is a significant risk from these effects. OSHA certainly makes no explicit or implicit finding or argument in this regard in the NPR. Therefore, (1) there can be no justification based on the NPR for a determination in any final standard that there is a significant risk from any alleged irritation, pulmonary and/or reproductive effects from ETS; (2) any such determination would be beyond the scope of and not justified by the NPR; and (3) any such determination could not be considered a logical outgrowth of the NPR. In any event, Reynolds disputes OSHA's preliminary conclusions concerning ETS and these health effects. For Reynolds' comments regarding the portions of the NPR addressing ETS and irritation, see Comments of Walker (Exhibit A); for pulmonary disease, see Comments of Smith - Pulmonary (Exhibit B) and Coggins - Animal Studies (Exhibit C); for reproductive disease, see Comments of Swauger/Rees (Exhibit D) and Christopher P. D'Allienne, Ph.D., and James S. Smith, Jr., Ph.D., "Review of the Scientific Literature on Environmental Tobacco Smoke and Reproductive Health," (hereinafter "D'Allienne"). This and subsequent references to "Comments" refer either to Comments by scientists or engineers employed by Reynolds that are attached hereto, or to Comments submitted to OSHA by scientific consultants to Reynolds. - 7 -