Tobacco Documents Online

24 Mr. RosE. Mr. Bishop. Mr. BISHOP. Thank you very much, Mr. Chairman. The release of this risk assessment of environmental tobacco smoke has really been harshly criticized by a number of independ- ent scientists. I have a serious concern with reports that the Agen- cy staff may have ignored the universally accepted standards of sci- entific evidence in order to justify its position. So I really want to focus on the scientific review process to make sure that we have a scientific determination here rather than just a policy review to rubber stamp a preconceived idea. I would like to ask a few questions and maybe get some re- sponses. There have been some allegations that the Science Advi- sory Board panel that reviewed the risk assessment was comprised of a number of individuals that had conflicts of interest and that they had obvious biases against smoking. There are also allegations that some of them had actually been involved in preparing the document that they were asked to re- view. Also there were allegations that a number of well-funded re- cipients of EPA grants were included in the Science Advisory Board panel that conducted the review. Also, there are allegations that the EPA staff had engaged in be- hind-the-scenes maneuvering in order to stack the panel in favor of the Agency's position. I don't know whether that is true or not, but if those kinds of Agency allegations are out there, I think it is incumbent, particularly with the tremendous ramifications of these findings, or these alleged findings, on the environment throughout our country, and not to say what the ramifications will be on the tobacco industry, which is a very important economic contributor, I just think that we need to look at the science of it. Let me just ask you a couple of specific questions. One, did members of the EPA Science Advisory Board assist in the development of your findings in the risk assessment report, yes or no? Mr. BAYARD. Mainly based on their comments at public reviews, with one exception that I can think of. On my first draft in 1990, before we released it for public review, we asked one fellow who subsequently became a member of the Science Advisory Board for his comments, David Burns from the University of California, San Diego, and he provided comments. David Burns had been the sen- ior editor in many of the Surgeon General's reports, including the 1986 report. So I asked for his comments. He gave me a lot of com- ments and subsequently became a member of the SAB. I actually recommended that he was suitable to review our re- port. Dr. BISHOP. Do you see a conflict of interest on having an advi- sory board of scientists reviewing a report which they contributed to the findings of? Mr. BAYARD. He was 1 of 18 members of the SAB. I rec- ommended Nathan Mantel a consultant to the Tobacco Institute, Dr. Gross-no not Gross-Joseph Fleiss, who is a well-known biostatistician. All these people had been recommended to me by the Tobacco Institute. 25 Dr. Kabot, I recommended, who had been recommended by the Tobacco Institute and Dr. Kabot and Dr. Burns were chosen. Those were the only two people that I recommended who were chosen. Mr. BISHOP. It is my understanding that there was a policy or an understanding established that if the data and the EPA's sci- entific guidelines did not show that ETS was a carcinogen, then the Science Adviso Board felt that the guidelines should be revised. Is that accurat~ Mr. FARLAND. Mr. Bishop, maybe I can address that point. The chairman of the Science Advisory Board took a question at the end of one of the public meetings about the sufficiency of the data for environmental tobacco smoke and the question was whether or not according to EPA's guidelines this could be classified as a class A carcinogen. His response was with this amount of information, if it could not be classified as a class A carcinogen, then the guidelines would have to be changed. That is, if the guidelines were getting in the way of the classification of this as a class A carcinogen be- cause they were somehow restrictive to certain types of data, and wouldn't allow you to use all the information available, then there would need to be a change. That was not a SAB finding. It was a comment of the chairman to a question of him. It specifically points to the fact that the guide- lines where not restrictive, that they are meant to be used with sci- entific ,~'udgment and to weigh all of the evidence available. Mr. BISHOP. The draft report of April of 1991, page 29, "If the guidelines for carcinogenic risk assessment can be used to cast doubt on a finding and inhalation of tobacco smoke by humans causes an increased risk of lung cancer, the situation suggests a need to revise the guidelines." Now, I really am not in a position to determine whether there is a risk of harm that is created or not. My concern is with the ef- fect of what you are doing here and the effect of this EPA study on our country that we ought to really be looking at a scientific process and that we ought to be acting on and be driven by science and not by a policy and trym g to come on the back end and justify a policy that has been predetermined. Tlus is just a lack of objectivity. It suggests a subjective approach here, to change the rules to fit what you expect-what you want the outcome to be. That concerns me because of the effect that what you are doing could have on my district. Mr. FARLAND. Mr. Bishop, I would like to try and convince you that this study was extensively reviewed by our scientific peers. We took a lot of comment and responded to those comments in terms of the development of this report. The Science Advisory Board se- lection of members was done in an open process, as is always done. I can't comment personally on the selection of those individuals since that is outside of my jurisdiction, but as Dr. Bayard men- tioned we did submit a series of names that were scientists who were well-qualified and had represented a number of sides of this particular issue. The question about the guidance that is put out there is clearly one that is an argument for the use of scientific judgment and not for allowing a strict framework in which one eliminates information so as not to consider it when you reach your conclusions. This is not looking at an issue of trying to change the

GU approach in order to fit the end. This is a strong endorsement of using good science, using good scientific judgment, and this is something that both we and our Science Advisory Board are com- mitted to. Mr. BISHOP. I appreciate the statement you are making, but it does not seem consistent with what has been brought out today and what I have seen. It seems to me that there is more of a cloud here than there is clarity. Based upon the process, it seems to me the integrity of the story of the study has certainly come into ques- tion. Based upon the objectivity of the contributors or the lack of objectivity, it seems there is more of a cloud than clarity. Mr. BAYARD. Did you know we are criticized more by the antismoking folks for the makeup of the Science Advisory Board? Did you know Dr. Lippman was attacked because he was chairman of the committee for indoor air research [CIAR.J, which is heavily funded by the Tobacco Institute. Did you know Dr. Woods was in the process of negotiating a $1.2 million grant with the Tobacco Institute at that time? In fact, 6 of the 17 members had financial ties to the Tobacco Institute. We were heavily criticized by the antitobacco folks. Mr. BisHop. It seems like what you are saying is you are agree- ing more of a fog has been created? I don't think what you need is to have- Mr. FAR[.ANn. I think what we are seeing is these are individuals with strong scientific credentials, and that is not an issue with re- gard to where their funding is coming from in terms of their ability to reach conclusions on this particular report. We were criticized from both sides because of the makeup of this report. It was a bal- anced report and very strongly scientifically staffed. Mr. BISHOP. I appreciate your comments. I am not certain that I have any more clarity now that I did before you started answer- ing. Thanks a lot. Mr. RosE. I told Mr. Lewis I was going to call on Mr. Barlow next, then our colleague here and then over here. Mr. Barlow. Mr. BARLOW. Doctors, thank you very much for being here. I have not had a chance to go through all your studies. I assume they will be in the record. I wonder if you might help me. You all obviously saw these and prayed over them for a long time. Mr. FARLAND. Once or twice. Mr. BARLOW. Give me help here, basic down-to-earth help and guidance, if you can. How many of these studies would you esti- mate deal with physical dimensions of the areas in which environ- mental tobacco smoke arises? In other words, to give you an exam- ple: Environmental tobacco smoke in a room the size of this may not be as omnipresent as the same amount of tobacco smoke in a closet might be. Can you give me an estimate on the numbers of studies that cover that? Mr. FARLAND. The 30 epidemiology studies we looked at were based on exposure of nonsmoking women, spouses of smokers in their home; so that the exposure to tobacco smoke in the home was Te ;~T8MV0Z 27 considered to be the exposure under test in these epidemiology studies. Mr. BARLOW. Now somebody smoking over in the far corner of that room, of this room here-which is a very large room, for the record-may not have an~ yimpact on me sitting up here on the op- posite side of the room. Would you agree? Mr. FAx[.ANn. The smoke will be diluted as it goes up into the room; that is correct. Mr. BARLOW. May not even reach this side of the room, depend- ing upon the drafts and so forth? Mr. FARt.ArID. It is possible; that is right. Mr. BARLOw. Are variations such as that taken into account? Mr. FAR[.Arm. Again the way this was done was to look specifi- cally at spouses of smo6rs; and in that particular case, as Dr. Bay- ard mentioned, these are very difficult epidemiology studies be- cause we all have some exposure to environmental tobacco smoke, so we were looking for groups of people, or the investigators were looking for groups of people, who had a higher than normal expo- sure to environmental tobacco smoke-still environmental levels but a higher than normal average exposure-so that they could de- termine whether an effect was produced. That is what all 30 of these studies tried to look at, smoking in the home. Mr. BARLOW. It is difficult to construct a case when you have an extreme variation in physical surroundings; you would agree, ri ght? Frequency of smoke, it is very difficult again to make a statis- tical-one statistical finding based on many variations and fre- quency of smoke. If someone were to smoke a cigar in the back of this very large room, it might not have any impact upon me; espe- cially if I am only coming in once a week, twice a week, and that person may be only smoking a cigar once or twice a week. Mr. BAYARD. Not only that, but for lung cancer you have to fig- ure out what the exposure has been over the past 30 years. Mr. BARLOW. You have a very difficult job coming up with a sta- tistical sample. Mr. BAYARD. It is hard to believe any of these studies a priori would show an effect. Mr. FAFU.A1vn. I think that is an important point with regard to the chart we showed. The first aeproach I showed analyzmg the 30 studies, was a question of "ever versus "never." ~t wasJ ust sim- ply whether or not you lived in a home with a smoker. There are vast differences in those types of studies. Mr. BARLOW. Wouldn't you say it might be unfair to tobacco farmers and the tobacco industry to make a one-phrase condemna- tion which-environmental tobacco smoke is kind of becoming a pejorative phrase, a negative phrase-when you have many vari- ations and conditions over many years? Wouldn't you say that might be unfair? Mr. FAR.ANn. Again, I think the point we are making is that from a scientific and a public health perspective, we can talk about the hazards, about the risks; but people have to make their own judgments and risk managers in local situations, in States, in res- taurants, fiave to reach conclusions with regard to how that risk is going to be managed.

r 28 This is not a general condemnation in the sense that a small amount of environmental tobacco smoke will cause all of these ef- fects we are talking about. This is a discussion about real people who have experienced effects. That has been measured in scientific studies both in lung cancer and in childhood respiratory effects. Peopie have to take that information and make informed risk management judgments, whether personal judgments or whether or not they are some sort of regulatory judgments at the State, local, or Federal level. Mr. BARLOW. You are making a judgment applying one short phrase, environmental tobacco smoke, which, as I say and many people feel, has gotten to have a negative implication connotation on the worse case situations. You are classifying the heavy smoke in a very confined space like a closet, equating that with very infre- quent smoke in a room as large as this? Mr. Fn[u.AND. In terms of a hazard call, it is equivalent to talk- ing about asbestos as a carcinogen based upon occupational expo- sures. It is based upon talking about other types of occupational ex- posures to chromium, or other types of inetals or benzidine dyes or any of those in terms of a hazard. There is a hazard to exposure to environmental tobacco smoke. Now, the risks depend upon the exposure. If you can avoid envi- ronmental exposure, or if you can lessen your exposure to environ- mental tobacco smoke, you are going to lessen your risk. Mr. BARLOW. Some of those hazards may be very minimal, right? Mr. FARLANn. Under certain conditions, absolutely. Mr. BARLOW. I can recall in my lifetime my father would smoke an occasional pipe or cigar. I enjoyed as a young child smelling the smoke as it wafted across the room; and any hazard I mi ght have incurred in that process might have more than been made up for the love of being with my father and enjoying his enjoyment of his pastime, true? Mr. FARLANn. Mr. Barlow, that is a personal choice. I think it is absolutely true. I think you ought to be informed about the po- tential hazards and then make that judgment. Mr. BARLOW. Thank you, Mr. Chairman. Mr. ROSE. Thank you. But the data would suggest that what you just described would be good for you. Mr. BARLOW. Yes. Mr. ROSE. Mr. Lewis. Mr. LEWIS. Mr. Kingston of Georgia. Mr. KINGSTON. Thank you, Mr. Chairman and Mr. Lewis. I want- ed to ask you about the integrity of the study, which I know you are getting pounded on a lot. I hate to be redundant; but my father was a college professor. He told me years ago when, unfortunately, Federal dollars got involved on university campuses, all the studies suddenly became politically correct, with the conclusion meaning that more money had to be spent on that particular item that they were studying. In other words, it was predestined, whatever their conclusion was going to be. Knowing the pressures you folks are under, do you feel in your heart of hearts that this study was objective and just-I am not trying to belabor. Just a simple yes. Z~l 8MV0Z I 29 Then, since you are shaking your heads yes, all three of you agree with that-another thing that we studied-I studied chem- istry in college. Whenever you do a study, you have to look at the prejudice of the scientist. With that in mind, do you folks smoke? Do any of the people in the study smoke? Or was it-maybe a little bit, as Mr. Bishop suggested, maybe there was some antismoking sentiment in the scientists, that was there. Is that the case? Do you think-did you have a mixture? That is very important to know what the prejudice of the scientist is. Mr. BAYARD. I think probably it is better I answer that. If you want Dr. Farland to- Mr. KINGSTON. No. I know you are the point man. Mr. BAYARD. I am the project manager for this. I had something to do with choosing the contracting support. In terms of money, my paycheck just doesn't change. It hasn't changed for 14 years other than cost-of-living allowances. I have never been offered money from the antitobacco folks or the tobacco folks. I don't know what that means. In terms of smoking, I smoke an occasional cigar or occasionally smoke a cigar-good ones, actually. In terms of the objective, a couple of my-I have five children. Four are adult children. Two smoke, two don't. My exwife smokes, but I don't blame smoking for her. In terms of objectives, this is what I do for a living. In terms of getting people, now, what I tried to do was get people-the most unbiased people, people who knew nothing about tobacco smoke be- cause it had been- Mr. KINGSTON. The reason I am concerned about that, as a new Member of Congress, I find almost whatever you are trying to find out about, if it comes from one party, it goes this way; if it comes from another party, it goes that way. It is not tobacco studies but budget, anything you ask. In fact, I don't know if you have seen the arGicle-I will be happy to share it with you-a book called Galileo's Revision: Junk Science in the Courtroom. In it, Peter Huber of the Manhattan Institute for Policy Research talked about how lawyers get scientists in a court- room, that come up with conclusions that basically back up their argument. I just see a real corruption in the pure aA of science these days. It really depends on who is writing the check for the grant, the study. I say that because the National Cancer Institute had done a very comprehensive study which they released in November that said there was no significant link; and, I don't know if this is-if we are getting politics in the lab. That worries me. Mr. BAYARD. I can tell you I am not antismoker. I don't get paid any more for what I say. No matter what I say today, I will get paid the same tomorrow. I get paid every other Tuesday. You are going to get some other contractors coming up here. You ask them if they will get paid tomorrow depending upon what they say.

30 I stand by my own personal objectivity. I am sure if I had had any links, they would have been discovered by now; you would have known about them. Mr. FARLAND. To follow up on that, I think it is important we understand the process used for this report. We had numerous sci- entists involved in reviewing the drafts, as it was developed. Cer- tainly, within our own Office of Research and Development, there was a lot of internal peer review, scientific peers who had no par- ticular biases or no particular connection with this report as they made their views known. The external scientists that we go to routinely-we go to a wide range of scientists. As Dr. Bayard mentioned earlier, there were some 18 scientists on the Science Advisory Board panel. I men- tioned in my opening remarks, in my testimony, that these were follow-up reports to the U.S. Surgeon General's report in 1986, the National Research Council, Academy of Sciences in 1986; the World Health Organization in 1986 and 1992; National Institute of Envi- ronmental Health Sciences in 1991; and the National Cancer Insti- tute in 1993. To suggest that all of those individuals would somehow be biased in a way that would argue wrongly that there is a public health impact of environmental tobacco smoke, I just don't believe it is possible. Mr. KINGSTON. I am glad to hear that. Let me ask one other question that gets to Congressman Barlow's question. In terms of the size of a room or whatever, when somebody is smoking and the smoke becomes diffused, I suppose the smoke chemically reacts with some other chemical in the air, that the smoke intermingles with oxygen or whatever is out there. In your study, did you study the effect of a household in the inner city, for example, or maybe one that never uses air-condi- tioning, compare those two, versus one that is outside with a good breeze? Did you weigh out differences like that? Mr. FARLAND. Again, these studies are from eight countries around the world. Some of them were actually looking at the use of smoky coal from a heating source. In fact, there appeared to be lung cancer associated with the use of that smoky coal as an indoor air pollutant. Mr. KINGSTON. Would that skew the study? Mr. FARLAND. In that particular case, it quite likely wiped out any small effect associated with environmental tobacco smoke. Mr. KINGSTON. You would eliminate that household from the study? Mr. FARLANn. In that particular study, the study was actually of a group of individuals who lived under those kinds of conditions. It looked at a number of different ways of evaluating lung cancer; and in that particular study, there was no indication of environ- mental tobacco smoke or tobacco use causing an increased risk. There was an association with the smoky coal. Mr. KINGSTON. Would you do the same or did those studies do the same thing if there was radon in the air or somebody living un- derneath an el train and somebody-in an inner city versus some- body living in suburbia. 31 Mr. FARLAND. This is why it is difficult using one study to reach conclusions. This is why we had to use all the information available to us. I think what it shows is some studies were much more pow- erful than others. That some showed no increase may very well have been because both the exposed and the control individuals were showing some lung cancer risk from some other source; and because it pushed both of them up, it washed out the effect of the smoking. We just do not know. That is the nature of these studies. Again, this is a fairly unique data base because these are results that are being looked at at environmental levels as opposed to very high levels in the workplace, or what have you. Mr. KINGSTON. In the 90 versus 95 percent, would that tie into that at all? Mr. FARLAND. Again, I really believe that the issue of 90 versus 95 percent is a red herring. It is a statistical test of significance that is a 95 percent test. It has a confidence interval of 90 percent; and that 90 percent, as we have agreed to here, is less than 95 per- cent. But if the results that one finds are significant at 90 percent, and not significant at 95 percent, in light of all of the biological infor- mation, in light of all of the other studies that have been done on animals and other groups like that, you are not going to be con- vinced by that simple statistic. You. are going to make a decision on the basis of biological plausibility, on the basis of consistency of results, of findings from other studies and so on. That is very important. Mr. KINGSTON. Thank you, Mr. Chairman. Mr. ROSE. Thank you very much. Any further questions? Mr. Inslee. Mr. INSLEE. Thank you, Mr. Chairman. As far as ETS, I am trying to get a grasp of what you are saying. It has been very educational. But is this a situation where you as- sume that ETS has carcinogenic properties, so the real question be- comes, is there enough exposure in real life to produce carcinogenic effects? Do you see what I am driving at? In other words, is that the real heart of this study? Mr. FARLAND. Mr. Inslee, you asked the right question: First of all, is there a hazard, a carcinogenic hazard, associated with ETS? The answer to that was brought up in the first discussion. We categorize it as a category A known human carcinogen. We are quite confident. The question of risk has to do with how much exposure we get. Now, it is-for chemical carcinogens where there is a substantial data base on mechanisms of action, we can generally assume that there is a linear response from high doses to low doses, so that the probability of a cancer response will decrease as the exposure de- creases. At some point, we are going to reach a level where that prob- ability of cancer risk is no longer significant. It may still be there. There may still be a very small cancer risk associated with it; but given our lifestyle, everything else we do, it will not be significant. ~I~IRMVn71

Ir 32 That is what we really have to take into account as we make risk management decisions. Are we dealing with a situation that is high enough to be of concern to us, or are we dealing with occasional ex- posures that are not necessarily going to be a risk? What we are telling you in this particular case is that in individ- uals who lived with smokers we were able to measure-our inves- tigators were able to measure an increased cancer risk using a very crude tool which is an epidemiology study. So it is quite likely that the significance of that cancer risk is sufficient to be a public health concern. Mr. INSLEE. The report, at least my brief scan of it, indicated there were not workplace studies, that we do not have good epide- miological studies on workplace environments. Is that accurate? Mr. FARLANn. We did not focus on the workplace environments. I will let Dr. Bayard explain why. We actually talked about that in our report, so it was clear why we used these other studies. Mr. INSLEE. I think I can assume the reason why. You had lesser levels of exposure, not as much stability in the exposure, those rea- sons. But does that indicate that in our public policy we should not use this study for workplace policym aking questions? Mr. BAYARD. That would be for my response. The fact that a workplace is a much tougher place to look at exposure of environ- mental tobacco smoke doesn't mean that-I am sorry. To me, if en- vironmental tobacco smoke is carcinogenic in the home, because you are exposed in the home and because you can measure it better in the home means you should be careful in the workplace, too. In fact, in the numbers in our report we look at both home and nonhome exposures and calculate a total number of potential can- cer cases from home and nonhome exposures. Mr. INSLEE. That is included in the number, then? Mr. FARLmD. That is right. Mr. INSLEE. As far as peer review, has this study undergone or should it undergo a peer review process that, customarily, the study would be exposed to? Mr. FAFUANn. This study has undergone as much or more than studies that we have been doing over the past 8 to 10 years. Mr. INSLEE. As far as this was just an academic study that, as I understand the process, will commonly go through peer review process, has this gone through a similar process? In other words, do you have a study coming out in a journal? Mr. BAYARD. We do a few articles for journals. There are usually two or three reviewers. This one had 18, plus all the public review. Mr. FARLAND. This one was extensively peer reviewed. Mr. INSLEE. Thank you. Mr. ROSE. We have been 2 hours on this. We have had our first witness. So we will be here until midnight. Dr. Farland, in March of 1990, you sent an internal review draft of the ETS risk assessment to various groups within EPA, includ- ing the Environmental Criteria and Assessment Office in Cin- cinnati. What is the role of this office and what type of expertise does it have? May I suggest an answer for you, and you tell me if I am riRht or wrong. If you want to add to it, tell me, and you can. VMMM 33 Is it not true the Cincinnati ECAO group is a team of scientists with complementary expertise that involves itself in health risk as- sessments? The group includes experts in epidemiology and toxi- cology ? Mr. FARLnND. Yes, Mr. Chairman. Mr. ROSE. Would you add to that? Mr. FARLAND. I would say, yes. That is part of my group. That is my Cincinnati office. Mr. ROSE. I have seen from the documents that you provided to us that in a letter to you from Dr. Sonic from-which one is that; all right-dated April 27, 1990, signed by Acting Director, Dr. DeRosa, the Cincinnati group questioned the use of meta-analysis to support the classification as a group A carcinogen and suggested that the epidemiologic studies more appropriately reflect limited evidence of human carcinogenity. They also stated that there are substantial differences between mainstream smoke and sidestream smoke, or ETS, and con- sequently it is difficult to generalize about the properties of one study from the other. Thegr oup also commented that, finally, there are tremendous scientific regulatory and political ramifications of categorizing a substance as a group A carcinogen. With all due respect to the epi- demiologists who produced the report, given the inherent limita- tions of the data and the comparative novelty of the approach used to interpret the data, I would recommend that this approach not be used as the basis of a group A classification. Now, none of the documents that you provided to us contained-~ ~ any response to these comments either by you or by Dr. Bayard. What response did you give le Cincinnati group? Here is the document, technical manuscript review form, recommendations, ac- ~eeptable after major revision. OK? Mr. FARLeNn.es, Mr. Chairman. That was the first draft. As I said, it went through extensive internal review. This is our group. My response to that was to have a conference call to follow up with the scientists who were involved in the Cincinnati review and with our own scientists here to discuss the data, to discuss the is- sues they had laid out in that memo; and we took to heart their perspectives on that. They are certainly not the only oneQ that have provided that type of a perspective. - -: , :-.. ^ . 1. . . And we indicated that we would further strengthen the docu- ment as we went through the draft, taking other comments into ac- count; and we would supply it to the Science Advisory Board to help us finalize it. So, again, I think it points out that there is a diversity of views on these particular issues, those within our own group. But we did very strongly address many of the issues that were raised simply because that is the purpose of peer review, trying to understand where the misconceptions or the changes might be. Mr. RosE. All right. On March 9, 1992, you sent a memorandum to several EPA grousm asking for a second internal review of a revised draft, I ~,r.a that would be the second draft of the risk assessment. How I.~n~ wr. the dr.ft document at this time and can you tell us how m.nr Mre ke were allowed for this review of your second draft?

r 34 My indication is that the document was between 300 to 600 pages long, and you gave them about 13 to 14 days to review it. Mr. FARLAND. That is not unusual, Mr. Chairman, when we are dealing with the second draft of a document. About 2 weeks is what I would have said, although I do not really know that for sure. Mr. RosE. Let me move on to my real question here. I gather the Environmental Criteria Assessment Group in Cin- cinnati was critical of the time allowed for the review. It so indi- cates? Mr. FARLAND. That is also not unusual. Mr. ROSE. I understand. Dr. Harvey, Director of the ECAO in Cincinnati at that time, wrote on March 24, 1992 in this document that "No one liked the 11-day time allotted for review." He contin- ued "I suggest that the document manager consider more time for ev2uation to balance the seriousness of this document as applied to the public health and the intrinsic value of doing it right on this key health topic." What was your response to this criticism? Do you think it appro- priate to allow 11 days for the thorough review of a 600-page-you already said that is typical. What was your response to Dr. Harvey's- Mr. FARr.AND. Again, Dr. Harvey is my director out there at ECAO; we did talk about that. This is certainly not the only docu- ment we worked on. We work on lots of different things. We put time pressures on our scientific staff to review these. This document was 4 years in the making. It went through a tre- mendous amount of peer review. While we may not have gotten the extensive commentary from Dr. Harvey's group, we feel it was ade- quately peer revie edw Mr. ROSE. Now, you told us affter this April 27, 1990, first draft, you had a conference telephone call, and you took things that they said into consideration; so here comes the-their review of the sec- ond draft which, as you observed, they had 11 days to look at and gave it-sent it out on March 24, 1992, almost 2 years later. The ECAO group.in Cincinnati seemed hi ghly critical of this re- draft of the risk assessment, as it existed in March 1992. While Dr. Murphy appealed for more time, she did in her limited review find plenty to criticize about the risk assessment. She wrote, "Was there any attempt made to include nonpublished studies which are likely to have nonpositive findings in the review?" She concluded later, "I, feel that the importance of the trend test and its associated prob- ability value is overstated. Misclassification and measurement error can mask a dose response trend, but can also sometimes cre- ate one." Moreover, she wrote, concerning the lack of consistency of the histologic-type of lung cancer that, "I feel that it distracts from the presumed causal relationship of lung cancer and ETS." Here again, we have no copies of your response to that. What was your response to these concerns? Mr. FARLANn.Ag am, these comments, along with the other com- ments from the other groups, were passed on to Dr. Bayard as the manager. He had the opportunity to address those comments. I can let Dr. Bayard speak to that directly. a~~~~~9~(17 35 Mr. ROSE. Make a note, Dr. Bayard. We will come back to you. Let me just get one more thing out here. Then I am about through. Dr. Harvey also wrote in his March 24 letter to Linda Bailey- Becht, "I suggest a full discussion of carcinogen category A versus B based on the absence of definitive data of passive ETS in hu- mans. Like it or not, EPA should live within its own categorization framework or clearly explain why we chose not to do so." Dr. Harvey also wrote that "ECAO-CIN will be most happy to spend further time improving the quality of this document." What was your response to these criticisms that the Agency's guidelines were not being followed? Did you have discussions on these matters with either Dr. Harvey or his staff? How was this matter resolved? Mr. FAxl,Alvn. I did have discussions with Dr. Harvey on that particular issue because Dr. Harvey had been with us for a little over a year at that time. He had not been involved in the develop- ment of the guidelines, and the discussion about how EPA's classi- fication system was dealing with various types of chemical cases was not particularly well known to him. We talked about the category A classification. We talked about efforts that were going on to revise our guidelines and to use a nar- rative-type of approach rather than an alphanumeric classification, a box. Mr. ROSE. Could you provide for the record any correspondence or notes you had on this subject, please? Mr. FARLAND. Mr. Chairman, I will look. I don't believe I have any notes on that. I think they would have been sent up before if they were available. Mr. ROSE. Take another look. We haven't had a chance to ask you all the questions that we have, but we would like to supply them for the record. [The information follows:] A copy of EPA's Carcinogen Riek Assessment Guidelines and a list of documents on compounds that have been evaluated for carcinogenicity were provided to you in your onginal reque®t. We ~~ill be happy to provide further discussion of such docu- ments if you with. Mr. ROSE. If you will shepherd that effort to get us that informa- tion, we would appreciate it. Mr. FA13.LANn. We will be pleased to deal with those questions. Mr. ROSE. Dr. Bayard, did you have anything to do with the se- lection of the companies that were contracted with by EPA to do your study, your evaluation? In other words, you were the Project Manager. Actually, how many contractors out there were working under you? Mr. BAYARD. Let me try to get to that the best I can. Mr. ROSE. That is what I want you to do, but quickly. Five, six, 10? 1±4r. BAYARD. ICF was a contractor; Battelle was a contractor. Those two, but most of those had subcontractors. Mr. ROSE. All right. And who made the choice about those two? Mr. BAYARD. I did not choose- Mr. RosE. Beg your pardon? You made the decision?

36 Mr. BAYARD. To choose ICF? I guess I did, but it was chosen mainly as an umbrella contractor. Mr. ROSE. Yes. Mr. BAYARD. I am not the Project Officer on any of those. Mr. ROSE. You were the Project Manager. Mr. BAYARD. I am the Project Manager, but not the Project Offi- cer on contracts. Mr. ROSE. Did they work for you? Mr. BAYARD. I am a Work Assignment Manager on contracts. I submit the work assignment I want done. Mr. ROSE. To ICF, is it ICI? Mr. BAYARD. ICF was one. Battelle was another. Mr. ROSE. They subcontracted the contract to the Institute for Smoking Policy? Mr. BAYARD. No. I had nothing to do with that. Mr. ROSE. That was my question. I didn't ask, did you have any- thing to do with it. I said the contractor that you picked. ICF, they assigned it? Mr. BAYARD. The Institute for Smoking Policy had nothing to do with the risk assessment. They had to do with the workplace policy guide. I am here to discuss the risk assessment. I also used con- tractors from ICF. If you want, I will be glad to talk about that. Mr. ROSE. No. You had experience with these people before, had you not? Mr. BAYARD. ICF? Mr. ROSE. Yes. Mr. BAYARD. A very good atatistician worked at ICF, we used for the first draft. Mr. ROSE. We have to go vote. We will be back within a few min- utes. I think Mr. Baesler has a few more questions. We will try to be back within 10 minutes. [Recess taken.] Mr. ROSE. The subcommittee will come to order. I will restate for the record that we have a lot of questions that have not been an- swered, that we would appreciate the Agency responding to us in the usual timely fashion. Mr. Baesler. Mr. BAESLER. We have talked about several studies during the day. I would like to talk about one particularly. I notice that the Fontham study-we note in this regard that the revised drafts of the risk assessment dates from fall of 1992, not long after the Fontham study appeared, take the study results fully into the ac- count. You did take the Fontham study results fully into account; is that correct? Mr. BAYARD. Yes. Mr. BAESLER. We see amonp the documents provided a memo stating that even more emphasis needed to be put on the Fontham because it was NCI funded and it was the largest study. Brownson was also NCI funded and even larger than Fontham. Finally and perhaps most disturbing we have seen a document in which Kenneth Brown promises Dr. 'Fontham to hold up the cir- culation of the revised risk assessment draft until her study ap- 37 pears in the scientific literature. On October 12, 1991, we see that Kenneth Brown wrote to Dr. Fontham and said of the inclusion, "The time element is of concern to EPA, but I will not violate your request for propriety. Without permission from you I will hold a re- vised version that includes your study until it appears in print and is thus publicly available." It is my concern and the committee's concern that you seem to spare no effort to take into account the studies that help your case, but you didn't give the same consideration to studies that didn't help your case. How do you respond to that? Mr. FARLANn. The report published in December of 1992 listed a number of studies that came out shortly after the Science Advi- sory Board had looked at our report and had essentially given us a final sign-off. They included Stockwell and Brownson. As you are probably aware, scientists who are involved in this particular field get preprints of the materials before they are fi- nally published. We had preprints of the Fontham study, preprints of the Brownson study as they were being developed, but there is a reason that one cannot cite certain studies until they have ap- peared in the published scientific literature. I will let Dr. Bayard speak specifically to the issue of how we dealt with the Fontham study. We have taken into account the two other studies that were men- tioned, particularly Stockwell and Brownson. Both are consistent with our results. They are not at odds with our results, and I think you will see that in our report there is a statement to that effect. So we have not ignored those studies at all. Mr. BAESLER. Another question on the Fontham study. The risk assessment classifies the Fontham study as tier 1, these being studies that are of greatest utility for investigating a potential as- sociation between ETS and lung cancer. You say use of dietary, oc- cupational, and other exposure data in that analysis, along with an additional 2 years of subject accrual, will make the completed study for this, which constitutes an interim report, even more valu- able. From this I understand that you realize that the Fontham study published in 1991 did not attempt to analyze the effect of confound- ing factors. Since your guidelines for carcinogen risk assessments require you to rule out the possibility that confounding factors might account for an observed increase in risk, I assume that you asked Fontham to undertake analysis of the effects of compounding factors such as diet and previous lung disease, yet I see no commu- nication between EPA and Dr. Fontham in these documents that you supplied to me that discusses this matter. Did you commu- nicate with Dr. Fontham on this matter? Mr. BAYARD. On the matter of confounding? Mr. BAESLER. Yes, sir. Mr. BAYARD. I communicated with Dr. Fontham a lot-not that much-and we talked about confounding. The way the study was designed made it really helpful because there were very few con- founders left. Mr. BAESLER. I will read from the study. Mr. ROSE. Do you have an answer, sir? 9~TSMfi0Z

38 Mr. BAYARD. Yes. We talked to her about workplace, about child- hood exposures, about the exposure measurements, about whether or not we should pick population controls or the colon cancer con- trols. We talked about-there was one more-I am sorry I forget. Mr. BAESLER. Fontham did appear before the SAB in July 1992 where she stated, the approximate 30 percent risk of lung cancer associated with s ousal ETS exposure persisted after adjustments for vegetable consumption which was the most significant food or nutrient factor, family history of lung cancer and employment in hi gh risk occupations or industries. Whether or not this is correct, these findings have not been pub- lished in the peer review literature; have they? Mr. BAYARD. I think that is correct. Mr. BAESLER. Moreover, there is no evidence that EPA even thought this analysis to be important, though clearly some of us would think it was. Also Dr. Wu Williams, a coauthor of the Fontham study, cur- rently has a grant with NCI to undertake the analysis that Fontham said were complete to the SAB; is that correct? Mr. BAYARD. I don't know. Mr. BAESLER. Dr. Wu Williams is a coauthor of the Fontham study. He applied to the National Institutes of Health in October 1991 for a grant to analyze the Fontham study data. D,r. Wu Wil- liams' project description includes a proposal to evaluate "the inde- pendent effect associated with each of the above factors, indoor air pollution, diet and instances of previous lung disease and its poten- tial confounding effect on the passive smoking lung cancer associa- tion." A summary statement of a special review committee that re- viewed the grant proposal of the National Institutes of Health stat- ed that, "given the relatively small risk found for passive smoking, it is critical that ETS possible confounding factors be investigated to determine whether the passive smoke effect is merely the result of confounding or of other variables." This grant was awarded with an initial project period from June 1992 to May 1993. Were you aware of this? Mr. BAYARD. No. Mr. BAESLER. Do you agree with the comments of the NIH Spe- cial Review Committee that it is critical that other possible con- founding factors be investigated to determine whether the passive smoke effect is merely the result of compounding smoke or other variables? Do you agree with that? Mr. BAYARD. In the Fontham study or all studies? Mr. BAESLER. How about the Fontham study and all-include ev- erything. About the Fontham study first. Mr. BAYARD. My answer is no. Mr. BAESLER. On both cases, right? Mr. BAYARD. No. The question was is it critical with respect to our finding whether or not the Fontham study might have had more confounders and the answer is no because the Fontham study would have been one of 33 studies on epidemiology Mr. BAESLER. What steps did you take to insure that the relettvr risks reported in the Fontham study were not merely the reMult 4 V.1~ I 8~V ~V 09 39 confounding of other variables? Did you take any steps in your study on that? Mr. BAYARD. We examined every one of the studies where con- founders were addressed. We couldn't identify any one confounder that could have possibly been responsible for dose response rela- tionships that we saw or the fact that we saw these increases in all different countries. For instance, diet changes in every country, and yet we saw the effect of passive smoking in all the countries we looked at, with the exception of China, where the studies were mainly to examine the effects of smoke and other indoor air pollutants. Mr. BAESLER. It seems to me that there are many doubts sur- rounding the Fontham study in terms of there being an incom lp ete report that did analyze for potential confounding factors. I tlunk' everybody agrees that there wasn't analyzing for confounding fac- tors. That the study should not have been mcluded in the meta- analysis of the U.S. studies. I gather that a meta-analysis of the U.S. studies, including the two new studies, but excluding Fontham, gives us a statistical nonsignificant summary risk of 1.04. This suggests to me a considerable instability in the data. How can you claim such confidence in your analysis when slight changes of studies considered have such effects? In other words, if we take out the Fontham study, it seems it changes the whole thing. Mr. BAYARD. Why would you take out the best study we have available? Mr. BAESLER. Maybe because they didn't take into account the confounding factors. Mr. BAYARD. It was a published study in a well-respected journal supported by the National Cancer Institute the same way the Brownson study and Stockwell study were. Why should we take that out? It seems to me you are just being critical of the studies that were positive. Mr. BAESLER. Do you think it is relevant that now we have a grant Oven to study this study and how confounding factors might affect it. The National Cancer Institute has given a grant to study the effect of confounding factors as pointed out by the Fontham study, which you indicate is the best study you had to make your ar ment and the results of the ETS problem. ur. BAYARD. It was the only tier 1 U.S. study, if I am not mis- taken. There were several tier 1 studies in other countries. Mr. BAESLER. Maybe I just don't know enough, but you admit that we didn't do an analysis on what the other confounding effects might have had on Fontham's results? Mr. BAYARD. If I had had the information on diet, if I had known it existed, I would have done the analysis on confounding. Mr. BAESLER. You wouldn't do it, but the National Cancer Insti- tute thinks it is relevant enough that they will give, a grant to a Qroup of people to do a study that you said wasn't relevant. ktr F kHt.A~.n Mr Baesler, I think it is very important that you un=1, r%tand that we would have follow-up studies on the majority „/ t he r{~iArmi~~lolry studies that are published in the literature t,-1~~ TIwow studies find information, they develop additional

40 hypotheses and they suggest additional studies that need to be done. That is the way that the epidemiology work is done. I think this was a well-designed study that took into account a number of these issues with regard to confoundinP that many oth- ers did not. While there was not a specific analysis of confounding published with that study, I wouldn't argue that the answer was in and that we shouldn't do any additional research. The authors of the study came into the National Cancer Institute to get an additional study. They may find when they collect more information that the relative risks may go up, they may go down, or are not clear. Mr. ROSE. One question. Did you do a one-tailed analysis of the Fontham study? Mr. BAYARD. Yes. Mr. RosE. And the issues around confounded hadn't been re- solved; is that correct? Mr. BAYARD. That is correct. Mr. ROSE. And if you had done a two-tailed study on Fontham, your-what is the- Mr. BAYARD. Significance level Mr. RosE. Your confidence level would have been 95 percent; right? Mr. BAYARD. That is correct. It is close enough. Mr. ROSE. But you did a one-tailed study and essentially your confidence level is 90 with the one-tailed study? Mr. BAYARD. Yes. Mr. ROSE. That is all I was trying to get at earlier, Dr. Farland. Why did you have so much trouble answering my question? Mr. FAItLArID. This is a very difficult statistical issue. Mr. ROSE. I know. Now, if the statistical confidence level of a one-tailed study produces a 1.04, could that number have gone down with a two-tailed study? ' Mr. BAYARD. The two-tailed test on the ever versus never, I think, was 0.99, with the 95 percent confidence interval-the lower 90 rcent interval was 1.04. Jr. ROSE. If you had done the two-tailed study, you could have had a confidence level of 95, but your statistical significance could have been below one?' Mr. BAYARD. I know what you mean. Mr. ROSE. Do you see why we are a little suspicious about this game? Mr. BAYARD. That is true for the ever versus never. Mr. ROSE. That is the only American study with any statistical sigm ficance and you do a one-tailed study with a 90 percent con- fidence level. If you had done a two-tailed study, it probably would have been below one. You haven't figured in the confounding fac- tors that we just talked about and yet you issued this report in a hurry between changes in administrations and drive policy in this country from every courthouse, every workplace, every building in America. I have nothing to argue about the fact that direct smoking can cause serious results in human beings, but when you are in such a hurry to reach a result like this on this kind of weakness, I think we have a right to ask some tough questions. 41 Dr. Farland, why didn't you go for a separate study, an American Congress funded, whatever, an EPA request for one massive study of this issue in America, do it right, do it long enough, get this issue cleared up once and for all? Why didn't you do that, sir? Mr. FARLAND. Mr. Chairman, again, these issues have been under discussion in the scientific community for many years. This was not done in a hurry. The evaluation of these 30 studies is a umque data base. We have very few situations where we have this many studies to work with. I would not argue at all with the need to get addi- tional data on this issue. As a scientist, I agree with you. Mr. ROSE. How many other countries in the world consider envi- ronmental tobacco smoke under their system equivalent to a class A carcinogen? How many other countries in the world? Mr. FARI,AND. I think it is probably most important that the World Health Organization has taken the position that environ- mental tobacco smoke causes lung cancer. Mr. ROSE. When did they take that position? Mr. FAxt.AND. They took that in 1986, and on the issue of other respiratory disorders in 1992. Mr. ROSE. How was that conclusion reached? Mr. FARLAND. The same way that the Surgeon General's conclu- sion and the National Academy of Sciences conclusion was reached. Mr. ROSE. Was that on environmental tobacco smoke? Mr. FAIU.A1vD. Yes, it was, with 14 epidemiology studies at that time, not 30, and not nearly of the power of Fontham and Brownson. This is an issue that has been dealt with in the sci- entific community for a long time. I am certainly willing to answer technical questions about the way that we went through this, but not that we hurried. Mr. RosE. I would appreciate it if you would wow me with some data, if you can provide for the record whatever you have on the way in which the World Health Organization reached its conclusion and why you think it was based on sound science. I am sure that is around EPA somewhere since this has been studied so long and so carefully. Mr. FAItLAND. We will provide you with a copy of the report. [The information follows:] Copies of the WHO documents are enclosed for your use. All of the WHO docu- ments receive .rorldwidepeer review before publication. The International Agency for Research on Cancer (IARC) document on Tobacco Smoking ham been used in nu- meroue countries in establishing policies on smoking.