Tobacco Documents Online

UNITED STATES ENVIRONMENTAL PROTECTION AGENCY COMMENTS OF THE TOBACCO INSTITUTE ON HEALTH EFFECTS OF PASSIVE SMOKING: ASSESSMENT OF LUNG CANCER IN ADULTS AND RESPIRATORY DISORDERS IN CHILDREN REVIEW DRAFT October 1, 1990

TABLE OF CONTENTS Page EXECUTIVE SUMMARY ........................................... 2 A. The Draft Report Improperly Uses the Risk Assessment Process for the Purpose of Restricting Tobacco Use Rather Than Objectively Assessing Risk ................ 5 1. The advocacy nature of the draft report undermines its utility as a risk assessment .................. 5 2. EPA's decision to prepare the ETS risk assessment, despite acknowledged data gaps, further undermines the document's credibility ........................ 12 B. The Data Summarized in the Draft Report Do Not Support the Conclusion that ETS Is a Cause of Lung Cancer ...... 16 1. The use of epidemiology to support EPA's position that ETS causes lung cancer is severely flawed.... 17 a. The Guidelines` criteria for the use of human studies have not been met .................... 18 i. Bias .................................... ii. Confounding factors ...................... 19 23 iii. Statistical significance ................ 25 b. Relevant studies are unfairly dismissed...... 25 C. EPA's meta-analysis is faulty ................ 28 2. EPA's dose-response assessment is similarly flawed 30 a. Dose is poorly defined in the case-control studies ...................................... 32 b. Response data vary from study to study ....... 34 3. EPA's weight-of-evidence analysis is inconsistent with the Guidelines ............................... 35 4. EPA's hazard assessment conflicts with its approach to other substances ............................... 40 C. Data Gaps and the Draft's Failure To Acknowledge Them 00 Undermine the Result ................................... ~ 43 1. Lack of biological and chemical information....... tW 44 tD 2. Lack of exposure assessment ....................... N 47

Page D. The Assessment of Respiratory Disorders in Children Is Also Written as an Advocacy Document and Not as an Objective Assessment of Risk ........................... 50 CONCLUSION .................................................. 53 TABLE 1 ..................................................... 56 EXHIBIT A ................................................... 57

UNITED STATES ENVIRONMENTAL PROTECTION AGENCY COMMENTS OF THE TOBACCO INSTITUTE ON HEALTH EFFECTS OF PASSIVE SMOKING: ASSESSMENT OF LUNG CANCER IN ADULTS AND RESPIRATORY DISORDERS IN CHILDREN REVIEW DRAFT The Tobacco Institute submits these comments on the draft report entitled "Health Effects of Passive Smoking: Assessment of Lung Cancer in Adults and Respiratory Disorders in Children, Review Draft" ("draft report"). The draft report has been submitted to the EPA Science Advisory Board ("SAB") for review and comment. These comments are being submitted to the SAB for consideration. In addition, we have asked a number of independent scientists with expertise in the relevant disciplines to comment on the draft report. These expert reviews also are being filed with EPA and will be submitted to the SAB. They will be referred to in these comments as appropriate.

Page 2 EXECUTIVE SUMMARY The draft report cannot accurately be called a risk assessment. If it is intended to be an objective, scientific assessment of risk, the draft is seriously deficient. If it is intended instead as an advocacy document, seeking to encourage smoking restrictions, the document and its short- comings can be understood. The selective inclusion of data in the draft, and the questionable techniques used to extrapolate a relative risk for environmental tobacco smoke ("ETS"), demonstrate that the draft was shaped to justify a prede- termined policy. The bulk of the draft report is presented as a carcinogen risk assessment. Its evaluation of the conflicting epidemiologic data fails, however, to meet the threshold requirements of EPA's Guidelines for Carcinogen Risk Assessment. The fact is that the leading United States epidemiologic studies have found no statistically significant association between exposure to,ETS and lung cancer. The authors therefore conduct a meta-analysis to merge selected, dissimilar studies and apply arbitrary criteria to arrive at a positive relative risk. The use of meta-analysis in these circumstances is grossly unscientific and conflicts with the Agency's established principles for the evaluation of human studies, as well as with the recent warning by the Office of Management and Budget ("OMB") against the improper merging of risk management and risk assessment by federal agencies.

Page 3 Among the core problems with the risk assessment are the following: The clear majority of the international epidemiologic studies have found no statistically significant increase in lung cancer incidence associated with ETS, yet the draft relies disproportionately on an early Japanese study whose methodology and conclusions have been widely questioned. The United States studies find no statistically significant association. A claimed dose-response trend is not supported by reliable exposure data and is indeed contradicted by the results of a number of studies, including the American Cancer Society's cohort study, which reported an inverse dose-response relationship. Studies of varying quality, involving widely varying definitions of exposure, procedures, methods of analysis, and underlying racial and cultural populations, are merged, while flaws due to probable bias, confounding and misclassification are ignored or arbitrarily adjusted. No biochemical or exposure assessment of ETS is presented, despite EPA's previous statements that it was preparing such assessments, nor are any positive animal data offered. The unprecedented nature of the draft report and the Group A carcinogen classification it advocates becomes unmistakable when the draft is compared with existing risk assessments for Group•A carcinogens and EPA's assessment of other agents it has refused to classify as such. No other Group A substance has been so classified in the face of a clear majority of epidemiological studies finding no statistically significant association, and no other substance has been so classified without positive data from relevant

Page 4 animal studies. On the other hand, EPA has refused to classify substances with a much larger -- and positive -- human study data base, such as electromagnetic fields, because it recognized that data that were far stronger than are available for ETS were nevertheless inconclusive. Acceptance of the draft would create a precedent under which a Group A carcinogen designation is called for whenever a meta-analysis yields a slightly elevated positive relative risk -- regardless of the quality of the underlying studies, even when only a small minority find a statistically significant association, and without any positive animal studies. Applying these criteria, literally thousands of substances encountered in everyday life would have to be classified as Group A carcinogens. The draft ETS risk assessment lacks the fundamental objectivity that has characterized EPA's prior risk assessment efforts and that has been called for by OMB. When the EPA Carcinogen Risk Assessment Guidelines are properly applied to the ETS data, the Agency must classify ETS in Group D -- that is, not classifiable as to human carcinogenicity. Consistent with that classification, we urge that EPA acknowledge that it would be premature to conduct an ETS/lung cancer risk assessment at this time. The final chapter in the EPA draft report is essentially a selective literature review of studies of respiratory disorders in children, presented again in advocacy

Page 5 form to support smoking restrictions. While recognizing that the studies are too equivocal and flawed to support a causal inference, the draft goes on to urge that ETS nevertheless be deemed a "risk factor" for these disorders. In view of the scientific flaws and inconsistencies in the studies, the draft's summary conclusion that ETS should be considered a "risk factor" is entirely inappropriate. A. The Draft Report Improperly Uses the Risk Assessment Process for the Purpose of Restricting Tobacco Use Rather Than Objectively Assessing Risk While claiming to be a genuine risk assessment, the draft report is instead another step in a predetermined strategy to support smoking restrictions. By preparing a risk assessment document for advocacy purposes, the authors have violated the fundamental principle that the process of risk assessment must be carefully distinguished from the process of risk management. The draft report also conflicts with EPA's previously expressed acknowledgment that much additional research in a number of areas is needed before any conclusions concerning ETS can properly be drawn. 1. The advocacy nature of the draft report undermines its utility as a risk assessment It is axiomatic that the cornerstone of a risk assessment is its scientific objectivity. The scientific and factual inquiry into the existence, if any, of a risk must be separated from the management and policy process of

Page 6 controlling risk. The National Research Council ("NRC"), in its 1983 report on Risk Assessment in the Federal Government: Managing the Process, reiterated this principle in its first two recommendations (pp. 151, 153): Recommendation 1 Regulatory agencies should take steps to establish and maintain a clear conceptual distinction between assessment of risks and the consideration of risk management alternatives; that is, the scientific findings and policy judgments embodied in risk assessments should be explicitly distinguished from the political, economic, and technical considerations that influence the design and choice of regulatory strategies. * * * Recommendation 2 Before an agency decides whether a substance should or should not be regulated as a health hazard, a detailed and comprehensive written risk assessment should be prepared and made publicly accessible. This written assessment should clearly distinguish between the scientific basis and the policy basis for the agency's conclusions. In its 1990 Report on the Regulatory Program of the United States Government ("OMB Report"), the Office of Management and Budget affirmed the need for close adherence to these NRC recommendations, reprinting them in full (see OMB Report, p. 14). OMB also stressed that "[t]he need to keep risk assessment and risk management separate has long been the objective of responsible public officials" (id. at 13). OMB then pointed out (id. at 14): Unfortunately, risk-assessment practices continue to rely on conservative models and assumptions that effectively intermingle important policy judgments within the

Page 7 scientific assessment of risk. Policymakers must make decisions based on risk assess- ments in which scientific findings cannot be readily differentiated from embedded policy judgments. This policy environment makes it difficult-to discern serious hazards from trivial ones, and distorts the ordering of the Government's regulatory priorities. In some cases, the distortion of priorities may actually increase health and safety risks. As former EPA Administrator Ruckelshaus.has noted, in comments repeated by OMB in its most recent Annual Report on the Regulatory Program of the United States (OMB Report) (1990) (p. 15; original emphasis): Risk assessment * * * must be based on scientific evidence and scientific evidence only. Nothing will erode public confidence faster than the suspicion that policy considerations have been allowed to influence the assessment of risk. Traditionally, EPA has applied these principles rigorously. Carcinogen risk management decisions have been made only after a risk assessment has been prepared. Indeed, under EPA's authorizing statutes, regulatory decisions must be based upon an objective evaluation of relevant factors, including risk, and policy decisions must be reserved until after the risk has been characterized and, in some cases, quantified. In the case of ETS, however, risk management decisions inappropriately have preceded the preparation of the ~ draft report. The quantitative risk assessment was prepared not to determine objectively whether there is a risk and if W ~ , G1 0