Document 0Jmm792qQokLLKoZ3qVp3Vvdm

1 IN THE SUPERIOR COURT OF THE STATE OF DELAWARE IN AND FOR NEW CASTLE COUNTY IN RE: ASBESTOS LITIGATION ) C.A. No. 77C-ASB-2 BEFORE: HONORABLE JOSEPH R. SLIGHTS, III, J. APPEARANCES: JOHN J. SPILLANE, ESQ. BARON & BUDD, P.C. for the Plaintiff CAMERON R. WADDELL, ESQ. LeBLANC & WADDELL, LLP for the Plaintiff JOSEPH BIDEN, III, ESQ. BIFFERATO, GENTILOTTI, BIDEN & BALICK for the Plaintiff THOMAS C. CRUMPLAR, ESQ. JACOBS & CRUMPLAR for the Plaintiff DAUBERT HEARING TRANSCRIPT THURSDAY, OCTOBER 20, 2005 - A.M. SESSION SUPERIOR COURT OFFICIAL REPORTERS 500 N. King Street, Suite 2609, 2nd Floor Wilmington, Delaware 19801-3725 2 APPEARANCES CONTINUED: SAMUEL L. TARRY, JR., ESQ. McGUIREWOODS LLP for the Defendant DaimlerChrysler JAMES M. KRON, ESQ. SOMERS S. PRICE, JR., ESQ. POTTER, ANDERSON & CORROON LLP for the Defendant DaimlerChrysler STEVEN T. JOHNSON , ESQ. PEPPLE, JOHNSON, CANTU & for the Defendant OI, SCHMIDT, Inc. PLLC TED GIANARIS, ESQ SIMMONS COOPER LLC for Defendant ALSO PRESENT: David deBruim Erin Farris William A. Kohlburg Kathleen D. Hadley Michael Angelides Melissa Crowe Perry J. Browder Amy Garrett Christopher Lyon Bernard Kuery Francis J. Gribbin Margaret England Joseph A. Gabay Christian J. Singewald C. Scott Reese Miranda D. Clifton Gary Kaplan Mark Reardon Kai Seelaus J. Michael Johnson Neal Glenn Christine Boyd Eric Henry PRESENT: As noted. OCTOBER 20, 2005 Courtroom No. 8B 9:45 a.m. 3 1 THE COURT: Good morning. 2 MR. TARRY: May it please the Court. 3 THE COURT: Yes. 4 MR. TARRY: Your Honor, Daimler-Chrysler would 5 call Dr Michael Goodman. 6 MICHAEL GOODMAN, M.D. 7 having first been duly sworn, was examined and testified 8 as follows: 9 DIRECT EXAMINATION 10 BY MR. TARRY: 11 Q. Good morning, Dr. Goodman. 12 A. Good morning. 13 Q. You' re an epidemiologist, sir? 14 A. Yes. 15 Q. And where are you currently employed? 16 A. I am on faculty of the Department of 17 Epidemiology, Emory University School of Public Health. 18 Q. And have you reached some conclusions on the 19 risk of asbestos-related disease in auto mechanics? 20 A. Yes, I have. 21 Q. And are you prepared to share those conclusions 22 with the Court today? 23 A. Yes. 4 1 Q. Can I get you to agree in advance that any 2 conclusions you express will be done so to a reasonable 3 degree of medical and scientific probability? 4 A. Yes. 5 Q. Do you have an opinion about whether or not 6 auto mechanics, people who work on cars, are at an 7 increased risk for the disease of mesothelioma? 8 A. Yes, I do have an opinion. 9 Q. Was that your opinion on that, your conclusion? 10 A. My conclusion is that persons involved in 11 automobile repair are not at increased risk of 12 mesothelioma. 13 Q. And have you reached a conclusion as to whether 14 those same individuals are at an increased risk for lung 15 cancer? 16 A. Yes, I have reached a conclusi on. 17 Q. What' s your conclusion on that question? 18 A. My conclusion is that persons involved in 19 automobile repair are not at increased risk of lung 20 cancer -21 Q. Finally - 22 A. -- due to their occupation. 23 Q. And have you reached conclusion as to whether 5 1 those individuals are at an increased risk for the 2 disease asbestosis? 3 A. Yes. 4 Q. And what is your conclusion on that question? 5 A. And my conclusion is that persons involved in 6 motor-vehicle repair are not at increased risk of 7 asbestosis. 8 Q. I'd like to talk very briefly about your 9 background and training for the Court. 10 A. Sure. 11 Q. You have a little bit of an accent. Where are 12 you from originally? 13 A. I was born in Lithuania. 14 Q. And did you do your medical training there? 15 A. Yes. I went to medical school there. 16 Q. Tell us about your medical training. 17 A. I went to medical school between '79 and '84. 18 And it's a six-year program in many countries. And then 19 I did internship in pediatrics and, then, I practiced 20 for three years in a small rural hospital in Lithuania. 21 And, then, I moved to the capital city, where I did the 22 fellowship in what's called academic pediatrics. And 23 then I left the country. 6 1 Q. You were a pediatrician over there? 2 A. Correct. 3 Q. When did you come to the United States? 4 A. 1990. 5 Q. You applied for an exit visa? 6 A. No, I was -- I was awarded a political refugee 7 8 Q. All right. And tell us about your medical 9 training, scientific training since you've been in the 10 United States. 11 A. When I entered the country, first thing, if you 12 want to be a physician in the United States, you have to 13 pass a number of exams, which I did. And, then, you 14 have to reach -- get another training -- if you pass the 15 exams and you qualify, then you can undergo additional 16 training in a medical discipline of your choice. In my 17 case, I was already pediatrician back in Lithuania, so I 18 did another residency, three years of residency, in 19 pediatrics and became board eligible in pediatrics. 20 And, then, I decided to move on and -21 Q. Let me stop you. 22 You're board certified in pediatrics? 23 A. Yes, I am board certified in pediatrics. 7 1 And then I decided to move on and doing the 2 program in preventive medicine. It was a second 3 residency, basically back to back, in Baltimore at Johns 4 Hopkins, and I did two years. 5 Q. Is that the School of Public Health at Johns 6 Hopkins University? 7 A. John Hopkins University. Now it's called 8 Bloomberg School of Public Health. And I spent two 9 years there as a resident in preventive medicine. 10 Q. All right. You're what is called a medical 11 epidemiologist now, or - 12 A. Well -13 Q. Physician epidemiologist? 14 A. Yes, you can call me that. That's probably the 15 most accurate way to characterize what I do. 16 Q. Can you tell us about your training at Johns 17 Hopkins. 18 A. The program in preventive medicine includes two 19 stages: You first get a degree in Public Health, 20 Masters in Public Health. You basically go back to 21 school. And after that, the second stage, second year 22 of it, once you have your degree, is some hands-on 23 training in various areas of practice of preventive 8 1 medicine. And after that, you became board eligible - 2 not right away. You have to work for a year, become 3 board eligible in preventive medicine. 4 Q. And what did you do upon your completion of the 5 program at Johns Hopkins? 6 A. At the end of my training, I had a six-month 7 appointment with the World Health Organization in 8 Jamaica, which did work on measles and polio 9 vaccination. 10 Q. And did you participate in a study of the - 11 for the United States government? 12 A. Yes. That was part of my training. I also was 13 another -- in another rotation that I had, I did a study 14 for the Department of Health and Human Services. It 15 actually was three -- a series of three studies on 16 vaccines and vaccine-related injuries. 17 Q. All right. And did you, in fact, obtain board 18 certification in the field of preventive medicine? 19 A. A year after graduation, you can sit for the 20 boards in preventive medicine, and that's what I did. 21 Q. I want to ask a quick question about your 22 training in pediatrics because, in the brief that was 23 submitted to the Court, there was a little bit of fun we 9 1 had, I think, I think we had you have a background in 2 pediatrics. 3 A. Right. 4 Q. Is it abnormal for people in the 5 epidemiological field to have had their medical training 6 in a wide variety of fields? 7 A. Those of us who go to -- entry epidemiologist 8 through clinical disciplines usually have a degree or 9 training or board certification in areas that are not 10 necessarily related to epidemiology. So, no, it's not 11 unusual. 12 Q. You've heard some of the testimony, you've 13 heard references to Dr. Rothman. 14 A. Yes, that's one of the most commonly cited 15 books in epidemiology, and Rothman is the editor. 16 Q. And Dr. Rothman is one of the leading 17 epidemiologists in the country; correct? 18 A. Perhaps, yes. 19 Q. Do you happen to know what his background is? 20 A. He's a dentist . 21 Q. He was a dentist before expanding into 22 epidemiology? 23 A. And if I may add, you know, the other book that 10 1 I like a lot personally is -- the one I was trained at 2 Hopkins, is introductory, more introductory course in 3 epidemiology, authored by Dr. Leon Gordis, who is a 4 pediatrician. 5 Q. Dr. Gordis is also a pediatrician? 6 A. He's a pediatrician. 7 Q. Dr. Gordis teaches epidemiology where? 8 A. At Hopkins. 9 Q. What was the first job you took after 10 completing all this training? 11 A. After my training was complete, my first job 12 was with company called Exponent. 13 Q. Tell us briefly about Exponent and what that 14 is. 15 A. Exponent is a large consulting and research 16 company. 17 Q. Publicly traded company? 18 A. I think so, yes. 19 Q. And what does Exponent do? 20 A. Exponent is really multidisciplinary group of 21 people. It includes several hundred experts in various 22 fields of knowledge, from engineering, to medicine, to 23 epidemiology, to behavioral sciences, and so forth. 11 1 Most of them Ph.D. or Masters-level training. 2 And the structure -- the way Exponent is 3 structured has three major groups. It has the -- what 4 is called the engineering group. They do various things 5 that are related to structures and a lot of physics, a 6 lot of engineer science. Then, there's a health group. 7 That includes physicians, biostatisticians, 8 epidemiologists, with a doctoral degree in Epidemiology, 9 Ph.D.s. And, then, there is an environmental group, and 10 that group includes toxicologists, people that are 11 well-versed in risk assessment and exposure assessments 12 and things of that nature. 13 Q. All right. You were in the health group? 14 A. I was in the health group. 15 Q. Can you tell the Court very briefly, just in 16 succinct terms, how you practiced epidemiology when you 17 were at Exponent. 18 A. It's really not that different than what I do 19 now at Emory. You conduct studies, you write reports, 20 you obtain funding for studies. I did quite a bit of 21 teaching while at Exponent, although not classroom 22 teaching, but residents. Preventive Medicine residents 23 from Hopkins would come to my office and just do 12 1 rotation, just like I did when I was a resident, while I 2 was still there. 3 And the only difference is that, at the 4 university, the bulk of your funding comes from the 5 government; whereas, at Exponent, the bulk of your 6 funding comes from industry. 7 Q. All right. Privately funded research? 8 A. Correct. 9 Q. Okay. But does Exponent also do work for the 10 government, for the federal government, for the state 11 government? 12 A. Yes, that' s true, yes. 13 Q. Now, finally, would you tell us about your 14 current responsibilities at Emory University. 15 A. I'm assistant professor. I started two years 16 ago. My charge is to develop a research program in 17 cancer epidemiology. So, first two years were very 18 busy, primarily writing grants to obtain funding for 19 research projects. 20 I also slowly increased my teaching load. I - 21 right now, I co-teach several courses but, you know, I 22 have a luxury of not being responsible for one single 23 course. So, I co-teach a course on Introductory 13 1 Epidemiology. I co-teach a course on Cancer 2 Epidemiology. I also teach the major part of the course 3 on Medical Research and Evidence Based Medicine for 4 people who are receiving their training in Pediatrics. 5 Q. There are specialties within the big field of 6 epidemiology; correct? 7 A. Correct. 8 Q. And within the field of epidemiology, are you 9 considered a specialist in what's called cancer 10 epidemiology? 11 A. This is what I do virtually hundred percent. 12 Well -13 Q. Virtually a hundred percent? 14 A. Eighty-five percent, 85 percent, virtually a 15 hundred. 16 Q. And just for the record, what sorts of cancer 17 studies are you now currently engaged in? 18 A. We have now seven research projects that 19 started between the time I -- I joined faculty at Emory 20 and today. I think it was seven at the last count. 21 Q. What types of cancer? 22 A. Breast cancer, lung cancer, prostate cancer, 23 colon cancer, tongue cancer. And there are also some 14 1 studies that deal with all cancers, more of looking at 2 accuracy of information about cancers that is collected 3 by registries. 4 Q. And I think you're teaching Introductory 5 Epidemiology, and you've also taught a class there on 6 prostate cancer; correct? 7 A. Prostate cancer, pediatric cancers are the 8 classes that I typically teach in the Cancer 9 Epidemiology course. 10 Q. And your current research or ongoing research 11 is funded by the American Cancer Society, CDC, the 12 National Canc er Institute, and the Department of 13 Defense, among others? 14 A. Correct. 15 Q. Have you also done research on particular types 16 of carcinogens like asbestos? 17 A. Yes, I have. 18 Q. All right. Tell us about that. 19 A. Well, I've done work that's related to 20 carcinogenicity of asbestos, of course. I've done some 21 methodological work that relates to carcinogenic 22 exposures for breast cancer. 23 Right now, most of work that I do, most of the 15 1 work that I'm hoping to do in the future as well, is 2 interaction, look at interaction between dietary 3 components, particularly the -- the modern diets and 4 genetic -- genetic factors. So, exposures in that case 5 would be components of what people eat. 6 Q. Now, on the screen, we have a paper I believe 7 you're lead author on. What is that paper? 8 A. That paper, I think it was my first publication 9 -- well , mayb e not first, but one of the earlier 10 publica tions with -- when I joined Exponent. And it's 11 metaana lysis of cancer in asbestos- exposed occupational 12 cohorts . 13 Q. And when did you complete that? 14 A. The paper came out in '99. 15 Q. Has that document or that study been cited by 16 the U.S . Gove rnment ? 17 A. Yes. 18 Q. Have you done other epidemiological research 19 into asbestos? 20 A. Yes. Last year, I coauthored two papers that 21 relate specifically to risk of mesothelioma and lung 22 cancer in motor-vehicle mechanics, people engaged in 23 motor-vehicle repair. 16 1 Q. All right. And did you present the results of 2 that research to any expert scientific group? 3 A. There was a meeting in December of 2003, if I'm 4 -- if I remember correctly, in Baltimore, where we first 5 presented those results. This is slide of poster. It 6 was a wall poster. 7 Q. Yeah. And what is the Society For Risk 8 Analysis? 9 A. Society For Risk Analysis is -- is a group of 10 people engaged in risk assessment. And it is a 11 multidisciplinary group. It includes epidemiologists 12 and toxicologists and regulators and so forth. 13 Q. And did you -- that's another slide, I guess, 14 from the presentation you gave. 15 A. Just second half of it, yes. 16 Q. Did the research actually get published? 17 A. Yes, it did. 18 Q. All right. And where was that published? 19 A. That came out in journal, in Annals of 20 Occupational Hygiene in June of 2004. 21 Q. June of 2004? 22 A. June, I think, yes. 23 Q. Any other asbestos-related research you've 17 1 published on this top ic? 2 A. Well, then, as I said, I wasn't the first 3 author, but I was co- author of another paper by Dr. 4 Patrick Hessel. 5 Q. And is that the study that's depicted on the 6 screen now? 7 A. Yes, that's the one. 8 Q. I want to di scuss asbestos in general. And 9 you've been here for some of -- or a lot of testimony 10 and you've heard a lo t of it, so I'd like go through it 11 as briefly as possibl e when there are points we agree 12 on. 13 You agree that there are different types of 14 asbestos and that word is sort of a generic term for a 15 wide variety of heterogeneous fibers; correct? 16 A. Yes, I agree with that. 17 Q. And do you agree that the circumstances of 18 exposure vary? 19 A. Yes, that's true. 20 Q. Have epidemiologists studied the relationship 21 between various exposure circumstances and the diseases 22 at issue? 23 A. Research started several decades ago, and it's 18 1 still ongoing. 2 Q. And just in summary, what have they found? 3 What have the researchers found? 4 A. Generally, they found that asbestos is 5 irrefutably, in many circumstances, causally related to 6 mesothelioma, lung cancer, and, of course, asbestosis. 7 Q. And how did they establish that? 8 A. Well, the reason why we know that asbestos is 9 related to, say, lung cancer is because there are 10 systematic epidemiological analytic studies. 11 Q. In other words, just to clear it up, was the 12 connection between asbestos fibers and disease 13 discovered in a laboratory? 14 A. The -- the connection was established in human 15 epidemiological studies. 16 Q. All right. Now, once you have epidemiology 17 establishing an increased risk, does that mean you know 18 everything you need to know? 19 A. No. 20 Q. All right. Why? 21 A. Well, because, as we touched upon just a minute 22 ago, the circumstances of exposure and the types of 23 materials that are called by term asbestos are extremely 19 1 heterogeneous and, therefore, the risks are - 2 heterogeneous, meaning discrepant, very different; and, 3 therefore, the risks also vary widely. 4 Q. Now, the Court has seen this quote from 5 Dr. Selikoff, who's one of the early pioneers in the 6 research, and I think you agree with what Dr. Selikoff 7 says here, don't you? 8 A. Yes. 9 Q. All right. And, specifically, when 10 Dr. Selikoff is talking about the differences not only 11 in the circumstances of exposure but, also, in the 12 fibers, he says, that -- about the middle there, 13 "Duration of exposure, variety, and a grade of asbestos 14 used." Do you know what he means when he says "variety 15 of grade"? 16 A. Grade is not my field. But as -- as I 17 understand it, grade is another term for fiber length. 18 Q. All right. And in the course of reviewing 19 documents in the asbestos litigation, are you aware of 20 some explanation of what grade means - 21 A. Yes. 22 Q. -- even though you never worked in the asbestos 23 industry? 20 1 A. Yes. There is a paper that's not necessarily 2 in my field but provides an explanation of what grades 3 mean and what -- how the various grades were used. 4 Q. All right. And what about that information is 5 important to you as an epidemiologist? 6 A. Well, if I look at applications of various 7 grades, and, particularly, if we are interested in 8 friction fibers, it appears that brakes and clutches 9 that would be sources of friction fiber exposures are, 10 at least in this publication, related to Grade 7, which 11 is the highest grade, the shortest fibers. 12 Q. Shortest fibers? 13 A. Yes. 14 Q. All right now, why is the difference in fibers 15 specially important when you're trying to figure out 16 whether there's an association between an exposure and a 17 disease? 18 A. As we discussed a minute ago, it is -- it is a 19 very heterogeneous group of potential exposures. So, 20 fiber length is only one component. But, clearly, there 21 is evidence that that may be an important component. 22 Chemical composition, of course, is another 23 important component. Various morphologic, meaning 21 1 structural, characteristics, as well as physical and 2 chemical characteristics of fibers themselves may -- I 3 don't know what they are, but they may affect the 4 biological effects. 5 Q. And as I think the other experts previously 6 testified to, there are various scientists who are 7 researching those differences to this present day; 8 correct? 9 A. As far as I know, it's an active area of 10 research. 11 Q. Right. And what we have up on the screen, I 12 believe, is taken from the EPA's peer review panel that 13 was put together by the contractor we've heard a lot 14 about. Is that a part of the report? 15 A. Yes, that's one of the statements in the 16 executive summary. 17 Q. All right. And do you agree with that? Let me 18 just -- "Factors that influence toxicity: Health 19 effects from asbestos and SVFs ultimately are functions 20 of fiber dose, fiber dimension, the length and diameter, 21 and fiber durability, or persistence in the lung, as 22 determined by the mineral type, the amorphous or 23 crystalline structure, and the surface chemistry." Do 22 1 you agree with that? 2 A. It is a very general statement, but I think 3 there's no reason to disagree with that. 4 Q. All right. Let's talk about epidemiology and 5 its application here. 6 What is -- in a broad sense, what is 7 epidemiology? 8 A. The textbook definition is that epidemiology is 9 a science, field of science, that concerns itself with 10 distribution and determinants of disease and other 11 health-related states in human populations. 12 Q. And how do you identify those associations? 13 A. Well, there are two parts to epidemiology. And 14 as the definition -- you know, the definition gives you 15 a clue: The distribution of how much, where, in what 16 circumstances of, say, disease. 17 Determinants is, what are other factors that 18 are associated and ultimately, what are the factors that 19 are causative of those diseases or health-related 20 states. 21 Q. Let me narrow it down to your field because 22 mesothelioma is a type of cancer, as is lung cancer, 23 obviously; right? 23 1 A. Correct. 2 Q. All right. Where does epidemiology fit 3 specifically my the matrix of research on cancer 4 investigation? 5 A. Well, cancer investigation, of course, is a 6 very broad -- broad term. One can look at cancer 7 survival, the treatment, or -- or access to care. But 8 one can also look at potential causes of cancer -9 Q. All right. 10 A. -- and what we call etiological research, 11 cancer research. And in etiological cancer research, 12 human epidemiological studies are the most important 13 tool in identifying potential risk factors and/or 14 causes. 15 Q. All right. Is that because we don't know all 16 of the causes? We can't - 17 A. Certainly. Most of the time, we don't know the 18 causes. It's an exception rather than the rule when we 19 do. 20 Q. And others, we don't have a full appreciation 21 and understanding of the complete biological mechanism 22 from which to predict outcomes; correct? 23 A. For the majority of cancers, we are just 24 1 scratching the surface now. 2 Q. All right. Now, the next slide -- go back. 3 The next slide is taken from something called the IARC 4 Monographs. What is IARC? 5 A. IARC stands for International Agency For 6 Research on Cancer. 7 Q. What is that? 8 A. IARC is a part of WHO, World Health 9 Organization. It's located in Lyon, France. And IARC 10 is -- one of the functions -- there's multiple 11 functions. But one of the functions of IARC is to 12 develop guidance documents for various potential 13 carcinogenic substances. 14 IARC does, but not much, regional research. 15 But most of what they do is convening panels of experts 16 and evaluating the evidence, weighting it, and, then, 17 determining whether or not a particular substance 18 provides exposure circumstances which entail increased 19 risk of cancer -20 Q. All right. 21 A. -- in humans. 22 Q. All right. Now, just let's explain this in a 23 little bit of detail. 25 1 IARC has a monograph on the methods in which 2 something can be classified as a carcinogen. 3 A. Correct. 4 Q. All right. 5 A. And -6 Q. Go ahead. 7 A. I should say, the preamble for IARC monographs, 8 it doesn't matter, it will be the same for all -- all 9 issues. 10 Q. Okay. All right. Thank you. 11 Group one, "The agent or mixture is 12 carcinogenic to humans," or "is considered carcinogenic 13 to humans." 14 A. Yes. 15 Q. That's what their group one would constitute: 16 "The exposure circumstance entails exposures that are 17 carcinogenic to humans." 18 A. Correct. 19 Q. You understand what they mean, then? 20 A. Yes. 21 Q. And the categories used, "When there is 22 sufficient evidence of carcinogenicity in humans, 23 exceptionally, an agent or mixture may be placed in this 26 1 category when evidence in humans is less than 2 sufficient, but there is sufficient evidence for 3 carcinogenicity in experimental animals and strong 4 evidence in exposed humans that the agent or mixtu re 5 acts through a relevant mechanism of carcinogenici ty." 6 You agree with that? 7 A. That's the most recent definition. Until about 8 a couple of years ago, the second part did not exi st 9 there. But, yes, that's the way they evaluate 10 carcinogens. 11 Q. All right. Two very important questions for 12 this proceeding that I want to ask you in a very 13 straight manner and get a succinct answer from you on. 14 If I ask the question, is asbestos -- I'm 15 asking for your conclusions on this, all right - is 16 asbestos carcinogenic in humans, using this rubric , what 17 is your conclusion? 18 A. Yes. 19 Q. If I ask you the same question about 20 chrysotile, meaning all of the heterogeneous fiber s that 21 we call chrysotile, what is your answer? 22 A. Yes. 23 Q. Yes, it is carcinogenic? 27 1 A. Yes. 2 Q. Now, how do epidemiologists investigate 3 associations? 4 A. The cornerstone of epidemiological research is 5 really the same as cornerstone of any analytical 6 research. And what -- what -- what's important is the 7 ability and the need to follow what's called a 8 scientific method. 9 Q. All right. Very briefly, because I think 10 Court's heard this, what is important about the 11 scientific method in epidemiology? 12 A. Scientific method is, of course, a concept that 13 caught in probably elementary school. 14 You begin with an observation that allows you 15 to generate a hypothesis. And, then -- then, the 16 research begins. Then, you set out to test that 17 hypothesis in a structured manner. 18 The most important thing about scientific 19 method that takes you from an observational -- from an 20 observation to a tested hypothesis is the ability to 21 utilize controls, or a comparison group. 22 You know, a very quick example from elementary 23 school would be, if I observe that plants thrive in the 28 1 presence of daylight, my hypothesis would be that plants 2 need light to grow. How I test that hypothesis is a 3 different matter. For instance, the way we did it when 4 I was in third grade, you take a jar, a glass jar, fill 5 it with cotton, add some water, put a bean on top, and 6 take an identical jar, an identical amount of water and 7 cotton and an identical bean; put one in the closet and 8 the other one on the windowsill. You come back ten days 9 later and, in that case, sure enough, one bean actually 10 develops sprouts and the other one was still dry. 11 Now, it was 30 of us in the class, and we 12 reproduced this experiment multiple times. That 13 scientific method is a very low elementary school level. 14 Interestingly, that conceptually, sort of from 15 cognitive point of view, research follows the same 16 approach. 17 Q. All right. And I appreciate that. But let's 18 move from it elementary school up to modern cancer 19 research. 20 What type of literature generates the 21 hypothesis that then gets tested? 22 A. Well, there is a variety of sources for 23 hypothesis. It can be mechanistic studies, for example, 29 1 using cell cultures or an animal experiment. Or, very 2 often, it is case reports. 3 Q. And we have heard that term run around. Just 4 define quickly what a case report is. And we've got a 5 slide on the screen to help you do that. 6 A. Case reports is observation, documented, a 7 series of observations documented by a, say, physician 8 or a researcher. Usually, it's the clinician, the 9 physician who treats patients, who documents and reports 10 those cases. 11 Let's say -- this slide shows a hypothetical 12 example which actually led to a -- there was a 13 hypothesis that was entertained in the literature for 14 quite a while, is that, let's say physician observes 15 one, two, maybe three cases of pancreatic cancer in his 16 or her practice, and he observes, also, that pancreatic 17 cancer, people who come to him, patients, are heavy 18 coffee drinkers. You know, the idea that legitimately 19 may occur, and it may become a hypothesis, is that 20 coffee drinking somehow increases the risk of pancreatic 21 cancer. 22 Q. Let me stop you right there. 23 It's not your opinion that that's illegitimate, 30 1 to have a concern raised if you see some cases come in? 2 A. No, no, that's how research begins, by 3 formulating the hypothesis. 4 Q. All right and is it, though, accurate science 5 to then reach the conclusion that some number must be 6 significant without a control group? 7 A. No. Well -8 Q. We're going to come to that. 9 A. Almost never, let's put it that way. Almost 10 never. There are exceptions. 11 Q. How do the epidemiologists then test the 12 hypothesis? 13 A. Well, so, we have a case report and, now, we 14 have a hypothesis. 15 The next question is, frequency. You know, 16 that's the first building block of an epidemiological 17 study. How common is pancreatic cancer among coffee 18 drinkers? We go out, recruit 1,000 volunteers at 19 Starbucks, say, in 1990 -- I don't know if Starbucks 20 existed back there. 21 Q. West Coast. 22 A. Yeah, in Seattle. 23 And follow them for ten years until the year 31 1 2000. And ten years later, we identified ten cases of 2 pancreatic cancer. So, of course, the numbers are 3 hypothetical. They're all rounded for simplicity. But 4 the point is that we took a group of people, we follow 5 them over time and, at the end of followup, we counted 6 cases. And now we know there are ten cases among 1,000 7 coffee, heavy coffee drinkers. 8 And the question that immediately arises after 9 that is, so what? Because we don't know whether, in 10 fact, that is a lot, or a few, or just what one would 11 expect. 12 Q. Because some people might have gotten the 13 disease without any coffee at all? 14 A. Because people get disease, no matter what. 15 Some people get disease, no matter what. But we don't 16 know. 17 So, the next step is to identify a similar 18 cohort. And I must say that this particular flow of 19 logic, flow of steps is a -- would be a cohort study. 20 Q. And let me just footnote this. We've talked 21 about two different kinds of studies: Predominantly, 22 cohort studies and case control studies, this is the 23 first, cohort study? 32 1 A. Cohort study, we begin identifying people with 2 and without exposure of interest -- in this case, coffee 3 consumption. So, what's missing is, we don't know what 4 happens in the absence of coffee consumption. And what 5 one can do is identify, say, a similar group of 1,000 6 people who don't drink coffee ever, follow them for ten 7 years, and count cases at the end of that followup. 8 Let's say we found ten cases. Now is the time 9 to put these two groups side by side and conduct the 10 analysis of what -- what does it all mean? 11 In this particular example, we have ten cases 12 that were observed where one can say it's ten cases per 13 thousand per -- per ten years, which is the same as one 14 case per thousand per year -15 Q. Okay. 16 A. -- per -- per thousand person years, in the 17 exposed group, exposed meaning coffee drinkers, and 18 exactly the same number in the nonexposed group. 19 Q. But we've oversimplified the math. It's 20 extremely rare that there's the same number of people in 21 the control group as in the study group; correct? 22 A. Almost never happens. 23 Q. None of the studies have equal numbers. It's 33 1 actually a ratio that you're looking at or a fraction; 2 correct? 3 A. It depends on the study design. It depends on 4 the study design. It can be a fraction in a cohort 5 study that calculates SMR. 6 Q. Okay. 7 A. But the concept is this: We divide the 8 observed -- cases among the exposed by the expected 9 number that would have occurred anyway and, then, if 10 that ratio is one, you conclude there is no difference, 11 or close to one. 12 Q. Now, very quickly, show us the same -- how the 13 math works, the same type of study when you might have a 14 positive finding, because what you just illustrated was 15 a negative finding, no increased risk. 16 A. Right. We conclude that coffee consumption is 17 not associated with pancreatic cancer. 18 A very similar conceptual example using 19 different disease and different exposure is lung cancer 20 among smokers. You begin by observing a lot of lung 21 cancer among people who tend to smoke. Hypothesis is, 22 is lung cancer caused by smoking or associated with it? 23 You begin -- always begin with an association before you 34 1 move on. 2 One thousand regular cigarette smokers, we 3 found 100 cases. Among non-smokers, you found only ten 4 cases at the end of cohort. You have put the two 5 results side by side, the difference is ten-fold, which 6 is actually not a hypothetical number. It's the number 7 one would expect in the regular smoker, or round 8 about -- can be much higher and that can be a little bit 9 lower, depending on how much. But the conclusion is, 10 smokers are ten times more likely to develop lung cancer 11 than do non-smokers. 12 Q. That's cohort. Briefly show us in reverse how 13 the case-control design works. 14 A. The case-control design is less intuitive. The 15 case-control design begins with disease in question. 16 Q. And we just need to make sure that this is 17 understood. Most of the studies we're going to talk 18 about on this topic are case-control studies; correct? 19 A. That is correct. 20 Q. All right. 21 A. The case-control studies begin with an illness 22 of interest. Let's take the same example of pancreatic 23 cancer and coffee and just play it backwards. 35 1 Let's say, in a hospital or a cancer -- or a 2 cancer registry, assuming we have a good cancer 3 registry, we identify 1000 cases of pancreatic cancer. 4 We interview these people and find out how many of these 5 are coffee drinkers. And let's say, in our example, 250 6 of those drink coffee regularly, which is 25 percent. 7 Then, the next step is, is it -- is this 8 percent similar or different in people without 9 pancreatic cancer, in healthy individuals, which one 10 would call controls. Let's say we identified 1,000 11 controls, we asked them about their coffee consumption, 12 or documented it any other way, and it turns out that 13 250 of those 1,000 healthy controls do not drink coffee, 14 do not -- do drink coffee and 75 don't. 15 You put these two results side by side. The 16 proportions of people who consume coffee -- in this 17 case, our hypothetical carcinogen -- are equal. 18 Q. All right. 19 A. And that allows you to conclude that there is 20 no increased risk. 21 Q. Now, that mathematically came out to 1.0; 22 correct? 23 A. Both instances, they mathematically, they 36 1 would be 1.0. 2 Q. That's a negative study? 3 A. Right. 4 Q. Statistically speaking, when epidemiologists 5 finish this research and analyze it, do they need it to 6 be 1.0? What conclusion do you draw if it comes out to 7 be .9 or .8 or 1.1, something statistically around that 8 1.0 bulls eye? 9 A. Each and every result, individual result, has 10 some uncertainty arising with what's called random 11 error. Random errors very simply means that there is a 12 chance that plays a role in what exact result you're 13 going to observe. So, that random error is quantified. 14 People usually use 95 percent confidence intervals, 15 around their best estimate. 16 Let's say your best estimate is one, but your 17 confidence interval goes from 0.7 to 1.3. That's fine. 18 What if your result is not 1, but, say, 1.5 and your 19 confidence interval comes from 0.4 to 2 or 2.2. The 20 confidence -- can I show it on the flip chart. 21 MR. TARRY: May he get down. 22 THE COURT: Yes, he may. 23 MR. TARRY: And, your Honor, if this is too 37 1 pedantic, we can move. 2 THE COURT: No, it's very helpful, thank you. 3 BY MR. TARRY: 4 A. We've already discussed that 1 means equal 5 risk in people who are exposed and unexposed. But the 6 real life study, the actual result is, say, 1.5. The 7 confidence interval. 8 Q. Confidence? 9 A. Confidence interval, what we called 95 percent, 10 by convention, confidence, CI, confidence interval, may 11 go from 0.4 to 2.2. The 1 is included in that interval 12 Therefore, researcher cannot conclude that his or her 13 results is different from what's called known, or 1. 14 Q. From that one study? 15 A. That one particular study. 16 Now, if the confidence interval, for instance, 17 goes from 1.2 to 1.7, the interval no longer includes 1 18 and, then, the conclusion would be that this result is 19 unlikely to be explained by chance alone. There may be 20 other explanations and other limitations but, by chance 21 alone, it's unlikely -- we have only 2.5 percent in each 22 way that your result is different than this interval. 23 Q. All right. Before you take the stand again 38 1 while you're up, let me ask you to do a couple of other 2 things. 3 We were talking about case-control studies. 4 And maybe using smoking and lung cancer again, can you 5 show the math, how the math works in a basic fashion, in 6 a case-controlled study where there's a positive 7 finding. 8 A. Yeah. 9 Q. And we'll agree, hypothetical - 10 A. The math is important because that's how - 11 they think, they think in two-by-two tables all the 12 time. 13 The premise is a two-by-two table, four cells. 14 You have disease and you have exposure. People can have 15 disease or they may be disease-free. They may have 16 exposure or they may get -- they may be exposed or they 17 may be nonexposed. This way, our study group is divided 18 into four categories. People that are sick and have 19 exposure or had exposure in the past; people who are 20 healthy, no disease, but have exposure; people who are 21 sick and have no exposure; and people who are healthy, 22 or don't have the disease of interest. 23 Q. They don't have the particular disease? 39 1 A. Control. 2 Q. Right. 3 A. And neither have the disease of interest, nor 4 the exposure of concern. They are labeled A, B, C, D. 5 Let's -- an example, we go to a -- to a 6 hospital and we identify a hundred cases of lung cancer. 7 And we take a hundred cases of trauma patients rolling 8 into the emergency room consecutively. We hope they're 9 representative of the general population. 10 Q. People from the trauma unit are going to be our 11 control group? 12 A. Our cases of lung cancers. Our controls, 13 maybe. That's one of the ways to do it, is, say, random 14 visits to the emergency room for trauma, for reasons 15 that are not cancer. 16 Q. I don't want to get you too sidetracked. But 17 is it important to actually try to get control groups 18 from the same area - 19 A. Yes, it is. 20 Q. -- if it's possible? 21 A. There's a variety of -- control selection is a 22 very important part of it. 23 Q. Okay. 40 1 A. So, we have a hundred cases. And we have 100 2 control. 3 We ask them, "Do you smoke?" or, "Did you smoke 4 in the past?" The most important thing is what happened 5 in the past. And -- which is not surprising. You might 6 find 80 people out of a hundred that are smokers with 7 the lung cancer, and maybe 20 that have lung cancer but 8 are nonsmokers. 9 Among your just people who roll in the 10 ambulance, for instance, you may have what's 11 representative of the general population. 25/75 is not 12 -- is not unreasonable. That's probably what -- number 13 of people who smoke among controls. 14 Q. Twenty-five people in the trauma unit smoke? 15 A. And 75 who don't. 16 Q. Right. 17 A. Now is the time to calculate the odds ratio. 18 But before do you the odds ratio, you do the odds. The 19 odds of smoking, if you have lung cancer, are 80 by 20 20 equal four to one. The odds of smoking, if you don't 21 have lung cancer, 25 by 75 equals one third. The ratio 22 of the two, that's the odds ratio. 23 Q. We need to verbalize this for the court 41 1 reporter. 2 A. I'm sorry. The ratio of the two, four divided 3 by one third equals 12. The conclusion is, exactly as 4 it is with the -- with the cohort study, although we did 5 it differently, is that people who smoke are 12 times 6 more likely to have lung cancer than people who don't. 7 Q. One last thing before you take the stand. 8 Another type of study we're going to discuss on 9 the friction product exposures is called a PMR, 10 proportionate mortality study. 11 A. Yes, it is a common design. 12 Q. Can you give a brief explanation of that study 13 design. 14 A. Yeah. PMR is proportion mortality, or 15 proportionate incidence study. 16 Q. Which would P-I - 17 A. P-I-R instead of P-M-R. 18 It deals with proportions. Typically, it's 19 mortality -- proportions of people who die from a 20 particular cause out of total number of deaths in a 21 given area and given period of time. 22 Let's say we're doing a study of -- our 23 hypothesis of interest is, do teachers develop stroke or 42 1 die from stroke more commonly than general population? 2 We take a certain -- you know, a well-defined area that 3 has death certificates. Death certificates, of course, 4 document two things: They document occupation and they 5 document cause of death. It's not the best way to 6 collect information, but it's available so people do 7 that. 8 Let's say we had a 100,000 deaths total in a 9 population of -- in -10 Q. Wilmington? 11 A. I don't know, Delaware, in five years. We have 12 that database, we have all these death certificates, 13 total 100,000 deaths. Of those, 2,000 were due to 14 stroke -- two percent. And we say, "Okay, among those, 15 let me just pull out all the teachers." Let's say 1,000 16 deaths out of our hundred thousand. 17 Q. One thousand stroke-related deaths? 18 A. No, total deaths. 19 Q. Total deaths, sorry. 20 A. Occurred among teachers. And these are 21 teachers. The other ones are everybody. And we counted 22 and we found that there were, let's say, 40 deaths due 23 to stroke. 43 1 Now, we know that the number in the general 2 populati on is two percent. Therefore , here, we would 3 have exp ected 20 deaths, 20 expected deaths, based on 4 two percent out of the 1,000. But observed number of 5 deaths is 40. Hence, the PMR, proportionate mortality 6 ratio, for teachers and stroke is 40 observed divided by 7 20 expected. 8 This is not necessarily a measure of relative 9 risk because we don't have people in the population. 10 This is a measure of proportionate mortality. 11 Q. Yeah. Are there other limitations of the study 12 design that make it, in your opinion, less reliable for 13 calculating the relative risk, or risk of an 14 occupational exposure? 15 A. Yes. The first, as I mentioned, is that you 16 don't have the underlying population of healthy people 17 from which to draw. Your underlying population are all 18 deaths as opposed to all people. 19 The second one is, usually, those types of 20 studies are limited by information. They just don't 21 have enough information. As I said, if it is a death 22 certificate study, you have the cause of death and you 23 have the occupation. You don't have a whole lot of what 44 1 happened before that. So, it certainly needs to be 2 considered as a major limitation. 3 Q. Those are some of the limitations that you 4 identified in the studies we're going to discuss later 5 when you performed the systematic review and 6 metaanalysis? 7 A. It's not that I identify. It's sort of well 8 known. 9 Q. That you identified in your paper that we're 10 going to talk about. 11 A. Right. 12 Q. You can take the stand. 13 Dr. Goodman, that's, by no means, all the 14 different study designs, is it? 15 A. No, there's a number of other studies that are 16 available. 17 Q. The last witness who testified yesterday, 18 Dr. Frank, mentioned in some detail a study by Nicholson 19 and others at Mount Sinai, up in New York City in the 20 early 1980s, pursuant to a NIOSH contract. Do you 21 remember that discussion? 22 A. Yes, I do. 23 Q. What kind of -- and there was -- I'm not sure 45 1 it was made clear on the record what kind of study that 2 was. What kind of study was the - 3 A. Yeah. That particular study, looked -- as 4 opposed to events -- you know, there are two things that 5 can happen to people. There are events, death, and 6 newly diagnosed diseases and death. We can have states, 7 where the onset is not known -- for instance, x-ray 8 abnormalities or blood pressure. Nobody knows, really, 9 when, at what point in time, somebody develops high 10 blood pressure. You know when person might be diagnosed 11 with blood pressure, but that's not necessarily -- I 12 mean, it's an event, but typically, blood pressure is a 13 long-lasting state that a person lives in. And because 14 you don't have the onset time, and because it's a 15 long-lasting event, those studies -- typically, there's 16 no other choice but to study them through what's called 17 cross-sectional designs. 18 Q. Cross-sectional? 19 A. Cross-sectional design. 20 Q. Right. Is that the design that was talked 21 about yesterday? 22 A. Yes. The difference between -23 Q. Do you need some water? 46 1 A. I'm fine. I have it. 2 The difference between a cohort study and a 3 case-control study is that, in a cross-sectional study 4 -- well, here's how I would explain it to students. 5 Let's say you're watching a who-done-it movie, 6 you know, the detective story. You -- a cohort study, 7 you start from the beginning and you watch all the way 8 to the end and you -- the event being the murder. A 9 cross -- a case-control study, the same movie is watched 10 backwards. You can still understand what's happening, 11 it's just backwards. 12 Q. You start with the murder and look to see who 13 had the gun? 14 A. Right. There is a school of thought, maybe 15 that's more logical to do it because nobody ever 16 investigates prospectively. This is a different issue. 17 A cross-sectional study, unlike the cohort and 18 a case control, is a snapshot. All we have is one frame 19 from a movie. You know at a given point in time, in a 20 two-by-two table, there was exposure and/or disease, or 21 none, but you don't know what preceded what. That's the 22 most important issue about cross-sectional studies, that 23 the issue of temporality, chicken-and-egg issue, is 47 1 unresolved in cross-sectional study. 2 Q. What are they used for? What are 3 cross-sectional studies? What's the usefulness of a 4 cross-sectional study? 5 A. The states -- as I said, the states of 6 conditions are -- are probably the only way to study. 7 You know, with cross sectional -- I mean, they're very 8 important and useful. The problem is because of absence 9 of temporality, you may have some trouble interpreting. 10 The only exception that textbooks usually say 11 is that, if you look at things that are present at 12 birth -- color of your eyes, that certainly doesn't 13 change, so you know that preceded whatever outcome 14 you're looking at. In other -- in other situations, 15 it's a snapshot. 16 Q. In your field -- and, in fact, what this Court 17 ultimately is going to have to do -- weight of the 18 evidence is important; correct? 19 A. Correct. 20 Q. And you've touched upon something called 21 consistency. I'd like you to explain. What does 22 consistency mean to an epidemiologist? 23 A. Well, even the scientific method in -- back in 48 1 elementary school requires that experiment should be 2 repeated. 3 Q. The experiment should be repeated? 4 A. Should be repeated. Same applies for 5 nonexperimental studies, observational studies, the ones 6 that are typically conducted in cancer research. 7 Q. And just let me footnote for the record again, 8 when we say observational studies and epidemiology, 9 they're sort of synonyms for each other; correct? 10 A. Well, you can have an epidemiological study 11 that are clinical trials, of course, too, which will be 12 an experiment. 13 Q. Right. 14 A. But in most circumstances, when you're looking 15 at potential harmful exposures, you're testing 16 hypothesis for disease causation, clinical trials are 17 not ethical and/or feasible. But -18 Q. What does -- yeah, I'm sorry for interrupting 19 you. 20 What would you define consistency to mean in 21 epidemiology? 22 A. What it means is, you have to collect all the 23 relevant information and evaluate it in its totality 49 1 and -- to see whether or not studies tends to agree or 2 disagree with each other. Studies that test the same 3 hypothesis, they may use different methods, can be done 4 at different places, can be done by different 5 researchers, but they have to test the same hypothesis. 6 That's important. 7 And that's where consistency -- I mean, that's 8 what people use before they draw conclusions, presence 9 or absence of an association. 10 Q. Can you explain it briefly to the Court, using 11 this slide that is a compilation of some different 12 studies and their results? 13 A. Yeah. Let me just take a minute and orient us 14 with respect to this -- the slide itself. 15 There's no cursor, is there? There's no 16 pointer or anything? It's all right. 17 MR. TARRY: Does the touch screen work? 18 THE COURT: It might work. 19 THE WITNESS: Oh, yeah. 20 BY MR. TARRY: 21 Q. There you go. 22 A. This is a graph showing real studies from - 23 conducted in real life. And they look at lung cancer, 50 1 relative risk of lung cancer associated with smoking 2 approximately one pack of cigarettes per day. The Y 3 axis -- the X axis, the horizontal axis, is a relative 4 risk estimate from those studies. Again, one is 5 important to find the one always because that's your 6 known elevation, meaning that risk is not elevated. 7 And, then, the Y axis, the vertical axis, is the one 8 that lists actual studies. It's the first author and, 9 typical fashion, how you show publication -- say 10 McLaughlin, Siemiatycki, and so forth. 11 And, then, what we do -- and that's one way to 12 do it. You can tabulate numbers. But it's -- visually, 13 it's frequently much easier to do it. You show the best 14 estimate. That's -- in this case, for McLaughlin, it's 15 roughly ten, and a confidence interval, because the 16 graph -- here, you see mostly the lower -- the lower 17 bound of the interval because we ran out of space on the 18 axis. 19 Q. Okay. But it's in the middle? 20 A. Yes, it's written, actually. It says ten and 21 the confidence interval, CI, between 8.8 and 11.2. So, 22 you know, that with a 95 confidence, you know that's 23 exactly 10. But with 95 percent confidence, you know 51 1 the result is somewhere between roughly 9 and 11. 2 Q. And then, statistically speaking, using the 3 methods employed by those researchers in those studies, 4 you would note that the statistical probability of there 5 being no association between smoking and lung cancer is 6 what? 7 A. It's very, very low, extremely low. But the 8 point is that, of course, there is a consistency of 9 findings. 10 Some of the studies don't have a confidence 11 interval because these are earlier studies done in the 12 seventies, that one done in the eighties, and they just 13 didn't calculate it, the confidence interval. 14 Q. We talked about -- just to juxtapose the 15 situation, we talked about the coffee, real-world coffee 16 example, bladder cancer. 17 A. Right. 18 Q. Can you explain what they found when they 19 studied that hypothesis, that there might be an 20 association. 21 A. Yeah. These are actually real-life studies of 22 coffee drinkers, and these are cohort studies. There 23 have been number of those over the years. These are 52 1 studies of relative risk of pancreatic cancer in coffee 2 drinkers compared to the general population or maybe, in 3 some cases, compared to nondrinkers. And, as you can 4 see, there is a -- there is a study, for instance by 5 Mills, M-I-L-L-S, 1988, that shows that almost more than 6 twofold increase in risk. 7 Q. Twice as many people he observed? 8 A. Correct. The number of observed compared to 9 the number of expected was more than twofold higher. 10 But, yeah, because the numbers were small, the 11 confidence interval is extremely wide, goes from -- all 12 the way from 8 to roughly 0.7, and includes 1. So, 13 that's -- we say, if this was the only study, you would 14 have to sit and wait for other evidence to come in. You 15 know, if other studies found the relative risk right 16 around 2, even though they may be nonsignificant, in 17 aggregate, you would have to say that, as the number of 18 studies increases, your overall precision increases. 19 In this case, of course, and when other studies 20 were put on the same graph, you know, that's just - 21 they just randomly scattered around 1 -- some above, 22 some below. 23 Q. Bottom line, does this present a picture of 53 1 consistency among the research or inconsistency? 2 A. Well, that would say consistent lack of 3 association. 4 Q. Even though there are a couple that 5 statistically identified what appeared to be to them 6 maybe a little bit of an increased risk? 7 A. It is typical to have one or two outliers. 8 Q. Okay. Thank you, Dr. Goodman. 9 Now, with respect to the testimony you've heard 10 in this courtroom about cancer and the biological 11 mechanisms of cancer, you agree that it's a broad term, 12 cancer? 13 A. Yes. It's a very broad term. 14 Q. Cancer occurs in different sites in the body? 15 A. Yes. 16 Q. And there are different histologies, which 17 means different cell types in cancer? 18 A. Right, different tissue characteristics. 19 Q. Why is that important to an epidemiologist? 20 A. Well, because cancer is certainly not one 21 disease. And even within a known category of cancer, 22 that's -- an example, good example, would be leukemia. 23 Not only leukemia can come from different cells in the 54 1 bone marrow, but even within the same cell category, 2 depending on the age, they behave as completely 3 different diseases. For example, acute lymphoblastic 4 leukemia is a different disease if it's contracted at 5 the age of one versus the age of four or the age of 12, 6 and completely different disease, or large different 7 disease, if it's contracted by a adult in the 40s or 8 50s. 9 Now, if say, lymphoblastic leukemia in a 10 four-year-old, it seems like a pretty narrow category. 11 But it isn't, because once people look at the actual 12 molecular features of those cancer cells, we know quite 13 a bit about it, they all have this fingerprint, a 14 chromosomal abnormality, fingerprint, that would 15 distinguish them even with that narrowly defined group 16 with extreme amount of variability with respect to 17 prognosis, treatment that's required -- some would 18 respond to certain chemotherapy and some will require 19 different chemotherapy -- and survival. 20 So, just to show how, you know -- cancer, of 21 course, is a -- is a category of diseases. But even if 22 you keep zooming in on them more and more specific 23 categories -- there's probably o end to it, there's no 55 1 end to our understanding of how different types of 2 cancer vary. 3 Q. Now, unlike Dr. Hammar, you don't diagnose 4 mesothelioma? 5 A. No, I don't. 6 Q. But mesothelioma being one of the cancers, does 7 that same analysis apply for mesothelioma in terms of 8 variety? 9 A. I'm not a pathologist, but I do know enough to 10 say that it's a very heterogeneous group of tumors. 11 MR. TARRY: Your Honor, at this time, we're 12 about to move in the nuts-and-bolts substances. If 13 there are questions about the fundamentals. 14 THE COURT: No. 15 MR. TARRY: We're ready to go. 16 THE COURT: Ready to go. 17 BY MR. TARRY: 18 Q. Dr. Goodman, have epidemiologists studied the 19 associations between circumstances of exposure to 20 asbestos and disease, in particular mesothelioma? 21 A. Yes, they have, extensively. 22 Q. All right. And can you just give us a quick 23 overview of some of the sorts of things they found? 56 1 A. Well, there are a umber of circumstances where 2 people exposed to asbestos have increased risk of 3 developing mesothelioma, and very large increase in 4 risk. 5 Q. Now, in the field of occupational epidemiology, 6 is the occupation often used as, what Checkoway terms, a 7 surrogate for exposure cat -- exposure circumstances? 8 A. Yes, that's one way to characterize exposure. 9 Q. But why do you do it that way? 10 A. Well, because, if you characterize occupation, 11 you may evaluate people who are unified by certain types 12 of exposures. 13 Q. Does that mean they all have exactly the same 14 exposure? 15 A. No, of course not. 16 Q. Okay. There is a variety of exposure even with 17 that cohort; correct? 18 A. Certainly. 19 Q. All right. Now, what are some of the cohorts 20 or some of the -- well, then, let's just clarify one 21 thing. 22 Is it your contention, or anybody else's, that 23 something about the job title inoculates anybody or 57 1 makes them more prone to disease? 2 A. Not that I know. 3 Q. It's a method of measure; correct? 4 A. Correct. 5 Q. All right. What are some of the occupations 6 where associations have ben strongly demonstrated? 7 A. Well, certainly, insulators is the first group 8 that comes to mind. They have been studied, of course, 9 since the 1960s. And I think we just jumped -10 Q. We just jumped ahead. That's all right. 11 A. And there is consistent evidence that the risk 12 is not only increased, but it increased very 13 dramatically for people who are characterized as 14 insulators. 15 Well, for instance, as I said, this is -- this 16 is the scale that goes from up until 10. But some of 17 the estimates are -- relative risk intervals are off the 18 charts. For instance, the study by McDonald gives you a 19 relative risk of 46. The study of Spirites gives you a 20 relative risk of three. The study by Ossiander -- it's 21 a Finnish study -- gives you a relative risk of 263. 22 Q. Now, Doctor, that sounds to me to be 23 inconsistent. That does sound like pretty widely 58 1 varying numbers. 2 A. Sure. They are variant, although, you know, 3 one thing that is consistent, the relative risk are all 4 elevated. So, one cannot say that the risk is not 5 increased -6 Q. And, in fact - 7 A. -- if a person in an insulator. 8 Q. Are those studies done by different people at 9 different times in different parts of the world, looking 10 at just different areas of exposure? 11 A. Yes. McDonald is Canadian. Spirites is United 12 States, Ossiander is Finland, Teschke is British 13 Columbia, and Milham ad Ossiander are from Washington 14 State. 15 Q. And back to this issue of consistency, is it a 16 good thing or bad thing that you have a diversity 17 amongst studies, when you accumulate that body of 18 evidence? Let me define what I'm talking about. 19 When I say diversity of studies, different 20 researchers, different studies looking at different 21 people? 22 A. Well, good or bad thing, I really don't know. 23 Q. Okay, that's a bad question. 59 1 When you have different studies by different 2 people, using potentially different designs at different 3 times, what does that do to aid you in determining 4 whether there's consistency or not? 5 A. Right. The conclusions are much easier when 6 you have consistency amongst studies. It depends what 7 conclusions we are ready to draw. 8 In this case, I'm certain that -- that 9 elevation of risk of mesothelioma among insulators is 10 present, and it's high -- you know, whatever you define 11 as high. 12 THE COURT: Can I ask one question - 13 MR. TARRY: Please do, at anytime. 14 THE COURT: -- so I'm clear. 15 So, the notion of consistency in 16 epidemiologi cal studies, is it driven by the hypothesis 17 that you're testing? In other words, if the hypothesis 18 is, insulato rs are at a higher relative risk of 19 developing disease, if all of the studies show that, in 20 fact, that is an accurate hypothesis, are those studies 21 then consistent, even though they may differ in regards 22 to the extent of the relative risk? 23 THE WITNESS Yes. 60 1 THE COURT: I mean, I guess what I'm asking is, 2 when you're looking at consistency, is that what you're 3 looking at, meaning are all of them showing a consistent 4 increase? 5 THE WITNESS: I understand the question. 6 The question can be posed differently. You can 7 say either evidence of increased risk. And the answer 8 is unequivocally yes. Now, the next questions could be, 9 is it 10? Then, I' not sure. It could be three, it 10 could be 46. But I know it's not one, that's for sure 11 I mean, it depends how you phrase it. But if 12 -- from a perspective of causality, it's just important 13 to know that the risk is elevated in those 14 circumstances. And that elevation is not trivial. It's 15 large. We can conclude that from these data. 16 BY MR. TARRY: 17 Q. Let's look at another cohort, shipyard workers. 18 Both with respect to insulators and shipyard workers, to 19 your understanding, is there any debate about the 20 association between exposure and disease in these 21 cohorts? 22 A. Yeah. As a group, as a group, shipyard 23 workers, by all means, are at increased risk of 61 1 developing mesothelioma. 2 Q. Are you aware of any debate about that in the 3 scientific community? 4 A. I don' t think so. 5 Q. Let's look at the next one. These are some 6 examples. But what about boiler installation and 7 repair? 8 A. Well, it's the same story, more or less. If 9 you look at the totality of evidence, the studies are 10 decisively on the right from the -- you do have an 11 outlier. I mean, Milham and Ossiander, 2001, shows no 12 increase in risk. I mean, if you're confronted with 13 that, if you're interested to find out what happened, 14 you might want to go into the actual study and dig 15 through it and find out what the reason is. Maybe 16 they're right and everybody else is wrong. That's 17 rarely the case. Typically, it's the other way around. 18 Q. How about plumbers and pipefitters? 19 A. Same -- really, the eyeball test is pretty 20 clear and unequivocal, that it's to the right from the 21 1. 22 Q. Now, let's talk about auto mechanics or car 23 repair. 62 1 Have epidemiologists studied auto mechanics and 2 mesothelioma? 3 A. Yes. 4 Q. And what have they found? 5 A. No study found an increase in -- a significant 6 increase in risk of mesothelioma among people engaged in 7 motor-vehicle repair. 8 Q. All right. And just to clarify, does that 9 finding mean that car mechanics aren't expected to come 10 down with mesothelioma from time to time? 11 A. No, it does not mean that. 12 Q. Does that mean that physicians and clinics are 13 not going to observe cases of mesothelioma in people who 14 work on cars? 15 A. No, it does not mean that. 16 Q. All right. Why is that? 17 A. Because we have epidemiology to show that the 18 risk is not elevated. 19 Q. But to reconcile these two things, you're 20 saying there's no association between the two -- why is 21 it that auto mechanics may still show up with 22 mesothelioma? 23 A. Because no occupations group, particularly 63 1 occupational group that is large, as large as people 2 engaged in motor-vehicle repair, will have zero cases. 3 You will have cases -- motor vehicle mechanics account 4 for several million people. Many of them are 5 blue-collar workers. 6 Q. How do you know that? You just stated -- how 7 do you know that there are that many people? 8 A. Well, there are estimates from the United 9 States, that number of people engaged professionally i 10 motor-vehicle repair is about five million. 11 Q. That's occupational? 12 A. That's an estimate. 13 Q. Wait a minute. Is that occupational, or is 14 that people who might go out in their yard? 15 A. Occupation. 16 Q. That just the occupational? 17 A. Yes. 18 Q. And o you have any estimate at all as to how 19 many people might be hobby mechanics? 20 A. No, no. 21 Q. It's a larger number than that; correct? 22 A. I would expect that's a large number. I don't 23 know what that number is. 64 1 Q. All right. So, in other words, you're not 2 talking about a small amount of people? 3 A. No. It's a -- it's a large occupational group. 4 Q. And even with a very rare disease, just 5 statistically, you're going to see cases; correct? 6 A. Correct. 7 Q. How many studies did you consider before 8 reaching your conclusion that you expressed to the 9 Court? 10 A. On mesothelioma, there are 18 studies that I've 11 counted to date. 12 Q. Let's talk about the first one. 13 THE COURT: Before we get into the individual 14 studies, maybe this might be a good tie to take our 15 morning break. 16 MR. TARRY: All right. 17 THE COURT: It seems like a logical break 18 point. Why don't we say we'll recess for ten minutes, 19 come back at ten past. 20 MR. CRUMPLAR: Your Honor, before you leave, 21 not related to this, but another asbestos matter, the 22 upcoming trial, I know that, originally, we were having 23 jury selection next Wednesday. I mean the 65 1 qualifications and then final selection on the 30th. 2 We've got a message that it moved to the 2nd. Mr. 3 Jacobs is going to try that, as me. 4 On the 2nd, which is Wednesday, will that be 5 final jury selection and opening statement, or is 6 everything moved back a week? 7 THE COURT: I honestly have no idea. 8 MR. CRUMPLAR: Okay. 9 THE COURT: I'm not driving the scheduling 10 train there. Commissioner White - 11 MR. CRUMPLAR: We'll check. We've called, and 12 we're trying to move our experts. 13 THE COURT: What I do gather, tough, is that - 14 because I understood that there was some concern about 15 the notice of the change, and Commissioner White says 16 that he gave notice of this change several weeks ago. 17 So, I'm not clear why that is a concern. In any event, 18 check with Commissioner White. 19 MR. CRUMPLAR: I will do that. Thank you. 20 (Short recess.) 21 MR. TARRY: May it please the Court. 22 THE COURT: Yes. 23 BY MR. TARRY: 66 1 Q. Dr. Goodman, all of the studies we're about to 2 discuss, almost all, are going to have cases of 3 mesothelioma where there was some history of work with 4 friction products or in an environment where that may 5 have occurred, correct? 6 A. Correct. 7 Q. All right. The Court will indulge me for maybe 8 beating a dead horse, but why is that not surprising 9 that there will be cases like that? 10 A. Yeah, any -- any occupational group that is 11 large will have cases. Cases will occur among teachers 12 and accountants. The question is whether or not 13 occupation puts a person at increased risk. You know, 14 everybody who's born on the day 1 of life is at risk for 15 disease, in this case mesothelioma. At some risk. 16 Mesotheliomas have been described in newborns. So the 17 risk starts immediately. And then over a period of life 18 you can do things or expose yourself to things that will 19 increase that risk, or you may be engaged in activities 20 that do not increase that risk and your risk can stay at 21 not increased level. 22 So in a group like garage mechanics, the risk 23 is exist. It's never zero. It's just wrong. I mean, a 67 1 newborn has a non-zero risk of mesothelioma, the moment 2 that person -- that newborn is breathing air and has 3 lungs. 4 And that's why, you know, it's not surprising 5 at all that groups like car mechanics or teachers or 6 accountants develop mesothelioma. 7 Q. And do you understand whether there's any 8 debate in the community about what percentage, albeit 9 hypothetically, of the mesotheliomas that occur may be 10 what we call spontaneous or may be caused by 11 mesothelioma? 12 A. There is a range of estimates, but there's no 13 way to measure it directly. 14 Q. All right . 15 A. But the range of estimates is somewhere between 16 10 percent and, you know, maybe 25 percent of -- well, 17 let me back up or a second. 18 There are several definitions of what one would 19 call spontaneous mesothelioma. There is a mesothelioma 20 that can occur in the absence of known occupational 21 exposure history. There is mesothelioma that can occur 22 without any exposure history, occupational or 23 non-occupational, and then finally is there a proportion 68 1 of mesotheliomas that can occur without asbestos. And, 2 you know, depending how you define it, you may have 3 different proportions in the total. But, you know, for 4 people without occupational asbestos history, it's about 5 20 percent. 6 Q. Well, and whether it's your opinion about it or 7 somebody else's opinion about it, is that a fact? Is 8 that something that we know and can observe or is that 9 an estimate or hypothesis? 10 A. Yes, that's been observed over and over again. 11 Q. It's beenobserved? 12 A. Right . 13 Q. The percentage has? 14 A. The fact that some people have unexplained, 15 unexplained cases of mesothelioma. 16 Q. Yes. Let me clarify because you might not have 17 understood the question. 18 A. Sorry. 19 Q. Any specific number that a researcher puts on 20 that, the percentage, is that a fact, is that a known 21 thing or is that more of an estimate or hypothesis? 22 A. Two things. One, it is an observed thing, if 23 you calculate a percentage of total cases. If your 69 1 denominator's the cases, then you can inquire about the 2 exposures and find out what proportion of people you 3 were unable to identify any exposures. 4 The other question is, what is the rate or the 5 incidence of mesothelioma among the unexposed? It's a 6 very different question. There are a measure now of 7 people without any exposure, and that's the trickiest 8 part of all. That's where you are in the estimate land. 9 You don't have measures directly. 10 Q. And that was one of the distinctions we drew 11 with Dr. Lemen, the difference between an incidence, 12 right, and a supposed background rate, right? 13 A. There is incidence and there is various 14 definitions of background rate. 15 Q. Last definitional thing before we go into the 16 studies themselves. The term "confounder" was used 17 yesterday by Dr. Lemen. What is a confounder? 18 A. A confounder is a factor. In a -- in a 19 epidemiological study evaluating association between the 20 exposure X and disease Y, a confounder is a third factor 21 Z that is associated with both X and Y. And the point 22 is that the best way to demonstrate a concept of 23 confounder is to show an example. 70 1 Q. Would you like to show it? Is that all right? 2 A. Yeah. Not even necessary. It1's fine. Some of 3 the studies that found a positive association, early 4 studies, between coffee drinking and pancreatic cancer, 5 they found an increased in risk in coffee drinkers, were 6 done -- were performed in a very crude way. They just 7 basically did a schematic thing that we showed on the 8 board. Do you drink coffee? Yes, no. Pancreatic 9 cancer; yes, no. 10 The point is that pancreatic cancer is known to 11 be associated with smoking. Relative risk is not as 12 high for lung cancer, but it is an established risk 13 factor for smoking. Now, smoking in this case also, 14 particularly in the older years, used to be associated 15 with cigarette -- with coffee consumption. So what you 16 have is a triangle. You have -- 17 Q. What do you - 18 A. I'll show you. 19 Q. When people who liked their coffee also used to 20 like their cigarettes? 21 A. Correct. This is pancreatic cancer. That's 22 your effect. This is coffee. That's hypothesized 23 carcinogen. If you don't take into account smoking, you 71 1 will observe an association between coffee and 2 pancreatic cancer because you fail to look at people who 3 smoke. More of coffee drinkers will be smoking than 4 non-coffee drinkers. And, of course, more pancreatic 5 cancer patients will be smoking than with people without 6 the disease. 7 And that's a confounder. That's what a 8 textbook of confounder definition is. 9 Q. Let me take -- go ahead. Take the stand. 10 But with regard to the studies we're about to 11 discuss, we showed slides where certain circumstances of 12 exposure to given products and a given environment 13 showed a strong association with the disease, correct? 14 A. Yes . 15 Q. Shipyard? 16 A. Yes . 17 Q. All right. Can those exposures, because 18 they've been demonstrated in the absence of anything 19 else to have an association consistently with the 20 disease ? 21 A. With the disease. 22 Q. Can those exposures become confounders in other 23 studies? 72 1 A. Well, same triangle. Same concept. You have 2 exposures that are clearly associated with disease, 3 insulation, shipyard work. You also know that car 4 mechanics are blue-collar workers. They're manual 5 workers. These manual workers will be more likely to 6 have exposures elsewhere than people, say, of 7 white-collar occupations. And that's the issue of 8 confounding. In this particular case, the most powerful 9 confounder of all is that third factor, is other 10 asbestos exposures that need to be considered, if 11 possible. 12 Q. And, in fact, are some of these studies going 13 to attempt to control for those confounders? 14 A. Yes. 15 Q. All right. And the reason you control for 16 those confounders gets back to the question about why 17 you see cases to begin with, correct? 18 A. Yeah. Can I explain a confounder? Just one 19 more thing. 20 Q. Yes. 21 A. Because I think it is important. There are 22 two -- two ways you can deal with it, with a confounder. 23 You can match. If I'm doing a study of coffee and 73 1 pancreatic cancer, I can make sure that my cases in my 2 controls are matched by smoking. For each smoking case 3 I have a smoking control, and for each nonsmoking case I 4 have nonsmoking control. Then the only difference 5 between the two would be coffee, but not cigarettes 6 anymore because cigarettes are controlled. They are - 7 they're fixed. The frequency of smoking will be the 8 same . 9 So you can match. The other thing you can do 10 is if the data set is unmatched, you have a mishmash of 11 smokers and nonsmokers in both groups and they're not 12 the same, then you can do what's called adjust. And 13 that's an analytical technique that allows you to 14 calculate the relative risk after taking into 15 consideration smoking, in other words, the more pure 16 result. 17 Q. And very briefly, how do you adjust for that? 18 A. Well, there's a number of ways to do it. There 19 is so called stratified analysis. 20 Q. Stratified? 21 A. Stratified analysis. When you -- I have a 22 slide of that, but it's printed. I can probably... 23 Q. Put it on the overhead. Can you -- Doctor, you 74 1 can get down. 2 A. We have our cases and we have our cont rols. 3 have people who drink coffee and those who don' t. If 4 one were to take this two-by-two table at face value, 5 they would find the odds ratio is 2.15. That's the 6 bottom bullet. In other words, if the numbers were just 7 taken as is, you would find a two-fold increase in risk 8 among people who drink coffee compared to those who 9 don't. 10 Now, the next step is what you do is stratify. 11 The s ame two-by-two table, now it's split into two 12 table s, smokers and nonsmokers. The numbers all add up. 13 It's the same numbers from -- but the distributions of 14 them within cells are different now depending on their 15 smoki ng status. 16 Now, if one were to -- so the left side is 17 smoke rs. The right side is nonsmokers. Smokers alone 18 gives you a relative risk of 1. The association went 19 away. Nonsmokers alone give you a relative risk of 1. 20 The a ssociation is no longer observed. And that's, you 21 know, the definition of confounding, and that's how you 22 deal with it in a stratified analysis. 23 Q. Dr. Goodman, just to wrap that up then. Given 75 1 the presence of confounders and given the fact that this 2 is a disease that, albeit rare, does appear to occur 3 naturally in the population even in the absence of 4 asbestos exposure, is it surprising for researchers to 5 find cases of mesothelioma among car mechanics or shade 6 tree mechanics or spouses of them? 7 A. No. 8 Q. McDonald and McDonald, is it an epidemiological 9 study? 10 A. Yes, it is . 11 Q. Which study design is the McDonald study? 12 A. That's a case control study. 13 Q. All right. And how did they do it? Where did 14 they do it? 15 A. The study was conducted in -- throughout the 16 United States and Canada. I think the Canadian data 17 were collected over a period of 13 years or so. And the 18 U.S. data were collected just for one year, 1973. 19 Q. Is it just a study of garage mechanics? 20 A. No. Well, yeah. An important feature of case 21 control studies, they allow you to test several 22 hypotheses at the same time because if you have 23 occupational information, for example, or exposure 76 1 information, you don't have to limit yourself to just 2 one exposure. One can say in a study of pancreatic 3 cancer, one can look at coffee, ice cream, and 4 cigarettes at the same time as long as you adjust for 5 them for each other. 6 So, therefore, rarely do you have case control 7 studies limited to a single exposure or single 8 occupational group. Usually, they look at several. 9 Q. So are they testing a hypothesis about one 10 circumstance of exposure or about a whole bunch of them? 11 A. A whole bunch. 12 Q. Tell us what they did. 13 A. What they did, they contacted members of the 14 pathology societies in the U.S. and Canada and asked 15 them to report each and every case of mesothelioma that 16 was diagnosed. As I said, in Canada about 12 years. In 17 U.S. just one year, bigger country. 18 And then -- this is a case control study, but 19 so-called hospital-based case control study. Then they 20 said, We'll compare them to people who have cancer in 21 the lungs, but metastasizing from other sites. And this 22 has advantages and disadvantages, of course. 23 Q. What is the disadvantage? 77 1 A. Well, the disadvantage is you don't have -- the 2 ideal case control study takes populations that's 3 representative of the general population. Your control 4 are representative of the general population. There has 5 to be a sample of people at large. People with 6 metastases into the lungs certainly are not 7 representative of the population at large. That's - 8 you know, that's an important issue. It's not a 9 confounder. It's a limitation. 10 Q. Right. That's a limitation that the authors 11 noted in their section called limitations? 12 A. I'm pretty sure they did, and if they didn't, 13 we probably did. 14 Q. That's a limitation of the study that you 15 observed in your meta-analysis paper? 16 A. Yeah, we noted that. 17 Q. You noted that. And just in case this wasn't 18 clear yesterday, all studies that go through the 19 peer-review publication process have limitations, 20 correct? 21 A. Correct. 22 Q. And, in fact, part of the paper is a section 23 entitled limitations, right? 78 1 A. Usually people say a thing or two about their 2 reservations or cautions. 3 Q. So if we wanted to throw some criticism at a 4 paper, it would be pretty easy because the authors 5 always tell us the limitations based on the hypothesis 6 and the way they went about trying to test the 7 hypothesis, correct? 8 A. Most of the time, you know, it's sort of part 9 of the discussion section. That has to be -- well, not 10 doesn't have to be, but typically it is. 11 Q. All right. What did the McDonalds find? 12 A. Well, they had an interesting analysis. They 13 took the occupational history in what they called 14 hierarchical or maybe what one would call an 15 hierarchical manner. If the person is known to have 16 occupation that's very high risk, say an insulator, they 17 would rank them very high. So if somebody was an 18 insulator and then a shipyard worker and then a piano 19 tuner, these people would be counted as insulators. 20 That's sort of -- it's a clever way of taking care of 21 these powerful confounders because it's much more 22 important to know that he was an insulator than that he 23 was a piano tuner. 79 1 If, for example, the person was a shipyard, but 2 never an insulator, he would be called a shipyard worker 3 and so forth until they got down to the category for 4 which they -- you know, the bottom of the hierarchy. 5 And there they have a total of 156 cases and 6 156 controls were matched by age to take care of 7 confounding by age, the matched analysis. And among 8 those 156, that's after removing the - 9 MR. TARRY: May he leave the stand? 10 THE COURT: Yes. 11 THE WITNESS: After removing the heavy exposed. 12 If you recall, the relative risk for insulators was 46 13 there. So after removing those people, they ended up 14 with 11 what they called garage workers, among cases, 15 and 12 among controls. The total number is 156 for 16 each. And that will leave us with 145 and 144 here. 17 This two-by-two table gives you an odd ratio of 0. 9 with 18 a confidence interval from . 35 to 2. 3. 19 BY MR. TARRY: 20 Q. So -- but did they conclude that a . 9 was a 21 protective finding? 22 A. No, not at all. 23 Q. All right. 80 1 A. Not at all. I mean, nobody would conclude 2 that . 3 Q. We've got the quote on the call-out on the 4 screen. What did they conclude? 5 A. They stated no increased in risk was found in 6 garage workers, certainly exposed to chrysotile from 7 brake linings. 8 Q. They considered that an exposed group that 9 potentially might have a risk? 10 A. Yes. They felt that was important enough to 11 mention that in the -- in the discussion of the results. 12 Q. Anything else particularly important he needs 13 to understand about the McDonald study? 14 A. Well, there's a lot of things that they've 15 really done. You know, there's some strengths of the 16 study as well. 17 Q. What is some of the strengths of the study? 18 A. Well, this hierarchical category situation, 19 it's a very clever way to do it. I mean, I think that's 20 the only -- the only study that did it that way, that 21 particular way, because what you're doing is you're 22 removing really powerful -- you're removing all of the 23 heavy artillery before you analyze other occupations. 81 1 There are other ways to do it, but that's one way to do 2 it and I think it's a very interesting way to do it. 3 The other thing is, because it's an interview 4 study, the problem -- and it's a next-of-kin interview, 5 of course. So the problem of doing healthy controls, on 6 the one hand, it's good because they are representative 7 of the general population. On the other hand, their 8 recall would be very different, if you have people that 9 are alive and healthy and say families of people who are 10 deceased from mesothelioma. So that may be another 11 consideration. One, they said, We'll take another fatal 12 disease that manifests itself in the lungs, but not a 13 lung cancer. And that way the recall from next of kin 14 may be not perfect, but it's comparable. It's not 15 skewed. I mean, that's the thinking behind it. 16 So it has -- you know, like anything, it has 17 its strengths and weaknesses. 18 Q. What's the next study you considered? 19 A. I think in - 20 THE COURT: Can I stop you on the McDonald 21 study? I just want to -- absorbing what you just said. 22 Next-of-kin interviews were done both for -- 23 THE WITNESS: Yes . 82 1 THE COURT: -- the control set and - 2 THE WITNESS: And for the cases. 3 THE COURT: So is that a matching sort of? 4 THE WITNESS: Yes, I mean -- right. That - 5 why do I say that? That's one way to make sure 6 that even if information is incomplete, it's equally 7 incomplete. You -- there is a tip of the iceberg, 8 almost always, but at least you're comparing the 9 equal -- the two icebergs have equal chance of showing 10 the tip. 11 BY MR. TARRY: 12 Q. It might be appropriate to talk about 13 differential or non-differential at this point. Do you 14 want to do that? We can wait. Okay. 15 THE COURT: My question was answered. 16 MR. TARRY: Okay. 17 THE COURT: Thank you. 18 BY MR. TARRY: 19 Q. What's the next study you considered? 20 A. The next study in chronological order was 21 published in 1980. And it's actually not on the screen. 22 Q. It's '83, correct? 23 A. No, 1980. That's a study by Peterson and 83 1 Milham. It's not on the screen, and the reason it's not 2 on the screen is it was a PMR study, large PMR study, of 3 deaths at -- the proportionate mortality ratio site of 4 all deaths in California that happened between 1959 and 5 1961. That's only two years, of course, but it's a 6 large state so it's quite a few deaths. 7 And they looked at various occupational groups 8 and evaluated proportionate mortality ratios for various 9 groups. For the category defined as mechanics and 10 repairman, automobile there was zero cases of pleural 11 mesothelioma and zero cases of peritoneal mesothelioma. 12 And there's nothing to show on the screen because they 13 just found nothing, but it is still a study that 14 addressed that issue. 15 Q. Just clarify that. They didn't find any cases 16 so there's no way to do a meaningful two-by-two 17 calculation, if it had been that design, and there's no 18 way to calculate PMR if you don't have -- 19 A. No, you can't. You just have no cases. 20 Q. What was the next study? 21 A. The next study is Teta. That11 s 1983. 22 Q. We've heard some about that. That's the 23 mesothelioma Connecticut study? 84 1 A. Yes . 2 Q. And what study design did they employ? 3 A. Teta is a case -- population-based case control 4 study. The mesotheliomas were taken from the 5 Connec icut Tumor Registry. 6 Q. We've heard about another registry. There's a 7 regist y also the State of Connecticut keeps? 8 A. Well, yes, that actually goes back to the 9 1950s, 1955. SEERR which was mentioned here early, 10 SEERR, Surveillance and Epidemiology End Results 11 Regist y, was established in 1973. So Connecticut was 12 basica ly an earlier study proceeding by 20 years. 13 Q. Let me just stop you. Who conducted this 14 study? Dr. Teta, obviously? 15 A. Dr. Teta. 16 Q. And her colleagues, where were they? 17 A. I think they were at Yale. 18 Q. Yale University? 19 A. Uh-hmm. 20 Q. And do you recall who funded that study? 21 A. No, I don't. 22 Q. Okay. We might get to that later. What did 23 they do and what did they conclude? 85 1 A. The controls were death from all causes, 2 regardless of diagnosis. And they again looked at a 3 variety of occupational categories. In this particular 4 group they used the category called automobile repair 5 and related services. The results do not show the 6 entire breakdown of cases and controls, but -7 Q. This is one chart from the study? 8 A. That's right. But the chart highlights the 9 result that's most pertinent to -- to the hypothesis at 10 question here. 11 And automobile repair and related service has 12 an odds ratio of .65. Actually, they call it relative 13 risk in the table, but it's odds ratio. And that's 14 based in one case and five controls in the total group 15 of six exposed individuals. 16 Q. Now - 17 A. And that's, of course, it's a tiny number, 18 really small. 19 Q. Now, how did they determine what somebody's job 20 was? 21 A. Yeah. I mean, the feature of Teta studies that 22 unlike, say, McDonald, they did not interview people. 23 Q. Okay. So there was no interview done? 86 1 A. There was no interview done. Connecticut has, 2 you know, an interesting feature. They have what's 3 called town directories. So people just keep logs of 4 their residence and what they do for a living. 5 Q. Is that like a phone book with occupation? 6 A. Yeah, I've never seen it, but Dr. Teta 7 described it to me. And you can -- you know, you find a 8 person. You go to find a town book that precedes the 9 death, of course. And you try and reconstruct the 10 occupational history. 11 Q. Will those towns put them out every year or 12 over a certain period? 13 A. I don't know the details, but I know those were 14 available and I'm pretty sure she describes it. The 15 thing here is, on the one hand, of course, if the record 16 is incomplete in the book, then you have incomplete 17 information and you may have to go to the death 18 certificate. On the other hand, there's no recall here 19 involved. This was information collected before. In 20 other words, it wasn't -- before anything happened to 21 these people. And, therefore, you don't expect it to be 22 skewed one way or the other by -- by just relying on 23 recall. 87 1 Q. Let me just give you an example. If you wanted 2 to know what that particular subject was doing 3 occupationally in 1964, did they ask them to recall what 4 they were doing in 1964 or did they go to a document 5 that showed what they were doing in 1964? 6 A. No, there were no interviews. They went to 7 these town registries. 8 Q. Is there anything else about the Teta study 9 that the Court should know about in your opinion? 10 A. Well, the main limitation, of course, is small 11 sample size. That's really the -- with a confident - 12 if this were the only study on earth, you know, and I 13 had -- one had to draw conclusions. You look at the 14 result of .65 confidence interval from .08, which is 15 tenfold decrease in risk to a 12-fold decrease in risk 16 to 5.53, which is almost six-fold increase in risk. You 17 don't know. The only way this study may be of -- needs 18 to be evaluated is as part of the totality of the 19 evidence. 20 Q. What's the next study you considered? 21 A. Spirtas. 22 Q. And who is Dr. Spirtas? 23 A. Spirta still works for National Cancer 88 1 Institute. And he's the one who started the study back 2 in the '70s. He's an industrial hygienist and 3 epidemiologist, works for the NCI. 4 This is the first publication of their study 5 that now had I think three different papers published 6 based on those data. It's an abstract that was 7 presented at the Society For Epidemiological Research 8 back in 1985. And as an abstract, you don't really have 9 much information. They just don't have enough space 10 there. But the point is that -- two things that are 11 important. One, it's -- it's a registry-based study 12 where they find cases in the population-based tumor 13 registry, Los Angelos and New York. 14 The other thing is that controls were selected 15 from V.A. Hospitals, Veteran Administration Hospitals. 16 The reason why you may want to get controls from V.A. 17 Hospital, it's much easier to contact patients and their 18 families if you have to conduct the next-of-kin 19 interviews. People don't leave the system. V.A., from 20 that point of view, it's a very good study population 21 because they're stable. They don't move a whole lot, 22 and they stay within the V.A. system even if they move. 23 So the controls were people from the V.A. 89 1 Hospitals as well as -- but the only thing they did they 2 removed cancers and other respiratory disease. 3 Q. How many subjects did they have in the study, 4 approximately? 5 A. In this particular study, they had 259 cases 6 and 259 age-matched controls. They made sure that the --i 1 0 u 7 cont is the same age as the case . 8 Q. Just so we're clear. 259 cases of what? 9 A. Malignant mesothelioma . 10 Q. Rare disease, is that a big number or a small 11 numb er? 12 A. For a case control study it's a nice number. 13 You know, it's a good, big number. 14 And it actually shows in the confidence 15 interval that the larger your sample size -- in the 16 two-by-two tables the bigger the number, the tighter 17 your confidence interval is. And here it's 1 and 18 it's . 6 to 1.6. If you compare it to Teta, for example, 19 it's a much larger sample. 20 Q. Now, you mentioned that there were two subject 21 papers based on the same National Cancer Institute data 22 set. Tell me about the next one. 23 A. The next study was published in '94 by Spirtas. 90 1 It did not address the hypothesis of individual 2 occupations. 3 Q. What did it do ? 4 A. They set out t o test a hypothesis for, or 5 answer the questi on as to what is the contribution of 6 asbestos exposure from all sources to mesothelioma in 7 terms of relative risk, exposed versus non-exposed, and 8 in terms of what' s call ed attributable risk. That is, 9 the proportion of peopl e whose disease can be attributed 10 to asbestos expos ures . 11 Q. This is sort o f back to that question about 12 what percentage o f meso theliomas might be related to 13 asbestos ? 14 A. Correct. I mean, people -- a lot of people 15 besides 20 percent -- in men. It's in men. Does not 16 apply to women. Without occupational exposures use the 17 Spirtas estimate. This actually comes from the Spirtas 18 '94 study. But the study did not -- did not calculate 19 odds ratio, relative risk estimates, specifically for - 20 the category's important, brake lining installation and 21 repair. It's not garage mechanics or motor mechanics. 22 They're much more specific in their questioning. 23 Now, another useful thing about Spirtas is that 91 1 they actually described the difficulty of exposures - 2 multiple exposures during a person's lifetime. If I 3 recall correctly, they said that 33 percent of -- they 4 use brake mechanics or brake installation repair people 5 as an example. 33 percent of those work in shipyards 6 and 55 percent of those had previous history of 7 working -- working as insulators. 8 Q. Let me ask you to repeat that again so we're 9 clear. Of the cases they were looking at where there 10 was a history of automotive work, right? 11 A. Well, I don't think it was cases. Everybody 12 who was involved in brake installation and repair. 13 Q. Right. Case end controls? 14 A. Case end controls. 15 Q. Okay. What percentage of them had high risk 16 occupations elsewhere? 17 A. If I recall correctly, 33 had ship building and 18 shipyard experience, and 55 percent had performed 19 insulation work. 20 Q. What was the next study you considered? 21 A. Woitowitz and Rodelsperger, '94. 22 Q. Now, we've heard about them. Who are Woitowitz 23 and Rodelsperger? Do you know those researchers? 92 1 A. Yeah, I know both of them. Woitowitz and 2 Rodelsperger run an occupational health clinic in a town 3 called Gechingen in Germany, not far from Frankfurt. 4 Q. Now, we've got a call-out on the screen, and 5 this is from the 1994 study. But so that we set the 6 historical record straight, before they published this 7 study, did they also publish sort of a case series 8 study? 9 A. Yes, they did. 10 Q. And what -- where they raised a hypothesis, 11 right? 12 A. Well, actually, in their mind they thought they 13 answered the hypothesis when they wrote the paper. They 14 say, We observed four mechanics in -- four cases of 15 mesothelioma in our clinic among car mechanics. And, 16 therefore, we think being a car mechanic puts one at 17 risk -- at increased risk of developing mesothelioma. 18 And that was a case report. 19 Q. Why did they think that? 20 A. I think it was published in '91, three years 21 before this paper. 22 Q. Why did they think that, just because they had 23 four cases? 93 1 A. Well, yeah. They shouldn't have said that. 2 Q. All right . 3 A. The point is they shouldn't have said that. 4 Q. All right. And did they later then go and get 5 a control group to test that hypothesis to see whether 6 there was an association? 7 A. Yes. Then they got two groups of controls. 8 One in the population, from the population. I don't 9 know how they selected them. It's not very well 10 described in the paper. And one in the hospitals when 11 they would pick patients who underwent lung resection, 12 in other words, people who had a piece of their lung 13 removed for whatever reason, maybe lung cancer. 14 Q. Right. Let me stop you. When they were 15 constructing their actual analytical study this time 16 around, did they do something interesting with respect 17 to this question that was raised yesterday about maybe 18 differences in the risk of people who did brake repair 19 versus people who were just employed as mechanics? 20 A. Yes, they had several categories. One category 21 was called motor mechanic or automotive mechanics, and 22 the other category was called definitely engaged in 23 brake installation and repair. Doesn't mean that others 94 1 were not, but those that they know for sure were engaged 2 in brake installation and repair. 3 Q. Why is that important? What does that allow 4 them to do, as I think you're going to show us? 5 A. What one can do this way is define your 6 exposure in two different ways. One is the sort of a 7 more vague category of motor mechanics, and the other 8 was much more specific of people who we know were 9 exposed to friction fibers. And the results were not 10 all that different. They were very comparable. If one 11 were to combine both sets of controls, you have odds 12 ratio for all mechanics, 0.87, and it's non-significant, 13 the confidence interval between the 0. 4 and 1. 7 -- 0. 43 14 and 1. 7. And if you take a subset with definitely 15 engaged in brake installation and repair, then your odds 16 ratios are 0.89. 17 MR. TARRY: Your Honor. 18 THE WITNESS: Virtually the same. 19 MR. TARRY: Have we defined statistical 20 significance? 21 THE COURT: Yes. 22 MR. TARRY: We did. 23 THE WITNESS: Virtually the same. Both result 95 1 in non- significant. And, of course, the number of 2 people definitely engaged in brake installation and 3 repair is smaller because it1' s a subset. So your 4 confide nce interval is a l ittle bit wider . It goes 5 from . 3 1 to 2.47. 6 BY MR. TARRY: 7 Q. What's important about the Woitowitz and 8 Rodelsperger analytical study? 9 A. Yeah, it's not the grea test study, but what 10 they did is they asked this ques tion twice in differe nt 11 ways. They asked car mechanics and they asked brake 12 mechanics. And the answer is re ally indistinguishabl e. 13 They couldn't tell the differenc e between one and the 14 other. 15 Q. What's the next study? 16 A. Well, it's important to know their conclusion, 17 though. 18 Q. Go back, because - 19 A. They're very cautious. They say, From these 20 results, there is no evidence that car mechanics are 21 exposed to an increased risk of mesothelioma even if 22 they do brake repairs. And then they can say, But 23 asbestos exposure in other employment is an important 96 1 confounding factor. So that if there is a mesothelioma 2 risk for car mechanics, but if it were small, it would 3 not be detectable. 4 Q. Do you agree with that conclusion? 5 A. It's very fair, yes. 6 Q. What was the next study you considered? 7 A. Coggon. 8 Q. Who is Coggon and what kind of study is that? 9 A. Coggon is -- I think he's from the University 10 of South Hampton in England. It's a PMR study, 11 proportionate mortality ratio study, that took place in 12 all of England and Whales between '79 and '80 and then 13 between 1982 and 1990. 14 Q. Does that mean they're looking at all of the 15 deaths and causes of deaths between 1979, and what was 16 the other -- what was the last one? 17 A. And ''90. 18 Q. And '1 90? 19 A. With a two-year interval in between. I don't 20 know why. Maybe they didn't have the data. 21 Q. Does that mean they're looking at all of the 22 causes of death within that time period within that 23 region? 97 1 A. Correct. 2 Q. And what did they conclude? 3 A. Well, first they found that they observed a 4 number of cases for category as they define it -- well, 5 the cases are defined not as mesothelioma. It's a 6 different definition, pleural cancer. It's important to 7 know that although the majority of pleural cancers are 8 mesotheliomas, there is -- there are some 9 non-mesothelioma pleural cancers. 10 Q. Would it be fair to say that this study defined 11 it by the site of the cancer; not by the histology? 12 A. That' s true. And it 's a death certificate 13 study. So, of course, that - - it s certainly -- it 14 certainly have limitation of a death certificate study. 15 So observed number of cases is 12, expected 26, based on 16 the general population. That would give the PMR of 17 0.46, and actually that one is significant. It's 18 significantly less than 1. Why? Because it's probably 19 confounded by other exposures. They were unable to take 20 it into consideration. 21 Q. All right. And from that calculation, what did 22 they conclude? 23 A. Well, they they state that notable for an 98 1 absence of significant risk are motor mechanics by whom 2 concerns have been raised because of the presence of 3 asbestos in brake linings. It seems that this exposure 4 has no important effect on mortality from mesothelioma 5 or asbestosis. 6 I have to mention that they did calculate PMR 7 for asbestosis. 8 Q. What was the PMR for asbestosis among this 9 cohort of motor mechanics? 10 A. I'll tell you in a minute. 11 0. 8. 12 Q. Is that higher or lower than expected? 13 A. That's less than 1. 14 Q. And, again, that's the occupation at the time 15 of death. It doesn't tell you what the person did when 16 they were 20 years old in this study? 17 A. That's right. It's problematic. Those kind of 18 studies are problematic. 19 Q. This was a study design that you told the Judge 20 earlier you consider less informative than the case 21 control cohort design? 22 A. Not that I consider. I mean, most people would 23 agree with that. 99 1 Q. What's the next study you considered? 2 A. Hodgson. 3 Q. What kind of study was Hodgson? 4 A. PMR. A follow-up of Coggon. It just came out 5 later and included larger geographical region as well as 6 more years of data. 7 Q. Did they find more or less than expected deaths 8 for the mechanics? 9 A. The thing with Hodgson, of course, is that they 10 don't give numbers. They provide you with a graph, but 11 you can still figure out what the results were. They - 12 to make it easier on -- for the reader, they grouped 13 people into three categories. The category -- and 14 ranked them with respect to PMR because they understand 15 very well, just like we just discussed, that the actual 16 estimate for PMR, it's not necessarily a very meaningful 17 measure. What is more meaningful is where various 18 occupations rank in -- in the scheme of things. 19 So you start with -- they started with plumbers 20 and gas fitters and carpenters. They have a little bit 21 different definitions. And electricians, construction 22 workers, and production figures and so forth. Then they 23 move into what's call group A: security workers, 100 1 protections maintenance, other service occupation, post 2 men, stove keepers. Then they move into group B: sales 3 reps, teachers and lecturers, managers of transport 4 facilities, bar men, and office managers. And then 5 group C with the lowest level of PMR are motor 6 mechanics, coal miners, and farmers. 7 Q. Right. There's a statistical anomaly we may 8 need to clarify. How can it be possible that with 1 9 being the line of delineation between positive and 10 negative that there are only a handful, my word, of 11 occupations that are positive and everything else is 12 negative? Is that because -- does this chart represent 13 an equal distribution of deaths or is that because there 14 are overwhelmingly -- an overwhelmingly disproportionate 15 of deaths among those handful of occupations? 16 A. Yes. Certainly, certainly. A big chunk of 17 mesothelioma conditions, diseases and death from 18 mesothelioma, are attributable to a relative finite 19 number of occupations. I would say so. Doesn't mean 20 that other people are not going to develop it. 21 Q. Right . 22 A. But the bulk will be contributed by these heavy 23 hitters. 101 1 Q. What we saw -- what this also tells us, 2 correct, is that there are deaths among sales reps and 3 managers and teachers and bar men and office managers 4 from this disease? 5 A. Certainly. And then, again, it's a PMR study, 6 so that one needs to take that into consideration. 7 Q. It has its limitations? 8 A. Right . 9 Q. Is the Teschke study the next one you 10 considered or is there one in between that? 11 A. Yes, Teschke is the next study. 12 Q. And Dr. Teschke and Dr. Checkoway, who we 13 discussed a little bit, are where? 14 A. Dr. Teschke is -- I think she's at University 15 of British Columbia now. Maybe she's at University of 16 Washington. 17 Q. How about Dr. Checkoway? Where is he? 18 A. He's definitely at University of Washington. 19 And -- 20 Q. What sort of study design did they set up and 21 what were they looking to test? 22 A. Well, this is a population-based control study 23 of mesothelioma that took place in British Columbia. 102 1 And that's the study that's probably worth spending just 2 a couple more minutes because it's more informative. 3 Q. Why is it more informative? What's especially 4 informative about the Teschke study? 5 A. Well, there's certain things about how they 6 went about doing the study that makes it more informed. 7 Q. All right. What were they setting out to test? 8 A. Again, it's a case -- population case control 9 study so they evaluated relative risks or odds ratios in 10 a variety of occupations. 11 Q. How did they do that? 12 A. The cases came from histologically-confirmed 13 pleural mesothelioma in -- diagnosed in British Columbia 14 for two years, 1990 through '92. Actually, that makes 15 it -- yeah, September through August will be two years. 16 Controls were very different in this study, and 17 that's, you know -- 18 Q. Who were the controls? 19 A. The controls were voters, people selected from 20 voter registries, living in British Columbia. Of 21 course, it's also not perfect because not everybody 22 votes. So it's only if you're a voter, you have a 23 chance of getting a control. But it's certainly a 103 1 different way to look at things compared to, say, 2 McDonald who used cancers from other sites or Woitowitz 3 who used resection and some undefined population control 4 group or death from other causes. So, you know, that's 5 one thing about them that's important. 6 The other thing about them that's important, 7 they actually were able to conduct interviews in person, 8 at least for some people. They were able to conduct 9 interviews with cases and not next of kin, but not for 10 everybody. 11 So now, of course, the problem now arises, what 12 do you do with controls because controls are healthy 13 voters? If you have next-of-kin interviews with cases, 14 or at least some of them -- they are probably a very 15 large proportion -- you can't interview each and every 16 control directly. You will be introducing bias, 17 information bias it's called. 18 So what they did, they matched next-of-kin 19 interviews with cases. Even though control is alive, 20 they will interview his wife. 21 Q. Slow down. Let's explain that. How did they 22 match? 23 A. Yeah . Well, if the cases it's frequency 104 1 match. If they saw that they ended up with 20 percent 2 of interviews, let's say, directly and 80 percent next 3 of kin, they will say it's a problem. So what are we 4 going to do? We'll interview 20 percent of alive 5 controls, next of kin nevertheless. 6 Q. Right . 7 A. And whatever, just to match it up. So that the 8 proportions of people that were interviewed directly 9 among cases and the proportion of people who were 10 interviewed directly among controls was the same. 11 Q. What information did they have pertaining to 12 motor vehicle mechanics? 13 A. They did three types of analysis. 14 Q. Would it be appropriate for you to show maybe 15 two-by-two tables of - 16 A. Sure. 17 Q. Of the numbers that they had? 18 A. Sure. 19 Q. Is this one study where they looked at more 20 than one thing? 21 A. Yes. 22 Q. Within the category? 23 A. Yes. 105 1 So the first cut is the definition of 2 exposures, vehicle mechanics. So the -- 3 Q. The occupation? 4 A. The occupation's vehicle mechanics. This is 5 disease. This is control. This is motor vehicle 6 mechanics. This is not. Ever never. Very simple - 7 very simple crude analysis. 8 Q. "Ever never" meaning they're taking -- 9 A. Mechanics. 10 Q. They're not mechanics ever? 11 A. Ever . 12 Q. Right. 13 A. This particular analysis, this is vehicle 14 mechanics, gives you an odds ratio, 0. 8. It's 15 non-significant. Confidence interval between 0.2 and 16 2. 3. 17 Q. What other type of hypothesis did they test 18 with respect to this type of work? 19 A. Then they said instead of doing all vehicle 20 mechanics, let's evaluate a subset of people who are 21 specifically engaged because they asked two types of 22 questions. They asked what did you do for a living and 23 also they asked what kind of activities you were engaged 106 1 in . 2 So they did brake installation and repair, a 3 category -- same as Spirtas. The numbers, of course, 4 are smaller because not all vehicle mechanics report 5 being engaged in brake installation and repair. 6 And, perhaps, they didn't have job descriptions 7 for everybody. So at these two-by-two tables, the 8 numbers, of course, get smaller, gives you an odds ratio 9 of . 3 with very wide confidence intervals, from 0 10 virtually to 1. 4. 11 Then they said, Well, the problem with this 12 analysis is that we cannot consider other asbestos 13 exposures. 14 Q. Is that another way of saying we haven't 15 considered confounders yet? 16 A. Correct. 17 Q. All right . 18 A. I mean, the confounders really -- I mean, 19 they've considered other confounders. They've 20 considered age, for example. That's controlled for age 21 and other things. But they did not consider other 22 asbestos exposures. 23 So what they said, Let us only take a look at 107 1 where vehicle mechanics and as far as we can tell, 2 vehicle mechanics only, no other exposures, and compare 3 them with people who had no exposures. Basically that's 4 where you compare vehicle mechanics that never worked as 5 insulators or shipyard workers to librarians and 6 lawyers. People -- that would be basically end up with 7 a majority of white-collar occupations. 8 Q. What did they conclude from that? 9 A. Well, first of all, there was only one case, of 10 course. That's -- you know, the funnel gets tighter and 11 tighter once you get to the bottom. 12 Q. Just explain. There was only one case of what? 13 A. A person who was a motor vehicle mechanic 14 without any other -15 Q. Only one case of mesothelioma? 16 A. Correct. 17 Q. In a vehicle mechanic without what? 18 A. Without any other identifiable occupational 19 asbestos exposure. 20 Q. Okay. 21 A. And there were 15 controls. 22 Q. So there were 15 people who -- healthy or 23 non-diseased vehicle mechanics who had no other 108 1 occupat ional exposure? 2 A. Correct. 3 Of course, with one case, you know, that gets 4 you .4, 0.3, from 0 to 3. It's a very wide c onfidence 5 interva l, but the best estimate -- it's impor tant to 6 remembe r that it's not the probability of any results 7 and bet ween is more or less equal. It's not. The best 8 estimat e is still this one. As you move away from the 9 best es timate, the probability of more and mo re extreme 10 results dropping. 11 So this is vehicle mechanic without any other 12 asbesto s exposure as far as the authors can t ell, of 13 course. 14 Q. All right. And what conclusion from this data 15 did the researchers draw? 16 A. Well, actually, this relates to something else. 17 Although they don't show the data, they also did the 18 same thing for brake installation and repair without any 19 exposures. And that's one -- and that's when they 20 say -- 21 Q. I switched on you. There are two conclusions. 22 You want this one or you want the first one? 23 A. They don't show the results, but they just 109 1 simply say at the very end of the paper that the results 2 were the same as for vehicle mechanics altogether, after 3 removal of other exposures. 4 Q. How many different hypotheses did they test? 5 A. Oh, I forgot to mention. They also did the 6 same thing twice for each analysis was done without 7 consideration of latency and with at least 20 years of 8 waiting, meaning if somebody was exposed, say, within 20 9 years, they would not count that person. They will only 10 count people, and appropriately one can do that, who 11 were exposed 20 years or earlier before they were 12 diagnosed with the disease. 13 Q. When they controlled that way, did the numbers 14 change significantly? 15 A. No, not at all. I mean, the result basically 16 stayed the same. 17 Q. It was negative either way? 18 A. Yes. And numbers were very small. Once you 19 cut, it's really -- you know, the cells become very 20 small. 21 Q. Now, relative to those relatively small 22 numbers, Dr. Lemen I think had a slide yesterday and, in 23 fact, in his chapter in the book references a statement 110 1 that the authors made that was presented as a cautionary 2 statement. Do you recall that? 3 A. Yes, I do. 4 Q. What he says -- what he said -- what he says in 5 the book was that Teschke, et al did not find an excess 6 of mesothelioma among vehicle mechanics, but because 7 their findings were based on small numbers, judgments 8 about any causal associations would be speculative 9 according to the authors. 10 Now, first of all, is that precisely what they 11 said? I don't want to nitpick, but causal associations, 12 is that the word -- is that the term they used in the 13 paper? 14 A. You know, it's easier. You want to just put it 15 up, what they said? 16 Q. Let's show what they actually said. Let me 17 have that. The last thing they said about the paper. 18 A. Several occupations and exposure scenarios were 19 identified that had not previously been linked to 20 mesothelioma. 21 And, indeed -- 22 Q. Yeah, stop there. What are they talking about 23 there? 111 1 A. They're talking about findings that surprised 2 them because they were positive. For example, they had 3 gardeners that had increased risk. 4 Q. But they reported -- although they didn't 5 conclude anything, they reported an increased risk about 6 gardeners. Is that right? 7 A. Yeah. There were some other -- 8 Q. That's all right. That's one example. 9 A. But gardeners is a good example because they 10 found an increased risk among gardeners. 11 Q. And when they said they found an increased 12 risk, that's finding an association. If there's no 13 increased risk, is that an association or a 14 non-association? 15 A. Yeah. An association means that you had a 16 departure from normal. You had a departure from 1. 17 That is significant, of course. I mean, all of them 18 would have some departure, but the point is if there is 19 a large departure that seems to be significant, you 20 start thinking. 21 So they said: Several occupation exposure 22 scenarios were identified that had not previously been 23 linked to mesothelioma. All except nonasbestos mining 112 1 involved small numbers. 2 So it seems like they also found an association 3 for nonasbestos mining and that association was 4 significant. 5 Therefore, judgments about whether these 6 associations were causal would be speculative. 7 Q. All right. But let me ask you a very simple 8 question. Was this cautionary statement made about all 9 of the findings or was it a specific statement about 10 taking caution not to leap to conclusions about some of 11 these new, quote, unquote, associations they technically 12 reported? 13 A. Yes. It's clearly -- they clearly talk about 14 associations that they didn't expect to see, and 15 sometimes you get those when you have multiple - 16 multiple analysis in one study. 17 MR. TARRY: Your Honor, we've got probably five 18 or six more of these. Do you want to break now or keep 19 pushing? 20 THE COURT: My conference this afternoon is at 21 2. 22 MR. TARRY: Would it make sense for us to take 23 a lunch break at 2 if people can survive? Is that 113 1 amenable ? 2 THE COURT: Let me just double-check with staff 3 and see if we can get some folks rotating through. But 4 for now, why don't we take a recess for 10 minutes, and 5 we'll see where we're going for the balance of the day. 6 (Recess taken.) 7 BY MR. TARRY: 8 Q. Dr. Goodman, what was the next study you 9 considered after Teschke? 10 A. Next one was Agudo. 11 Q. Tell us very quickly, if you could, who 12 Dr. Agudo is and what sort of studies he did. 13 A. This was a study done in two areas of Spain, 14 one in up north in Barcelona, and the other one south in 15 Cadiz. And they identified mesotheliomas from 16 hospitals. It wasn't a population-based, a 17 registry-based study. And they studied hospital 18 controls but in a sort of innovative way. 19 Q. How's that? 20 A. The way they did it is they selected a sample 21 of general population first as control, randomly. And 22 they said for each control selected in the hospital, in 23 the general population, they will match a control in the 114 1 hospital. It sort of halfway between true population 2 control and a true hospital control. 3 In other words, they say, okay, I've 4 identified, let's say, 50 people in the general 5 population, and they have certain characteristics; age, 6 sex, whatever. But I'm not going to interview them. 7 I'm going to interview exactly the same person at the 8 hospital. I've never seen control selection done quite 9 that way. They actually published a paper just on that 10 particular issue. Seems to be working for me. It's an 11 interesting way. Basically, they're bridging between 12 the hospital control and population control. 13 Q. What did they find with respect to the 14 hypothesis? 15 A. You have to really dig through the paper before 16 you found the results pertinent to this question. But 17 they do have results for mechanics, motor vehicles, as 18 they defined that category in Spain. There were three 19 cases and 14 controls. And the reference category was 20 formed -- by reference category here they say unexposed 21 people -- were the same 51 cases and 148 controls who 22 never worked in any of what they called listed 23 occupations. Listed occupations are the ones they 115 1 thought a priori. Would be occupations that may pose a 2 risk. And that a priori including mechanics. 3 Q. Are those numbers presented there on the 4 screen? 5 A. Yes. If you put in the numbers in a table, 6 again, 314, 51, and 148, you would get a relative risk, 7 an observation of 0.62. And again, confidence 8 intervals, .1 and 2.3. 9 Q. A handful of other studies. Is there a study 10 called Milham and Ossiander from 2001? 11 A. Yes. 12 Q. Did you consider that study? 13 A. Yes. 14 Q. What about it did you consider? 15 A. It's a PMR study done in the state of 16 Washington. One interesting thing about it, the study 17 collected cases in 1950. So now they have 50 years 18 worth of deaths, I mean, thousands and thousands of 19 deaths. And they do PMR analysis. Again, it's 20 imperfect. But an interesting thing is that it expands 21 such a huge interval of time. 22 Q. What information, if any, pertained to auto 23 mechanics or car repair? 116 1 A. They have a category called automobile 2 mechanics and repair workers. And they had seven 3 observed and nine expected cases, nine expected based on 4 general population, seven observed in that particular 5 category. And that's a PMR of .75. 6 Q. Now, the Court has heard about NIOSH. Does 7 NIOSH also collect mortality data? Did they? 8 A. Yes. What NIOSH has is what they called 9 occupational mortality surveillance database. And it's 10 unlike Milham and Ossiander -- or previously we 11 discussed Petersen and Milham -- these are state-wide 12 registries; NIOSH selects 28 states. 13 Q. 28 states? 14 A. 28 states throughout the count ry. And 15 basically the same analysis. Few years , of course. 16 They only have -- let's see - - 14 years worth of data 17 1984 through 1998. 18 Q. The NIOSH data, the U. S. federal government 19 looked at how many states for how many years? 20 A. 28 states, 14 years. 21 Q. And what data do they present? 22 A. Same thing. They have -- the total number of 23 observed mesotheliomas is 15, expected 19. And, you 117 1 know, this is a nonsignificant PMR of .8. 2 Q. Continuing along with other studies on 3 mesothelioma, did you consider a study done by Hansen 4 and Miersen in 2003? 5 A. Yes. 6 Q. What is the Hansen and Miersen study of 2003? 7 A. Hansen and Miersen study is a case controlled 8 study which is a follow-up of another Danish study 9 called Olsen and Jensen, not to confuse with a Hansen 10 study that's a different study. 11 Q. All right. Yesterday Dr. Lemen and I had a 12 discussion that might have been muddled on the record 13 about Hansen 1989. That's a different study than this; 14 is that correct? 15 A. A different study, different Hansen. 16 Q. Right. Now, Olsen and Jensen, I think you 17 testified used some next of kin interviews that Hansen 18 ' 89 relied on; is that correct or not? 19 A. No. 20 Q. All right. Tell us about the Hansen and 21 Miersen study of 2003. 22 A. Well, let's begin with Olsen and Hansen. It's 23 easier to do that. Olsen and Jensen is a PIR, instead 118 1 of PMR, PIR study. That was published in '87. And what 2 they have is this: They have a cancer registry data and 3 -- so it's not a death certificate any more. It's 4 actually a newly diagnosed case that's reported to the 5 registry. Person can be alive. 6 So -- and then what they do, they match it, the 7 link. You see, in Denmark, unlike in the United States, 8 everybody is accounted for. People don't really move. 9 It's about six million population country. And 10 everybody has records. So what they do they would link 11 incidence data, say a case was diagnosed with a certain 12 cancer, and they link those data with what's called 13 Denmark supplemental pension fund and the central 14 population registry. 15 So, you know, if I'm diagnosed with lung cancer 16 tomorrow, they will be able to tell what kind of - 17 because I pay taxes, so forth -- they can tell me what 18 occupations I've held. 19 Q. Just in it interest of time, we'll submit these 20 studies for the Court's consideration. But what was the 21 conclusion? 22 A. The point is Olsen and Jensen found no cases in 23 '87. They kept collecting the data. And just two years 119 1 ago, 2003, they published in Danish this Hansen and 2 Miersen paper. They had ten cases by then. So that it 3 now went for 27 years, from 1978 to 1997, the data set. 4 So unlike '87, ten years later they had ten cases. 5 And they observed number was ten, expected was 6 12. And that's a PIR, relative risk of point A. 7 Q. Just to wrap this up on mesothelioma, was there 8 two more studies published in this year, 2005, that you 9 considered? 10 A. Yes. Actually, there was one more study 11 published in 2004. That would be Hessel. And two more 12 in 2005. 13 Q. Let's come back to Hessel, because there's 14 another issue I want to discuss in a little bit more 15 detail. 16 A. Okay. 17 Q. Tell us about the McElvenny study that was 18 published earlier - 19 A. Yeah. The two most recent studies, a 20 McElvenny, out of England -- Great Britain, rather - 21 that -- we discussed Coggin followed up by Hodgson. 22 This is a third follow-up. 23 Q. The same data set 120 1 A. PMR, British. Results are identical .48. A 2 lot more cases. 60 cases now. 3 Q. Now, what about Welch? What is Welch? 4 A. Welch is a study of peritoneal mesothelioma, 5 specifically. It's a hospital-based study. 6 Q. Let's make that distinction. Peritoneal 7 mesothelioma was discussed earlier before the Court 8 But that s a different site in the pleura? 9 A. Yes . 10 Q. Is peritoneal more rare or more common 11 mesothelioma? 12 A. It's more rare, in general. 13 Q. Yeah. Very quickly, what did Dr. Welch attempt 14 to do? 15 A. Dr. Welch basically borrowed the Spirtas 16 questionnaire 17 Q. I am not sure she's a doctor. 18 A. Neither am I. 19 Q. What did Welch do? 20 A. Took the Spirtas questionnaire, the one that 21 was used by the NCI study and applied to a 22 hospital based case controlled study of mesothelioma -23 and as controls they use another rare tumor, the cancer 121 1 of the appendix -- and applied the same questionnaire. 2 The problem is, just like Spirtas '94, there 3 are no numbers specifically, you know, for each 4 category. They do have -- they do report numbers of 5 people, but they don't tell us who was doing what in 6 terms of occupational histories. 7 But if you plug those numbers, the number of 8 exposed mesothelioma cases would be eight. The 9 number -- I'm sorry. The category is exactly the same, 10 brake installation and repair; eight exposed cases, six 11 exposed controls, 16 nonexposed cases, 18 controls. 12 It's a total of 24 cases, 24 controls. 13 They give you an odds ratio of 1.5. 14 Q. Did the authors calculate that or did you 15 calculate it from the data presented? 16 A. Yeah. They did not report that number. 17 Q. Even though they didn't report it, you went 18 ahead and calculated it. Am I correct that you have 19 included it on the compendium slide that is now up on 20 the screen? 21 A. Right . 22 Q. Let's summarize the results of the studies you 23 considered mesothelioma in auto mechanics. Does this 122 1 slide present all of those studies plotted on the graph 2 you explained earlier? 3 A. Yes. These are -- I think that's 15 out of 18 4 studies. 5 Q. I'm going to ask you about Hessel in just a 6 moment when we talk about lung cancer and on one other 7 issue. But what is your conclusion based on this 8 depiction of the body of evidence from epidemiology? 9 A. Well, if one looks at this set of results, 10 there is no evidence whatsoever to conclude that risk of 11 mesothelioma in people engaged in -- this is a broadest 12 category -- in people defined as garage mechanics, motor 13 vehicle mechanics, or brake installation repair workers 14 is elevated. It's indistinguishable from one. 15 Q. Now, Dr. Frank said yesterday in response to a 16 question about cohort study design -- let me back up. 17 The purple are case control studies, and the green are 18 PMR studies on this chart? 19 A. And there are two studies that didn't have an 20 estimate because they have no cases . 21 Q. You have not identified any cohort studies. In 22 fact, Dr. Frank yesterday, in response to a question, 23 said there were no cohort studies of garage mechanics or 123 1 brake mechanics. Is that true? 2 A. No, that's not true. 3 Q. But yet you don't have any cohort studies up 4 here. Why is that? 5 A. Because none of them were able to calculate 6 relative risk or SMR. However, they provide data that 7 are very much in keeping with this model. 8 Q. In particular, let me ask you about a study 9 called Gustavsson. 10 A. Right. 11 Q. Is Gustavsson a cohort study? 12 A. Yes, it is. 13 Q. And what did Gustavsson do in their study and 14 what findings did - 15 A. Gustavsson is a Swedish done in Stockholm 16 garages. They primarily focussed on lung cancer. And, 17 you know, lung cancer is -- there is really a lot of 18 information that can be discussed on lung cancer. They 19 did say they had two mesotheliomas in their workforce. 20 Although, it's just in the discussion phase. But that's 21 an interesting thing. 22 One mesothelioma case was a garage worker, but 23 he was an electrician. And electricians clearly have an 124 1 elevated risk of -- he was never a -- fixing vehicles. 2 He was an electrician. About the second one, they don't 3 say what he was doing, so we don't know. We also know 4 that both of them had exposures or may have had 5 exposures, to precisely to quote Gustavsson, elsewhere 6 besides the garage, so -7 Q. A simple question: Are the cohort studies that 8 have been done, the moving going forward studies on car 9 repair, garage mechanics, however defined, are those 10 studies consistent with an association between exposure 11 to friction fibers and mesothelioma? 12 A. They're consisted with no association. 13 You know, one more thing. And another thing 14 that is really important, I think, is Jarvholm. 15 Literally it's important because -- 16 Q. Jarvhom, J-A - 17 A. J-A-R-V-H-O-L-M. 1988. And that study, the 18 reason why it's so important because it includes almost 19 20,000 motor mechanics as recorded in the Swedish 20 census. 21 Q. How many years did they follow? 22 A. 20 years of follow-up. 23 Q. 20,000 mechanics followed for 20 years. And 125 1 what did they find? 2 A. That gives you roughly 400,000 person years of 3 follow-up. They found one case of mesothelioma about 4 which they say that this case was a construction worker 5 Q. Who also worked in a garage? 6 A. Who also was at some point during a census a 7 car mechanic. 8 Q. Just for the sake of comparison so the Court 9 can understand, 20,000 persons followed for 20 years 10 resulted in one case? 11 A. Didn't result in one. It was one case. 12 Q. It was the result of an observation of one 13 case . How does that compare to similar work, similar 14 size cohorts like what Dr. Selikoff did in the 1960s? 15 A. Yes. They Selikoff cohort is very similar 16 size. I think it's just under 18,000. 17 Q. Insulators? 18 A. Insulators. And followed for around 20 years. 19 And there was 180 cases. 20 Q. Dr. Selikoff in a similar cohort found how many 21 cases of mesothelioma? 22 A. 180. 23 Q. You are a coauthor of the study listed in the 126 1 middle of the upper part of the chart, Hessel, 2004. 2 You've also told the Court about that? 3 A. Yes. 4 Q. What did that add to the literature? 5 A. Well, the Hessel is additional analysis of the 6 NCI data set. 7 Q. Why did you do additional analyses? 8 A. When we were reviewing the literature, a couple 9 of questions remained unanswered. And we were hoping to 10 answer those by analyzing the National Council Institute 11 data set. That's why we contacted Spirtas. And he just 12 sent it to us. 13 The first question was, if you recall, we have 14 brake installation and repair as an exposure definition, 15 which is a valuable definition of exposure. But all we 16 had was this very crude estimate from Spirtas '85, which 17 was 1.0 with a confidence interval between 0.6 and 1.6. 18 Then you had this '94 pretty long study, pretty 19 extensive study, but it did not look at mechanics. And 20 it also told us that 33 percent of motor vehicle 21 mechanics were engaged in ship building or ship repair, 22 shipyard work, and 55 percent were engaged in 23 installation work. 127 1 So we felt that that might be an opportunity to 2 do additional analysis. One is to look at whether or 3 not one can evaluate risk associated with brake lining 4 installation and repair and adjust for other exposures. 5 That's typically not done. People just split it up and 6 say you compare vehicle mechanics only to people who 7 were not engaged in any asbestos-related jobs. 8 So the other way is to do it in model in what's 9 called logistic regression model where you adjust 10 analytically for other exposures. 11 Then we also did the traditional way, where you 12 just eliminate people with mixed exposures and look at 13 people who were brake installation and repair workers 14 only, compare them who had no known history of 15 occupational exposure. 16 The third analysis that was never done before 17 is to look at duration of job. Because most studies - 18 the only exception is Gustavsson -- do not have direct 19 industrial hygiene measures. Duration of exposure is a 20 surrogate, I mean, albeit imperfect, but a surrogate for 21 exposures level for dose response. You would expect, 22 say, that people who work more than ten years have a 23 statistically significant difference in terms of risk 128 1 compared with people who work less than ten years. Or 2 you have more cases. You can do five, ten, or 25 years, 3 for instance. 4 Q. Dr. Goodman, we'll submit this study to the 5 Court, also. But quickly, did the data indicate that 6 the people who engaged in this work longer had a higher 7 risk of disease? 8 A. No. 9 Q. All right. Did the data indicate a dose 10 responsive relationship at all? 11 A. That's one way to look at dose responsive. And 12 there was no dose responsive. It stayed flat, 13 regardless of whether people were less than ten years or 14 more than ten years. 15 Q. On this hypothesis that was expressed regarding 16 possibly synergy between chrysotile and other types of 17 fibers, did the data indicate that there's a synergy 18 between working in a garage and having other 19 occupational exposures? 20 A. No. They did not indicate. 21 But the important thing that -- we actually 22 tested the hypothesis of synergy by using what's called 23 interaction term. An interaction term would look at 129 1 synergy or departure from a regular logistic regression 2 model in -- between other exposures and the brake 3 mechanic. And none of those interaction terms -- I 4 don't want to overload you with technical -- were 5 significant. Basically, they did not -- were not 6 present in the model. 7 Q. Now, you also performed a systematic review of 8 the occupational epidemiology; correct? 9 A. Yes, I have. 10 Q. That's the meta analysis paper that we've 11 discussed? 12 A. Yes. 13 Q. I'm going to ask you a couple questions about 14 that. But do we have a slide that's -- we'll go through 15 this very quickly. This is a slide that you compiled 16 that shows where the research was done and how it was 17 funded; correct? 18 A. Yes. 19 Q. It's actually a second page to it. We'll just 20 go through that. Go back. Up one to Teschke. I think 21 the Court has already seen both those slides. Go down 22 to -- right there. 23 We're talking about your meta analysis. Was it 130 1 just a calculation or were you doing two different 2 things? 3 A. Yes. The meta -- we set out to do the meta 4 analysis to do two things. One is just a qualitative 5 evaluation and, you know, critique of each study 6 individually in a systematic way. 7 Q. What is the process for doing that? 8 A. Yeah. For meta analysis or in general? 9 Q. For a systematic review. What's the process 10 for conducting a systematic review? Because that might 11 mean something different to me than to an 12 epidemiologist. 13 A. Yeah. It's very important. A systematic 14 review, you begin like any study. You begin with a 15 hypothesis. You state your hypothesis. In our case the 16 hypothesis was just like the hypothesis set out to 17 tested by any of the individual studies. What is the 18 evidence of the association between mesothelioma and 19 motor vehicle repair using various definitions using a 20 epidemiological data, data that's either reports, 21 relative risk estimates, or allow you to calculate 22 relative risk based on data included by authors? 23 So once you state the hypothesis, then the 131 1 first thing you do, you identify all relevant study. 2 The two key words, "all" and "relevant." You know, you 3 don't want to miss anything, but you also don't want to 4 include studies that are irrelevant to the hypothesis, 5 to the specific hypothesis you are testing. 6 After you've done that, you go through each 7 study one by one. And there is multiple ways to do it. 8 Laden, for example, just described them. 9 Q. Laden is at Harvard. That's a study that was 10 also submitted? 11 A. Right. Laden just described them. And it's a 12 perfectly reasonable way to do it just one by one in a 13 sort of systematic academic fashion. Many of you put 14 together tables and show what individual studies showed. 15 We decided, because studies are not created 16 equal -- many of them have strengths and many of them 17 have weakness and all of them have different strengths 18 and weaknesses -- we decided that if we set certain 19 number of criteria -- they have well-accepted criteria 20 for study quality -- and score the studies so that we 21 can understand the relative trustworthiness -- that's a 22 good word -- trustworthiness of the result, then you can 23 do -- take it to the next step and take the studies and 132 1 combine them in what's called a pooled analysis or meta 2 analysis. You take several risk estimates. And there's 3 a statistical significance that allows you to say, 4 what's the best average estimate for these studies -- 5 Q. Even though there's a heterogeneity? 6 A. Well, if there was a heterogeneity, I wouldn't 7 have combined them. Because you can't combine 8 heterogenic studies. Your job is to go find out why the 9 heterogeneity occurs. But if you do that in tiers, 10 there was no heterogeneity. Studies that were, say, 11 tier one were very similar to each other. And studies 12 in tier two were somewhat similar although not exactly 13 the same as tier one. And tier three we did not even 14 analyze -- 15 Q. In fact, is there a formula for heterogeneity? 16 A. Of course , yeah. There's a test for 17 heterogeneity. 18 Q. Did you p resent that in the paper? 19 A. Yes. One of the tables shows heterogeneity. 20 Heterogeneity is a lay term, but it's also a 21 mathematical term. You can test whether or not there is 22 a disagreement. 23 Q. Dr. Goodman, in the interest of time I want to 133 1 move into another topic. Before I do that, tell the 2 Court what's important about -- in your opinion, about 3 the result of the meta analysis. 4 A. There are three rows. It's not necessarily the 5 three tiers that we use, but the three informative rows. 6 The first row are what we called tier one studies. And 7 we do not insist that that's necessarily the best set of 8 studies. This is the set of the studies that happen to 9 be tier one compared to every other study. I mean, 10 there may be other ways to do it. But these are studies 11 that are, in our opinion, and based on our scoring, were 12 the most accurate. 13 Then the second row, the green color, are 14 studies that we're able to eliminate other exposures. 15 In other words, look at -- compare people who were 16 engaged in brake repair to those -- again, with a 17 caveat -- to the best of their knowledge that did not 18 have any exposure. So essentially comparing motor 19 vehicle mechanics to essentially white collar workers or 20 blue collar workers that are engaged in things that are 21 not related to asbestos. 22 And the third tier -- it's not a tier in terms 23 of quality, but it's a way to look at this, third 134 1 hypothesis, is studies that define exposures of brake 2 workers as opposed to a garage work or motor vehicle 3 repairs. 4 And the bottom of each of these colors gives a 5 meta relative risk, meta RR. That meta RR for tier one 6 is -- you see what happens with meta analysis; it takes 7 care of the frequency of sample size issue. We have 8 small studies with very wide confidence intervals. The 9 estimates are very imprecise. So a meta analysis allows 10 you to combine them as if it's one study. So the 11 confidence interval becomes much more precise. 12 So for tier one you get a one, basically, for 13 studies to look -- eliminated other exposures and 14 compared mechanics without other known asbestos 15 exposures to people without any asbestos exposure is 16 basically one. If you do it for brake work, if you 17 define your exposure as brake work as opposed to motor 18 vehicle work, you get basically one. 19 Q. Did you hear the testimony from some of the 20 previous experts that in their opinions on, again, all 21 of the things being equal, just on a volume dose basis, 22 it takes the most exposure to get asbestosis, less for 23 lung cancer, and even less for mesothelioma? Did you 135 1 hear that? 2 A. Yes, I heard that. 3 Q. Is that something you agree with or disagree 4 with? 5 A. Well, it's a very general statement. But 6 overall there is a hierarchy, I think, that mesothelioma 7 would be, as many of the experts pointed out, would be a 8 signal tumor if there is an increase incidence. 9 Q. If I understand right, of the three diseases 10 we're talking about, mesothelioma, in your opinion, like 11 you heard, is the one of the three that requires on a 12 volume of dose basis the least? 13 A. It's the most sensitive end point. That's the 14 one that you will see a red flag earlier than others. 15 Q. Stated another way, you would expect that lung 16 cancer, you require a larger dose of whichever kind of 17 fibers to result in lung cancer; is that correct? 18 A. Yes. I mean, the reason I'm saying is that is 19 just because there are studies that show just that. 20 Q. I hope I phrased that as a hypothesis and 21 opinion. I know you're not establishing that. 22 With respect to lung cancer, let me ask you, 23 how does the lung cancer epidemiology differ from the 136 1 epidemiology of mesothelioma? 2 A. Well, they're different -- it's very different 3 challenges. When you do studies of lung cancer, the 4 challenges that you're facing are very different. The 5 problem with mesothelioma is that it's rare and, 6 therefore, even large cohort studies, unless people are 7 insulators, will give you no cases, so you cannot really 8 evaluate the risk. All you can say whether it's rare or 9 common. 10 That's why the mesothelioma just like, you 11 know, other rare cancers suits most -- lends itself best 12 to case control design. Because you can collect cases 13 from multiple sources and take controls and analyze 14 that, particularly -- case control studies are 15 particularly useful if you have rare disease and common 16 exposure. And being a brake mechanic is a common 17 exposure. Mesothelioma is a rare disease. 18 The reverse is true for cohort studies. If you 19 have common disease but rare exposure, you must pare 20 off, rounding out that exposed people even though you 21 have maybe several hundred, follow them over time. But 22 with a common disease you can end up seeing quite a few. 23 Prostate cancer lends itself perfectly to a cohort study 137 1 or even lung cancer. 2 Q. How many causes potentially are there of lung 3 cancer that are under study? 4 A. Yeah. So challenges of lung cancer and 5 mesothelioma, as I said, are different. 6 Now, nonconfounders. For mesothelioma age is 7 important confounder, like any other cancers. The other 8 confounder is other exposure. That's more or less it. 9 You know, you don't have to worry a whole lot about 10 other things because it's so specific. 11 With lung cancer there are multiple causes and, 12 therefore, there are multiple confounders. 13 Q. Your meta analysis was not just for 14 mesothelioma, but you also did a systematic review of 15 the literature for lung cancer, also? 16 A. Yes. 17 Q. We'll discuss that in a moment. 18 Do you agree that the first and foremost risk 19 factor for lung cancer is smoking? 20 A. Yes. 21 Q. With respect to asbestos, do you believe that 22 asbestos is a risk factor for lung cancer? 23 A. Absolutely. 138 1 Q. Is the risk smaller or larger than for 2 mesothelioma ? 3 A. Smaller. It's a general statement. But, 4 generally speaking, smaller. 5 Q. Should smoking be taken into account if you're 6 going to study the potential relationship or association 7 between auto work and lung cancer? 8 A. Yes. 9 Q. And, succinctly, why is that? 10 A. Because automobile mechanics in studies have 11 been shown to be an occupational group that ranks among 12 the highest with respect to smoking habits, both with 13 respect to intensity -- how many cigarettes per day - 14 as well as frequency of habit in terms of prevalence, 15 percent of people who smoke versus those who don't. 16 I think the previous slide ranked them. The Y 17 axis is the cigarettes per day. So mechanics on 18 average, averaging both smokers and nonsmokers, smoke 19 about eight cigarettes per day. The other people who 20 beat them to it are supervisors and production 21 occupations. And then -- next slide. And here it's 22 prevalent, basically, the percent of people that smoke. 23 And there they rank number four. 139 1 Q. Now, do you have a graph that shows studies of 2 lung cancer where smoking was controlled for - 3 A. Yes. You have to control for smoking. And you 4 have to control for smoking and control for smoking 5 adequately. If it's the ever never category, you just 6 use pack years; that won't do. So these are the studies 7 that adequately control for smoking. 8 Q. We won't go through each one of these studies 9 like we did for mesothelioma. We'll have those 10 submitted on the record, also. But what is your 11 conclusion from that presentation of the literature? 12 A. Well, the smoking adjusted relative risk for 13 lung cancer among motor vehicle mechanics are no 14 different than one as a group. And this is old studies. 15 I think there are about 30 of them. 16 Q. Let me ask you about one study. Those are all 17 the studies whether they controlled for smoking or not? 18 A. Only those on the first slide did a good job of 19 doing that. 20 Q. These are all the studies? 21 A. All, each and every. 22 Q. I'm going to ask you about one study in 23 particular. But just with respect to all of the studies 140 1 taken together, is that a consistent or an inconsistent 2 finding to you? 3 A. Yes. This is one. 4 Q. Okay. Now, Gustavsson we talked earlier. Is 5 Gustavsson also concerned with a hypothesis about lung 6 cancer? 7 A. Gustavsson, yeah, it is. It's a cohort study 8 that we discussed a little bit earlier. 9 Q. That followed all of those auto mechanics for 10 all these - 11 A. These are not auto mechanics. These are people 12 who worked in municipal garages in Stockholm. 13 Q. So all sorts of people? 14 A. Yes. 15 Q. What was particularly important about the 16 Gustavsson - 17 A. Well, they set up industrial hygiene 18 measurements. They actually set out pumps. And they 19 would calculate concentrations of fibers per CC and then 20 convert it into dose . 21 Q. So there was a question from the Court either 22 yesterday or the day before about how you account for 23 exposure. Am I correct that Gustavsson actually had 141 1 measurements of whatever kind of fibers it was they were 2 being exposed to during the study? 3 A. Yes, they had measurements. But of the most 4 important thing is not only did they measure, but they 5 actually analyzed by exposure level, just for lung 6 cancer Because, of course, with mesothelioma they 7 couldn' t do much. 8 Q. Is that important because you could test the 9 hypothesis whether more exposure in terms of volume 10 equals more risk of disease? 11 A. That's right. They tested the dose responsive. 12 Q. What did they conclude on that? 13 A. Well, they conducted the analysis... 14 (Pause 15 BY MR. TARRY: 16 Q. Did they find people with more exposure were at 17 increased risk of lung cancer than people with less 18 exposure ? 19 A. Yes. Let me give you the actual numbers. It 20 will be easier. That way I just need a second because 21 it's... I need to find it. 22 (Pause.) 23 THE WITNESS: Here we go. 142 1 They conducted what's called a nested case 2 control analysis within a cohort. That's probably the 3 best of all if you can do it. You have a cohort. You 4 collect data prospectively. You follow people all of 5 the time. But then your analysis is a case control 6 study within a cohort. So everybody is counted for you. 7 You don't just run around sampling or recruiting people. 8 Nested case control study, if possible, is one of the 9 best designs. And you control people within the same 10 occupational group. So chances are that they're more 11 homogenous with respect to social economic status, 12 lifestyle habits, and whatnot. 13 Q. Okay. 14 A. So here we have -- they calculate what's called 15 an asbestos index, which is a measure based on both 16 duration of employment and levels that they had at the 17 garage, the actual fiber levels. That asbestos indices 18 basically provide them with four categories; zero to 20, 19 20 to 40, 40 to 60, and 60 plus, more than 60. That 20 would be the highest. 21 The lowest categories, zero to 20, would be 22 your unexposed in the case control study. That's 23 relative risk of one by definition. Because it's a 143 1 reference group. Then if you go up the level -- can I 2 just draw it? 3 THE COURT: Yes. 4 THE WITNESS: So zero to 20 is a comparison. 5 And that's -- always will be one. So everybody is 6 compared to them. Then the next category is 20 to 40, 7 then 40 to 60, and then 60 plus. 8 So 20 to 40 had a relative risk of 1. 6, which 9 was not significant, 1. 67 actually, which was not 10 significant, from 0.5 to five. Then as exposure 11 increased, the risk decreased to 1.26. And, again, it 12 was not significant. And then as exposure increased 13 further, the risk decreased further to 1. 2. 14 BY MR. TARRY: 15 Q. Dr. Goodman, yesterday Dr. Frank was asked a 16 question at the very end I want to ask you about. He 17 drew a line like this (indicating), talking about dose 18 response. He was asked, do you mean assume, Dr. Frank, 19 a linear non-threshold dose responsive curve for these 20 diseases? And I believe with the exception of 21 asbestosis, he answered yes to the first two. 22 Is this consistent -- is this data consistent 23 with a hypothesis of a linear no threshold dose 144 1 responsive curve? 2 A. For this particular type of exposure, the data 3 are most consistent with an average of a flat line. 4 Because it's -- you would expect some fluctuations, but 5 overall you have no increase with increasing dose. 6 Q. So if there's a theory that friction fibers 7 ought to contribute to disease in linear no threshold 8 manner, would the Gustavsson report study support that 9 hypothesis? 10 A. No. 11 Q. All right. Your conclusion on lung canc er is 12 that auto me chanics are not at an increased risk? 13 A. Yes 14 Q. All right. And is that just your conclu sion or 15 have others who did systematic reviews of the lit erature 16 come to the same conclusion? 17 A. Yes 18 Q. Dr. Laden we talked about earlier. We'l l 19 submit that paper. Go ahead to the next one. 20 Is it surprising to you, given that the data 21 don't suppor t an increased risk for mesothelioma, that 22 there would not be one for lung cancer? 23 A. No . 145 1 Q. Why? 2 A. Well, these are the studies selected using two 3 important a priori criteria. One, is that these are all 4 cohort studies. So these are the studies that reported 5 and calculated relative risk for both lung cancer and 6 mesothelioma. 7 And the studies for which reported unequi vocal 8 increase in lung cancer, you know, the relat ive ri sk 9 of -- all of those are significant. All of them a re 10 ones you can 't miss, you know, 2 .8, roughly three -- and 11 one even has a relative risk of 17. That's the as bestos 12 sprayers in Finland. And then side by side is the 13 analogous result for mesothelioma. 14 And even though lung cancer, of course, is 15 elevated in each those studies, this is a very different 16 circumstances of exposure. Lung cancer, of course, is 17 elevated in all of these studies. Mesothelioma is in 18 double digits -- actually, in triple digits in one 19 particular case. The relative risk for mesothelioma for 20 the Finnish studies is 263. 21 Q. The Court was told that smoking and asbestos 22 exposure, just mentioned generically, are either 23 additive or synergistic or multiplicative for lung 146 1 cancer. Do you agree with? 2 A. There are circumstances when smoking and 3 asbestos clearly act synergistically. And there are 4 circumstances where they don't. And there's a hole 5 bunch of in between. 6 Q. Let me ask you very briefly to come down, and 7 using the basic formula both for additive hypothesis and 8 assuming a multiplicative hypothesis, show what the 9 results of the epidemiology would be for mechanics, auto 10 mechanics. 11 Additive. 12 A. Yes. 13 Q. If it's additive, how do you add smoking and 14 whatever contribution from friction products? 15 A. There are two ways of looking at how the two 16 factors, both of which are risk factors, interact. And 17 with respect to asbestosis and smoking, probably the 18 very first paper that came out on this was conducted 19 within the group of insulators. And first off was 20 Hammond, the American Cancer Society head of research 21 back then. 22 And what he reported -- I think it's in ' 79 - 23 is that asbestos alone in the absence of smoking, people 147 1 who they thought were not smokers -- it's a bit 2 unclear -- gave a relative risk of five. Smoking alone 3 in people that, according to author's information had no 4 asbestos exposure, was about ten. But people who had 5 exposure to both -- no room here -- but it goes high. 6 It goes to 50. 7 The point is -- and that's the model that this 8 data fits the best, meaning if I'm already five fold 9 increase in risk, my risk went up five fold, and on top 10 of it I start smoking, then it's ten times more. 11 An alternative model where data sometimes fit 12 and actually in cases of asbestos exposed workers fits 13 as well is so-called additive model. That would be 14 five, and ten, and 14. The reason how math works is, 15 for multiplicative, not 15, but 14. 16 Q. Ten plus five is 14 in the additive model? 17 A. Well, that's how it works. If the model is 18 multiplicative, then the two factors, say, factor A and 19 B together, relative risk -- this is relative risk for 20 factors A and B together -- compared to people who are 21 exposed to neither, is relative risk of A times relative 22 risk of B. That's easy. 23 Now, for additive molds, relative risk of A 148 1 plus B equals relative risk of A minus one. Because 2 relative risk minus one is excess risk, above and beyond 3 what should have happened anyway, plus relative risk of 4 B minus one, plus one. Because there is a background. 5 So that if it's five and 15, that's four -- five and 6 ten, it's four and nine and one, it's 14. 7 Q. Very quickly, though, for auto mechanics, based 8 on epidemiology, what is the contribution of smoking to 9 risk on the above -- 10 A. It's ten. Smoking gives you a relative risk of 11 ten. 12 Q. Smoking gives you a relative risk of ten. And 13 you add what contribution for working in a garage? 14 A. You add a one. 15 Q. Or you subtract one -16 A. Well, if it's additive, then you have a 17 relative risk ten minus one plus one minus one plus one 18 equals ten. If it's multiplicative, regardless of what 19 model you pick, ten times one equals ten. In other 20 words, if one of the components does not increase risk, 21 unless the data indicates otherwise -- you have to 22 always say that -- unless the data indicates otherwise, 23 one cannot conclude there is a synergy. 149 1 Q. Let me ask it in a simple non-mathematical way. 2 If a smoker goes to work in a garage, does the 3 epidemiology suggest that the garage work adds to that 4 person's risk of lung cancer? 5 A. No. We have studies that actually looked at 6 that issue by separating smokers and nonsmokers and 7 looking at that risk separately, the study by Blair; and 8 there is a study by, Hrubec, H-R-U-B-E-C. Both of them 9 found no increase in risk whether a person is smoker or 10 nonsmoker when you looked at them separately, otherwise 11 you would have expected difference. 12 Q. Before we wrap up on asbestosis, we promised 13 the Court we would show the chart from the Hessel study 14 where this synergistic effect was tested. Can you just 15 very briefly explain what that is? 16 A. Yes. That is additional analysis of one of 17 several that we did. They have hypotheses we test with 18 these analyses. If you add brake work to other known 19 risks, you would expect a sum total risks that's higher. 20 People with no -- the comparison is for people 21 with no history of brake work and insulators who were 22 both insulators and have history of brake work; and 23 similar for ship builders, comparison of people who are 150 1 ship builders but never worked on brakes and people who 2 are ship builders and worked on brakes. And at least 3 from these data -- this, of course, is a small data set 4 -- at least from this data there's no evidence that 5 you -- there is a contribution by brakes. 6 Q. If you don't see risk for mesothelioma and you 7 don't see risk for lung cancer, would you expect to see 8 risk for asbestosis? 9 A. No. 10 Q. Why not? 11 A. Because asbestosis -- and I think there was 12 agreement in this room -- that asbestosis would require 13 much higher doses. 14 Q. Have there been studies published to support 15 the conclusion that auto mechanics do not have increased 16 risk of asbestosis? 17 A. Well, there -- the trick with asbestosis is 18 it's hard to study with regular design other than 19 cross-sectional studies. Because it's a state as 20 opposed to an event. Although the Coggon analysis, the 21 PMR analysis from the U. K, they did analyze proportion 22 mortality from asbestosis and found no increase. 23 Q. How about the other PMRs? 151 1 A. I don't remember seeing episodes anywhere, but 2 that's the one that is just because it's there; they 3 discuss it in the same breath as mesothelioma. 4 Now, cross-sectional studies, there were 5 several of them. There's one in Switzerland, 6 Lobboilog -- 7 Q. Spell that for the court reporter. 8 A. L-O-B-B-O-I-L-O-G. There's one in Germany, 9 Eilenhausen, E-I-L-E-N-H-A-U-S-E-N, et al. And then, of 10 course, there is one in Sweden, study by Marcus, 11 M-A-R-C-U-S. 12 And, of course, there's two Mount Sinai 13 studies. There is a Lorimer, 1976, a study that's a 14 descriptive study that did not use controls that 15 reported -- and actually that was one of the earlier 16 reasons why people got concerned about asbestos-related 17 disease, not just asbestosis, of course, primarily 18 mesothelioma. It was triggered by the Lorimer report in 19 the seventies. Because they evaluated a group of brake 20 mechanics and found, or reported rather, reported x-ray 21 changes that in their mind were consistent with 22 asbestosis and, therefore, concluded asbestosis may be 23 elevated in that group. 152 1 Then to follow up, 1984 -- '84 by Nicholson 2 that used a control group. And I've read this paper 3 multiple times. I haven't seen anywhere mentioning of 4 the word "asbestosis." 5 Q. That's the question I was going to go ask you 6 Dr. Frank testified, if I heard correctly, that 7 Dr. Nicholson reported cases of asbestosis; is that 8 correct? 9 A. The word "asbestosis" was not seen as a 10 diagnosis identified in any of these papers. 11 Q. Does the literature suggest, in your opinion, 12 an increased risk of asbestosis among car mechanics, 13 brake repair workers, or people who work in garages? 14 A. No. I have not seen studies that would be - 15 any of those cross-sectional studies that were found a 16 true diagnosis of asbestosis. 17 Q. Now, Dr. Frank said, I think also, that the 18 risk of asbestosis was shown to increase with exposure 19 in those garages. Do you believe that's the conclusion 20 of the Nicholson report? 21 A. No. 22 Q. All right. And I believe Dr. Frank also said 23 that 153 1 A. I'm sorry. The conclusion may be there, but I 2 know the data well enough to -- 3 Q. I'll ask you to explain that to the Court in a 4 second. 5 Dr. Frank also said that Nicholson had found 6 out it was latency. Is that accurate? 7 A. No. 8 Q. Based on the data? 9 A. No. I don't think there is even a conclusion 10 about latency. 11 MR. TARRY: Your Honor, about to wrap up. 12 If you could come down. And I'm going to put 13 these tables up and ask you to explain these conclusions 14 to the Court. 15 (Pause.) 16 THE WITNESS: This is Table 11. They looked at 17 the issue in several different ways. The Table 11 looks 18 at x-ray changes in people who did brake work and did no 19 brake work. And the critical numbers here, these two 20 lines (indicating), people who did no brake work, the 21 percent with 1./0 or greater parenchymal change, which 22 is not specific of asbestosis, is 20.7. A similar 23 percent for people who did brake work is a little bit 154 1 lower, 18.5. This actually continues for various 2 categories. 3 An important thing, of course, these -- they 4 don't report these levels of significance. 5 BY MR. TARRY: 6 Q. How is that usually reported? 7 A. You can easily report a p-value. P-value is 8 the probability your differences are observed due to 9 chance. P-value less than .05 is considered to be 10 significant. And this is not significant at all. From 11 statistical point of view these results are the same. 12 By no means I'm saying that brake work have lower 13 proportion of change. I'm just saying there is no 14 evidence of increase. 15 Now, the next table, that's very important, 16 because this is not adjusted for age. And one needs to 17 adjust for age. That's Table 13. Again, brake work, no 18 brake work. There are other categories that are 19 important, of course, and maybe they're informative. 20 But really, the gist of this analysis is brake work 21 versus no brake work. 22 And here, after adjustment for age, there is a 23 difference but in the opposite direction. There's 155 1 18.6 percent for people with no brake work, and 2 24.5 percent, people with brake work. But these 3 differences are not statistically significant. 4 There's a quick way to eyeball it. And that 5 is, if the intervals overlap, instead of looking at the 6 number, one needs to look at the intervals. And if 7 those intervals overlap, then statistically one cannot 8 conclude the difference is real, not due to chance 9 alone. So 24.5 minus 2.5 is 22. 18.65 plus 4.1 is 10 22.7. The intervals are overlapping. 11 Q. Table 16 also is an important table. Why is 12 that? 13 A. Table 16 is very important. There you see a 14 big difference among people who did brake work and 15 people who did not do brake work. That difference is, 16 in fact, statistically significant because here you get 17 as far as 23.4. And the intervals are not overlapping. 18 That's the real difference. 19 But the most important thing here is the lower 20 row, the percentage of participants with other asbestos 21 exposure. And the percentage of participants is higher 22 than any other category. So we don't know whether 23 that's an explanation -- an explanation that was offered 156 1 is that these particular categories involved in 2 grinding. But it couldn't just as well be that the 3 difference is explained by differences in other 4 exposures. 5 Most important thing, though, most important 6 thing, though, if there is no increase in risk in 7 mesothelioma and no increase in risk of lung cancer, it 8 appears inconceivable that asbestosis as disease, not as 9 x-ray changes -- because it's pretty nonspecific 10 finding -- would be elevated. But anyway, this is one 11 of the important conclusions that comes from that table. 12 Q. What about their analysis of the percentage of 13 individuals who had the x-ray abnormalities when they 14 grouped them by years of garage work? What can you make 15 of that, and are there some inconsistencies there? 16 A. There were two tables that both of which are - 17 could be potentially very useful. One is looking at 18 duration of work. In other words, people -- in this 19 case, they use more than 30 years and less than 20 30 years. Duration is not the same thing as latency. I 21 know this. This is just not right. Duration is the 22 measure of dose. It's not a latency. 23 The problem with this table is that all we know 157 1 is that, compared to within the group that did the brake 2 work, we don't know whether the same pattern -- and 3 there is a difference, of course -- 19.4 versus 28.7. 4 But we don't know whether or not the same pattern would 5 have occurred among people who did not do the brake 6 work. That's the right way to do the analysis, before 7 you conclude that that's an occupation that doesn't. 8 All you know is that people who work longer tend to have 9 worse lungs or more changes on the lungs. 10 One second. And then it's possible perhaps to 11 recalculate, do the same thing for known brake work. 12 Because the number is a given. It's a bit of a problem, 13 though If one -- and it'' s ju st -- you just don't know 14 what numbers are showing. Bec ause three out of 48 is 15 not 27 percent. And it's not a matter of opinion. 16 Similarly, if you add 94 plus 170 plus 18 plus 17 31, you'll never get 413. That's why it's very hard to 18 recalculate -- it's possible sometimes to just take the 19 data from a study and recalculate to answer a question 20 that you're interested in, but here I sort give up. 21 Now, the latencies is the next table, 17(b). 22 They don't call it latency. It's years since onset of 23 exposure; not the same thing, a completely different 158 1 thing than duration of exposure, and has nothing to do 2 with duration follow-up. There's a different thing. Be 3 happy to explain it. But the point is that, if one were 4 to look at latency, these differences are no longer 5 significant. The confidence intervals largely overlap. 6 MR. TARRY: Questions, Your Honor? 7 THE COURT: No. 8 BY MR. TARRY: 9 Q. Dr. Goodman, in summation, is the data 10 presented by Nicholson to NIOSH in 1984 consistent with 11 a hypothesis that garage work, whether brake work in 12 particular was involved or not, increases the risk of 13 asbestosis? 14 A. No. 15 Q. We talked about epidemiology of what I'm going 16 to call friction fiber exposure, whether brake work was 17 involved or not. Why don't you put greater weight than 18 you have on other asbestos exposures or chrysotile 19 exposure from other settings? 20 A. Well -- 21 Q. Let me back up. Withdraw that question. 22 Have you disregarded other epidemiology and 23 other scientific literature? 159 1 A. No. I have reviewed other epidemiologic 2 studies as well as literature from other fields of 3 knowledge. 4 Q. Why then have you not put more weight on those 5 studies of other exposure settings than on the garage? 6 A. Yeah. I think it takes us back to the issue of 7 systematic literature review. And if you recall, the 8 first step was to formulate a hypothesis and then go 9 through the literature and pick studies that are 10 relevant studies that actually test the hypothesis that 11 you are interested in. 12 And, you know, if one can make assumptions in 13 the absence of the data, one can make assumption -- and 14 that's a reasonable thing to do, and, you know, 15 researchers do it all the time -- when they don't have 16 the data, they might assume that a similar set of 17 circumstances applicable to this particular set of 18 circumstances, no problem with that. 19 But if you have the data that are relevant, you 20 can't operate based on assumptions . And that 's why, you 21 know, the relevance of studies of miners and millers in 22 China is questionable at best with respect to studie s of 23 brake mechanics. 160 1 Q. And if you consider evidence outside of 2 epidemiology like the Langer paper that we've heard - 3 the most recent Langer paper we've talked about -- first 4 off, is that a paper that you are familiar with? 5 A. Yes. I cite it in my review. 6 Q. Even if you're not an expert in mineralogy, as 7 an epidemiologist who has to make determinations about 8 the degree of relevance in certain types of 9 epidemiologic studies, how do those studies form your 10 opinion on how to make a systematic review? 11 A. Well, if they're relevant, then, you know, 12 they're certainly important. The problem with studies 13 of other fields is that there's not much out there that 14 would be specifically relevant to the exposure to 15 friction fibers and risk of disease in either humans or 16 animals, or at least I'm not aware of any studies that 17 would test that hypothesis specifically. 18 Two studies that come to mind that came out 19 recently -- I both cite in my review when I discuss 20 epidemiology -- because it is puzzling, I must say. 21 Most people who set out to look at this occupation were 22 -- had an expectation that they will find an increased 23 risk. That's why, you know, you often see those 161 1 comments in the discussion sections as to the authors, 2 you know, they didn't expect it to happen. 3 So the next question is: Are all these studies 4 wrong or is there an explanation that we just don't 5 have? And, you know, the two studies that I mentioned 6 that are relevant is the study by Langer, Reduction of 7 Biological Potential of Chrysotile Asbestos Arising From 8 Conditions of Servicing Brake Pads, and the other study 9 that is relevant from sort of a biological is a study by 10 Roggli when he compared -- and I know the shortcomings 11 of comparing pleural tissue to -- expressed shortcomings 12 of comparing pleural tissue to lung tissue. But that's 13 the study that's relevant. Because it's associated with 14 motor vehicle mechanics and people who have no exposure. 15 I just don't have any other relevant studies from other 16 sister disciplines. 17 Q. Dealing with friction fibers? 18 A. Dealing with friction fibers. 19 MR. TARRY: Your Honor, I think, in the 20 interest of time, we cut out a lot where we were 21 originally going to discuss about the data and claims 22 made about data from Australia. I think, in the 23 interest of time, if there were many questions about 162 1 that data, I would proffer the witness, who is familiar 2 with it and has reviewed all of the raw data from 3 Australia, if there were any lingering questions about 4 it; otherwise, I turn the witness at this point for any 5 other questions. 6 THE COURT: I think that you posed a question 7 when you were discussing other studies about attempting 8 to explain whatever differences may exist there in 9 conclusions from the epidemiological studies we've been 10 discussing; are they wrong or is there perhaps some 11 other explanation for the differences that -- did you 12 call them in other system disciplines? 13 THE WITNESS: Sister disciplines. 14 THE COURT: Have you undertaken to do that as 15 it relates to Langer and is it Roggli? 16 THE WITNESS: No. Unfortunately, it's not my 17 field. But I must say I'm more than anybody else 18 curious to find out what actually happens when you take, 19 honest to God, brake fibers and use it even in animal 20 tissue or cell cultures. I just don't know what 21 happens. And if the results are contradicting 22 epidemiologists, I don't have to scratch our head and 23 think about it. But it's not my field, regrettably. 163 1 But you can't use just studies of short fibers 2 because not all short fibers are the same. You can't 3 just describe chrysotile because it's a heterogenous 4 group of minerals. I don't know what exactly happens to 5 these fibers before they get into the airway. Sure, 6 they get into the airway of all these people, but 7 something is happening. 8 THE COURT: And it's correct to say -- and I'm 9 sure that this is fundamental -- but what you are able 10 to draw from the epidemiology that exists, based on your 11 interpretation of it, there is no increased risk. 12 THE WITNESS: Correct. 13 THE COURT: The reason why that is is not 14 something that epidemiology tries to answer. The reason 15 why there is no increase in risk is not something that 16 epidemiology is going to answer; that would be left to 17 one of the sister disciplines - 18 THE WITNESS: Well - 19 THE COURT: -- if they want to test that 20 hypothesis? 21 THE WITNESS: The traditional epidemiology, the 22 one that we discussed today can only talk about presence 23 or absence of association. Of course, if you have an 164 1 association, then you can move onto Bradford Hill 2 criteria. If you don't have an association, then you 3 have to look for biological plausibility of those 4 findings. 5 And the future of epidemiology is something 6 that I more or less started practice now is molecular 7 epidemiology studies, which is sort of marriage of lab 8 science with population studies. These are sort of the 9 20th century epidemiology. But it is time to move into 10 21st century with occupational study. 11 THE COURT: Has there been an effort to take 12 the studies that you reported in your data study and 13 look for a biological plausibility? 14 THE WITNESS: Well, the only relevant studies 15 are Langer -- that I can find are Langer, and some with 16 Roggli, too. There are people who hypothesize. But the 17 point is explanations may be half a dozen. Is it short 18 fibers? Could be. Is it because it's the lowest grade 19 chrysotile? Could be. Is it because those fibers 20 undergo all those kinds of changes before they become a 21 brake? Could be. Is it because they undergo additional 22 changes after -- when the brake is replaced, after the 23 actual braking occurs? Could be. 165 1 What we do know, they cannot be the same as 2 fibers they got from the ground, the ones that you find 3 in insulation. We know they're different. But the only 4 way to explain why they're different biologically is 5 perhaps to marry lab science with epidemiology. 6 THE COURT: That hasn' t been done yet? 7 THE WITNESS: No, not that I know. 8 THE COURT: Those are all the questions I have 9 for you. 10 Here's what I'm thinking, my two o'clock 11 conference is going to not take any more than a half 12 hour, which would put us to 2:30. If we took an hour 13 luncheon recess, that would bring us back here at 2:45. 14 Would that be adequate to complete cross-examination and 15 do the brief sort of wrap it up that we discussed 16 yesterday? 17 MR. WADDELL: Yes, sir, I believe. 18 THE COURT: Why don't we do that then. And if 19 it looks like we're running short on time, then we can 20 discuss what our options are at that point. 21 All right. We stand in recess then until 2:45. 22 (Whereupon a lunch recess was taken at 1:46 23 p.m.) 166 CERTIFICATE OF COURT REPORTER We, Lynne B. Coale, Registered Diplomate Reporter and Certified Realtime Reporter, and Domenic Verechia, Registered Professional Reporter, and Patricia Ganci, Certified Realtime Reporter, Official Court Reporters of the Superior Court, State of Delaware, do hereby certify that the foregoing is an accurate transcript of the proceedings had, as reported by us, in the Superior Court of the State of Delaware, in and for New Castle County, in the case herein stated, as the same remains of record in the Office of the Prothonotary at Wilmington, Delaware. WITNESS our hands this 18th day of October, 2005. Lynne Bell Coale, RDR, CRR Official Court Reporter, Cert. # 165-PS Domenic Verechia, RPR Official Court Reporter, Cert. # 162-PS Patricia Ganci, CRR Official Court Reporter, Cert. #163-PS 1 ' '70s - 88:2 '79 - 5:17, 96:12, 146:22 '80 - 96:12 '83 - 82:22 '84 - 5:17, 152:1 '85 - 126:16 '87 - 118:1, 118:23, 119:4 '89 - 117:18 '90 - 96:17, 96:18 '91 - 92:20 '92 - 102:14 '94 - 89:23, 90:18, 91:21, 121:2, 126:18 '99 - 15:14 0 0 - 106:9, 108:4 0.2 - 105:15 0.3 - 108:4 0.4 - 36:19, 37:11, 94:13 0.43 - 94:13 0.46 - 97:17 0.5 - 143:10 0.6 - 126:17 0.62 - 115:7 0.7 - 36:17, 52:12 0.8 - 98:11, 105:14 0.87 - 94:12 0.89 - 94:16 0 9 - 79:17 05 - 154:9 08 - 87:14 1 1 - 36:18, 37:4, 37:11, 37:13, 37:17, 52:12, 52:21, 61:21, 66:14, 74:18, 74:19, 89:17, 97:18, 98:13, 100:8, 111:16, 115:8 1,000 - 30:18, 31:6, 32:5, 35:10, 35:13, 42:15, 43:4 1./0 - 153:21 1.0 - 35:21, 36:1, 36:6, 36:8, 126:17 1.1 - 36:7 1.2 - 37:17, 143:13 1.26 - 143:11 1 3 - 36:17 1.4 - 106:10 1.5 - 36:18, 37:6, 121:13 1.6 - 89:18, 126:17, 143:8 1.67 - 143:9 1.7 - 37:17, 94:13, 94:14 10 - 50:23, 57:16, 60:9, 67:16, 113:4 100 - 34:3, 40:1 100,000 - 42:8, 42:13 1000 - 35:3 11 - 51:1, 79:14, 153:16, 153:17 11.2 - 50:21 12 - 41:3, 41:5, 54:5, 76:16, 79:15, 97:15, 119:6 12-fold - 87:15 13 - 75:17, 154:17 14 - 114:19, 116:16, 116:20, 147:14, 147:15, 147:16, 148:6 144 - 79:16 145 - 79:16 148 - 114:21, 115:6 15 - 107:21, 107:22, 116:23, 122:3, 147:15, 148:5 156 - 79:5, 79:6, 79:8, 79:15 16 - 121:11, 155:11, 155:13 162-ps - 166:19 163-ps - 166:21 165-ps - 166:17 17 - 145:11 17(b - 157:21 170 - 157:16 18 - 64:10, 121:11, 122:3, 157:16 18,000 - 125:16 18.5 - 154:1 18.6 - 155:1 18.65 - 155:9 180 - 125:19, 125:22 18th - 166:13 19 - 116:23 19.4 - 157:3 1950 - 115:17 1950s - 84:9 1955 - 84:9 1959 - 83:4 1960s - 57:9, 125:14 1961 - 83:5 1964 - 87:3, 87:4, 87:5 1973 - 75:18, 84:11 1976 - 151:13 1978 - 119:3 1979 - 96:15 1980 - 82:21, 82:23 19801-3725 - 1:21 1980s - 44:20 1982 - 96:13 1983 - 83:21 1984 - 116:17, 152:1, 158:10 1985 - 88:8 1988 - 52:5, 124:17 1989 - 117:13 1990 - 6:4, 30:19, 96:13, 102:14 1994 - 92:5 1997 - 119:3 1998 - 116:17 1:46 - 165:22 2 2 - 36:19, 52:16, 112:21, 112:23 2,000 - 42:13 2.15 - 74:5 2.2 - 36:19, 37:11 2.3 - 79:18, 105:16, 115:8 2.47 - 95:5 2.5 - 37:21, 155:9 2.8 - 145:10 20 - 1:18, 2:18, 40:7, 40:19, 43:3, 43:7, 68:5, 84:12, 90:15, 98:16, 104:1, 104:4, 109:7, 109:8, 109:11, 124:22, 124:23, 125:9, 125:18, 142:18, 142:19, 142:21, 143:4, 143:6, 143:8 20,000 - 124:19, 124:23, 125:9 20.7 - 153:22 2000 - 31:1 2001 - 61:11, 115:10 2003 - 16:3, 117:4, 117:6, 117:21, 119:1 2004 - 16:20, 16:21, 119:11, 126:1 2005 - 1:18, 2:18, 119:8, 119:12, 166:14 20th - 164:9 21st - 164:10 22 - 155:9 22.7 - 155:10 23.4 - 155:17 24 - 121:12 24.5 - 155:2, 155:9 25 - 35:6, 40:21, 67:16, 128:2 25/75 - 40:11 250 - 35:5, 35:13 259 - 89:5, 89:6, 89:8 26 - 97:15 2609 - 1:21 263 - 57:21, 145:20 27 - 119:3, 157:15 28 - 116:12, 116:13, 116:14, 116:20 28.7 - 157:3 2:30 - 165:12 2:45 - 165:13, 165:21 2nd - 1:21, 65:2, 65:4 3 3 - 106:9, 108:4 30 - 28:11, 139:15, 156:19, 156:20 30th - 65:1 31 - 95:5, 157:17 314 - 115:6 33 - 91:3, 91:5, 91:17, 126:20 35 - 79:18 4 4 - 108:4 4.1 - 155:9 40 - 42:22, 43:5, 43:6, 142:19, 143:6, 143:7, 143:8 400,000 - 125:2 40s - 54:7 413 - 157:17 46 - 57:19, 60:10, 79:12 48 - 120:1, 157:14 5 5.53 - 87:16 50 - 114:4, 115:17, 147:6 500 - 1:21 50s - 54:8 51 - 114:21, 115:6 55 - 91:6, 91:18, 126:22 6 6 - 89:18 60 - 120:2, 142:19, 143:7 65 - 85:12, 87:14 7 7 - 20:10 75 - 35:14, 40:15, 40:21, 116:5 77c-asb-2 - 1:3 8 8 - 36:7, 52:12, 117:1 8.8 - 50:21 80 - 40:6, 40:19, 104:2 85 - 13:14 8b - 2:19 9 9 - 36:7, 51:1, 79:20 94 - 157:16 95 - 36:14, 37:9, 50:22, 50:23 9:45 - 2:19 A ability - 27:7, 27:20 able - 103:7, 103:8, 118:16, 123:5, 133:14, 163:9 abnormal - 9:4 abnormalities 45:8, 156:13 abnormality - 54:14 absence - 32:4, 47:8, 49:9, 67:20, 71:18, 75:3, 98:1, 146:23, 159:13, 163:23 Absolutely - 137:23 absorbing - 81:21 abstract - 88:6, 88:8 academic - 5:22, 131:13 accent - 5:11 accepted - 131:19 access - 23:7 according - 110:9, 147:3 account - 63:3, 70:23, 138:5, 140:22 accountants 66:12, 67:6 accounted - 118:8 accumulate - 58:17 accuracy - 14:2 accurate - 7:15, 30:4, 59:20, 133:12, 153:6, 166:7 act - 146:3 active - 21:9 activities - 66:19, 105:23 acts - 26:5 actual - 37:6, 50:8, 54:11, 61:14, 93:15, 99:15, 141:19, 142:17, 164:23 acute - 54:3 ad - 58:13 add - 9:23, 28:5, 74:12, 126:4, 146:13, 148:13, 148:14, 149:18, 157:16 additional - 6:15, 126:5, 126:7, 127:2, 149:16, 164:21 additive - 145:23, 146:7, 146:13, 147:13, 147:16, 147:23, 148:16 Additive - 146:11 address - 90:1 addressed - 83:14 adds - 149:3 adequate - 165:14 adequately - 139:5, 139:7 adjust - 73:12, 73:17, 76:4, 127:4, 127:9, 154:17 adjusted - 139:12, 154:16 adjustment 154:22 Administration 88:15 adult - 54:7 advance - 4:1 advantages - 76:22 affect - 21:3 afternoon - 112:20 age - 54:2, 54:5, 79:6, 79:7, 89:6, 89:7, 106:20, 114:5, 137:6, 154:16, 154:17, 154:22 age-matched - 89:6 Agency - 24:5 agent - 25:11, 25:23, 26:4 aggregate - 52:17 ago - 12:16, 17:23, 18:22, 20:18, 26:8, 65:16, 119:1 agree - 4:1, 17:11, 17:13, 17:16, 17:17, 19:6, 21:17, 22:1, 26:6, 38:9, 49:1, 53:11, 96:4, 98:23, 135:3, 137:18, 146:1 agreement - 150:12 Agudo - 113:10, 113:12 ahead - 25:6, 57:10, 71:9, 121:18, 144:19 aid - 59:3 air - 67:2 airway - 163:5, 163:6 al - 110:5, 151:9 albeit - 67:8, 75:2, 127:20 alive - 81:9, 103:19, 104:4, 118:5 allow - 75:21, 94:3, 130:21 allows - 27:14, 35:19, 73:13, 132:3, 134:9 Almost - 30:9, 32:22 almost - 52:5, 66:2, 82:8, 87:16, 124:18 alone - 37:19, 37:21, 74:17, 74:19, 146:23, 147:2, 155:9 alternative - 147:11 altogether - 109:2 ambulance - 40:10 amenable - 113:1 American - 14:11, 146:20 amorphous - 21:22 amount - 28:6, 54:16, 64:2 Amy - 2:15 analogous - 145:13 analyses - 126:7, 2 149:18 Analysis - 16:8, 16:9 analysis - 32:10, 55:7, 73:19, 73:21, 74:22, 77:15, 78:12, 79:7, 104:13, 105:7, 105:13, 106:12, 109:6, 112:16, 115:19, 116:15, 126:5, 127:2, 127:16, 129:10, 129:23, 130:4, 130:8, 132:1, 132:2, 133:3, 134:6, 134:9, 137:13, 141:13, 142:2, 142:5, 149:16, 150:20, 150:21, 154:20, 156:12, 157:6 analytic - 18:10 analytical - 27:5, 73:13, 93:15, 95:8 analytically 127:10 analyze - 36:5, 80:23, 132:14, 136:13, 150:21 analyzed - 141:5 analyzing - 126:10 Anderson - 2:5 Angelides - 2:13 Angelos - 88:13 animal - 29:1, 162:19 animals - 26:3, 160:16 Annals - 16:19 anomaly - 100:7 answer - 26:13, 26:21, 60:7, 90:5, 95:12, 126:10, 157:19, 163:14, 163:16 answered - 82:15, 92:13, 143:21 anytime - 59:13 anyway - 33:9, 148:3, 156:10 appear - 75:2 Appearances - 1:7, 2:1 appeared - 53:5 appendix - 121:1 applicable - 159:17 application - 22:5 applications - 20:6 applied - 6:5, 120:21, 121:1 applies - 48:4 apply - 55:7, 90:16 appointment - 8:7 appreciate - 28:17 appreciation 23:20 approach - 28:16 appropriate - 82:12, 104:14 appropriately 109:10 area - 21:9, 39:18, 41:21,42:2 areas - 7:23, 9:9, 58:10, 113:13 arises - 31:8, 103:11 Arising - 161:7 arising - 36:10 artillery - 80:23 asbestos - 3:19, 14:16, 14:20, 15:11, 15:19, 16:23, 17:8, 17:14, 18:4, 18:8, 18:12, 18:23, 19:13, 19:19, 19:22, 21:19, 26:14, 26:16, 55:20, 56:2, 64:21, 68:1, 68:4, 72:10, 75:4, 90:6, 90:10, 90:13, 95:23, 98:3, 106:12, 106:22, 107:19, 108:12, 127:7, 133:21, 134:14, 134:15, 137:21, 137:22, 142:15, 142:17, 145:11, 145:21, 146:3, 146:23, 147:4, 147:12, 151:16, 155:20, 158:18 Asbestos - 1:3, 161:7 asbestos-exposed - 15:11 asbestos-related 3:19, 16:23, 127:7, 151:16 asbestosis - 5:2, 5:7, 18:6, 98:5, 98:7, 98:8, 134:22, 143:21, 146:17, 149:12, 150:8, 150:11, 150:12, 150:16, 150:17, 150:22, 151:17, 151:22, 152:4, 152:7, 152:9, 152:12, 152:16, 152:18, 153:22, 156:8, 158:13 assessment 11:11, 16:10 assessments 11:11 assistant - 12:15 associated - 22:18, 33:17, 33:22, 50:1, 69:21, 70:11, 70:14, 72:2, 127:3, 161:13 association - 20:16, 33:23, 49:9, 51:5, 51:20, 53:3, 60:20, 62:20, 69:19, 70:3, 71:1, 71:13, 71:19, 74:18, 74:20, 93:6, 111:12, 111:13, 111:14, 111:15, 112:2, 112:3, 124:10, 124:12, 130:18, 138:6, 163:23, 164:1, 164:2 associations 22:12, 27:3, 55:19, 57:6, 110:8, 110:11, 112:6, 112:11, 112:14 assume - 143:18, 159:16 assuming - 35:2, 146:8 assumption 159:13 assumptions 159:12, 159:20 attempt - 72:13, 120:13 attempting - 162:7 attributable - 90:8, 100:18 attributed - 90:9 August - 102:15 Australia - 161:22, 162:3 author - 15:7, 17:3, 50:8 author's - 147:3 authored - 10:3 authors - 77:10, 78:4, 108:12, 110:1, 110:9, 121:14, 130:22, 161:1 auto - 3:19, 4:6, 61:22, 62:1, 62:21, 115:22, 121:23, 138:7, 140:9, 140:11, 144:12, 146:9, 148:7, 150:15 automobile - 4:11, 4:19, 83:10, 85:4, 85:11, 116:1, 138:10 automotive - 91:10, 93:21 available - 42:6, 44:16, 86:14 average - 132:4, 138:18, 144:3 averaging - 138:18 awarded - 6:6 aware - 19:19, 61:2, 160:16 axis - 50:3, 50:7, 50:18, 138:17 B background - 5:9, 9:1, 9:19, 69:12, 69:14, 148:4 backwards - 34:23, 46:10, 46:11 bad - 58:16, 58:22, 58:23 balance - 113:5 Balick - 1:13 Baltimore - 7:3, 16:4 bar - 100:4, 101:3 Barcelona - 113:14 Baron - 1:9 Based - 13:3, 153:8 based - 43:3, 76:19, 78:5, 84:3, 85:14, 88:6, 88:11, 88:12, 89:21, 97:15, 101:22, 110:7, 113:16, 113:17, 116:3, 120:5, 120:22, 122:7, 130:22, 133:11, 142:15, 148:7, 159:20, 163:10 basic - 38:5, 146:7 basis - 134:21, 135:12 bean - 28:5, 28:7, 28:9 beat - 138:20 beating - 66:8 became - 6:19, 8:1 become - 8:2, 29:19, 71:22, 109:19, 164:20 becomes - 134:11 begin - 27:14, 32:1, 33:20, 33:23, 34:21, 72:17, 117:22, 130:14 beginning - 46:7 begins - 27:16, 30:2, 34:15 behave - 54:2 behavioral - 10:23 behind - 81:15 Bell - 166:16 below - 52:22 ben - 57:6 Bernard - 2:16 best - 36:15, 36:16, 42:5, 50:13, 69:22, 108:5, 108:7, 108:9, 132:4, 133:7, 133:17, 136:11, 142:3, 142:9, 147:8, 159:22 between - 5:17, 13:19, 15:2, 17:21, 18:12, 20:16, 45:22, 46:2, 50:21, 51:1, 51:5, 55:19, 60:20, 62:20, 67:15, 69:11, 69:19, 70:4, 71:1, 73:5, 83:4, 94:13, 95:13, 96:12, 96:13, 96:15, 96:19, 100:9, 101:10, 105:15, 108:7, 114:1, 114:11, 124:10, 126:17, 128:16, 128:18, 129:2, 130:18, 138:7, 146:5 beyond - 148:2 bias - 103:16, 103:17 Biden - 1:12, 1:13 Bifferato - 1:13 big - 13:5, 89:10, 89:13, 100:16, 155:14 bigger - 76:17, 89:16 Biological - 161:7 biological - 21:4, 23:21, 53:10, 161:9, 164:3, 164:13 biologically - 165:4 biostatisticians 11:7 birth - 47:12 bit - 5:11, 8:23, 11:20, 24:23, 34:8, 53:6, 54:13, 95:4, 99:20, 101:13, 119:14, 140:8, 147:1, 153:23, 157:12 bladder - 51:16 Blair - 149:7 block - 30:16 blood - 45:8, 45:10, 45:11, 45:12 Bloomberg - 7:8 blue - 63:5, 72:4, 133:20 blue-collar - 63:5, 72:4 board - 6:19, 6:22, 6:23, 8:1, 8:3, 8:17, 9:9, 70:8 boards - 8:20 body - 53:14, 58:17, 122:8 boiler - 61:6 bolts - 55:12 bone - 54:1 book - 9:23, 86:5, 86:8, 86:16, 109:23, 110:5 books - 9:15 born - 5:13, 66:14 borrowed - 120:15 bottom - 74:6, 79:4, 107:11, 134:4 Bottom - 52:23 bound - 50:17 Boyd - 2:16 Bradford - 164:1 brake - 80:7, 90:20, 91:4, 91:12, 93:18, 93:23, 94:2, 94:15, 95:2, 95:11, 95:22, 98:3, 106:2, 106:5, 108:18, 121:10, 122:13, 123:1, 126:14, 127:3, 127:13, 129:2, 133:16, 134:1, 134:16, 134:17, 136:16, 149:18, 149:21, 149:22, 151:19, 152:13, 153:18, 153:19, 153:20, 153:23, 154:12, 154:17, 154:18, 154:20, 154:21, 155:1, 155:2, 155:14, 155:15, 157:1, 157:5, 157:11, 158:11, 158:16, 159:23, 162:19, 164:21, 164:22 Brake - 161:8 brakes - 20:8, 150:1, 150:2, 150:5 braking - 164:23 break - 64:15, 64:17, 112:18, 112:23 breakdown - 85:6 Breast - 13:22 breast - 14:22 breath - 151:3 breathing - 67:2 bridging - 114:11 brief - 8:22, 41:12, 165:15 briefly - 5:8, 10:13, 11:15, 17:11, 27:9, 49:10, 73:17, 146:6, 149:15 Briefly - 34:12 bring - 165:13 Britain - 119:20 British - 58:12, 101:15, 101:23, 102:13, 102:20, 120:1 broad - 22:6, 23:6, 53:11, 53:13 broadest - 122:11 Browder - 2:14 Budd - 1:9 builders - 149:23, 150:1, 150:2 building - 30:16, 91:17, 126:21 bulk - 12:4, 12:5, 100:22 bullet - 74:6 bulls - 36:8 bunch - 76:10, 76:11, 146:5 busy - 12:18 C Ca - 1:3 Cadiz - 113:15 calculate - 40:17, 51:13, 68:23, 73:14, 83:18, 90:18, 98:6, 121:14, 121:15, 123:5, 130:21, 140:19, 142:14 calculated - 121:18, 145:5 calculates - 33:5 calculating - 43:13 calculation - 83:17, 97:21, 130:1 California - 83:4 3 call-out - 80:3, 92:4 Cameron - 1:10 Canada - 75:16, 76:14, 76:16 Canadian - 58:11, 75:16 Cancer - 13:1, 14:8, 14:11, 14:12, 24:6, 53:14, 87:23, 89:21, 146:20 cancer - 4:15, 4:20, 12:17, 13:9, 13:16, 13:21, 13:22, 13:23, 14:6, 14:7, 14:22, 15:11, 15:22, 18:6, 18:9, 22:22, 23:3, 23:5, 23:6, 23:8, 23:11, 24:19, 28:18, 29:15, 29:17, 29:21, 30:17, 31:2, 33:17, 33:19, 33:21, 33:22, 34:10, 34:23, 35:1, 35:2, 35:3, 35:9, 38:4, 39:6, 39:15, 40:7, 40:19, 40:21, 41:6, 48:6, 49:23, 50:1, 51:5, 51:16, 52:1, 53:10, 53:11, 53:12, 53:17, 53:20, 53:21, 54:12, 54:20, 55:2, 70:4, 70:9, 70:10, 70:12, 70:21, 71:2, 71:5, 73:1, 76:3, 76:20, 81:13, 93:13, 97:6, 97:11, 118:2, 118:12, 118:15, 120:23, 122:6, 123:16, 123:17, 123:18, 134:23, 135:16, 135:17, 135:22, 135:23, 136:3, 136:23, 137:1, 137:3, 137:4, 137:11, 137:15, 137:19, 137:22, 138:7, 139:2, 139:13, 140:6, 141:6, 141:17, 144:11, 144:22, 145:5, 145:8, 145:14, 145:16, 146:1, 149:4, 150:7, 156:7 cancers - 14:1, 14:2, 14:7, 23:23, 39:12, 55:6, 89:2, 97:7, 97:9, 103:2, 136:11, 137:7 cannot - 37:12, 58:4, 106:12, 136:7, 148:23, 155:7, 165:1 Cantu - 2:7 capital - 5:21 car - 61:22, 62:9, 67:5, 72:3, 75:5, 92:15, 92:16, 95:11, 95:20, 96:2, 115:23, 124:8, 125:7, 152:12 carcinogen - 25:2, 35:17, 70:23 carcinogenic 14:21, 24:13, 25:12, 25:17, 26:16, 26:23 carcinogenicity 14:20, 25:22, 26:3, 26:5 carcinogens 14:16, 26:10 care - 23:7, 78:20, 79:6, 134:7 carpenters - 99:20 cars - 4:6, 62:14 Case - 29:6, 91:13, 91:14 case - 6:17, 15:4, 28:9, 29:2, 29:4, 30:13, 31:22, 32:2, 32:14, 34:13, 34:14, 34:15, 34:18, 34:21, 35:17, 38:3, 38:6, 46:3, 46:9, 46:18, 50:14, 52:19, 59:8, 61:17, 66:15, 70:13, 72:8, 73:2, 73:3, 75:12, 75:20, 76:6, 76:15, 76:18, 76:19, 77:2, 77:17, 84:3, 85:14, 89:7, 89:12, 92:7, 92:18, 98:20, 102:8, 107:9, 107:12, 107:15, 108:3, 117:7, 118:4, 118:11, 120:22, 122:17, 123:22, 125:3, 125:4, 125:10, 125:11, 125:13, 130:15, 136:12, 136:14, 142:1, 142:5, 142:8, 142:22, 145:19, 156:19, 166:10 case-control 34:13, 34:14, 34:15, 34:18, 34:21, 38:3, 46:3, 46:9 case-controlled 38:6 Cases - 66:11 cases - 29:10, 29:15, 30:1, 31:1, 31:6, 32:7, 32:8, 32:11, 32:12, 33:8, 34:3, 34:4, 35:3, 39:6, 39:7, 39:12, 40:1, 52:3, 62:13, 63:2, 63:3, 64:5, 66:2, 66:9, 66:11, 68:15, 68:23, 69:1, 72:17, 73:1, 74:2, 75:5, 79:5, 79:14, 82:2, 83:10, 83:11, 83:15, 83:19, 85:6, 88:12, 89:5, 89:8, 91:9, 91:11, 92:14, 92:23, 97:4, 97:5, 97:15, 102:12, 103:9, 103:13, 103:19, 103:23, 104:9, 114:19, 114:21, 115:17, 116:3, 118:22, 119:2, 119:4, 120:2, 121:8, 121:10, 121:11, 121:12, 122:20, 125:19, 125:21, 128:2, 136:7, 136:12, 147:12, 152:7 Castle - 1:2, 166:9 cat - 56:7 categories - 25:21, 38:18, 54:23, 85:3, 93:20, 99:13, 142:18, 142:21, 154:2, 154:18, 156:1 category - 26:1, 53:21, 54:1, 54:10, 54:21, 79:3, 80:18, 83:9, 85:4, 93:20, 93:22, 94:7, 97:4, 99:13, 104:22, 106:3, 114:18, 114:19, 114:20, 116:1, 116:5, 121:4, 121:9, 122:12, 139:5, 143:6, 155:22 category's - 90:20 caught - 27:13 causal - 110:8, 110:11, 112:6 causality - 60:12 causally - 18:5 causation - 48:16 causative - 22:19 caused - 33:22, 67:10 causes - 23:8, 23:14, 23:16, 23:18, 85:1, 96:15, 96:22, 103:4, 137:2, 137:11 caution - 112:10 cautionary - 110:1, 112:8 cautions - 78:2 cautious - 95:19 caveat - 133:17 Cc - 140:19 Cdc - 14:11 cell - 29:1,53:17, 54:1, 162:20 cells - 38:13, 53:23, 54:12, 74:14, 109:19 census - 124:20, 125:6 central - 118:13 century - 164:9, 164:10 Cert - 166:17, 166:19, 166:21 certain - 42:2, 54:18, 56:11, 59:8, 71:11, 86:12, 102:5, 114:5, 118:11, 131:18, 160:8 Certainly - 23:17, 56:18, 100:16, 101:5 certainly - 44:1, 47:12, 53:20, 57:7, 77:6, 80:6, 97:13, 97:14, 100:16, 102:23, 160:12 Certificate - 166:1 certificate - 43:22, 86:18, 97:12, 97:14, 118:3 certificates - 42:3, 42:12 certification - 8:18, 9:9 Certified - 166:3, 166:5 certified - 6:22, 6:23 certify - 166:6 challenges - 136:3, 136:4, 137:4 chance - 36:12, 37:19, 37:20, 82:9, 102:23, 154:9, 155:8 chances - 142:10 change - 47:13, 65:15, 65:16, 109:14, 153:21, 154:13 changes - 151:21, 153:18, 156:9, 157:9, 164:20, 164:22 chapter - 109:23 characteristics 21:1, 21:2, 53:18, 114:5 characterize - 7:15, 56:8, 56:10 characterized 57:13 charge - 12:16 chart - 36:20, 85:7, 85:8, 100:12, 122:18, 126:1, 149:13 charts - 57:18 check - 65:11, 65:18, 113:2 Checkoway - 56:6, 101:12, 101:17 Chemical - 20:22 chemical - 21:2 chemistry - 21:23 chemotherapy 54:18, 54:19 chicken - 46:23 chicken-and-egg 46:23 China - 159:22 choice - 6:16, 45:16 Christian - 2:12 Christine - 2:16 Christopher - 2:15 chromosomal 54:14 chronological 82:20 chrysler - 3:4 chrysotile - 26:20, 26:21, 80:6, 128:16, 158:18, 163:3, 164:19 Chrysotile - 161:7 chunk - 100:16 Ci - 37:10, 50:21 cigarette - 34:2, 70:15 cigarettes - 50:2, 70:20, 73:5, 73:6, 76:4, 138:13, 138:17, 138:19 circumstance 25:16, 76:10 circumstances 17:17, 17:21, 18:5, 18:22, 19:11, 22:16, 24:18, 48:14, 55:19, 56:1, 56:7, 60:14, 71:11, 145:16, 146:2, 146:4, 159:17, 159:18 cite - 160:5, 160:19 cited - 9:14, 15:15 City - 44:19 city - 5:21 claims - 161:21 clarify - 56:20, 62:8, 68:16, 83:15, 100:8 class - 14:5, 28:11 classes - 14:8 classified - 25:2 classroom - 11:21 clear - 18:11, 45:1, 59:14, 61:20, 65:17, 77:18, 89:8, 91:9 clearly - 20:20, 72:2, 112:13, 123:23, 146:3 clever - 78:20, 80:19 Clifton - 2:13 clinic - 92:2, 92:15 clinical - 9:8, 48:11, 48:16 clinician - 29:8 clinics - 62:12 close - 33:11 closet - 28:7 clue - 22:15 clutches - 20:8 co - 12:21, 12:23, 13:1, 17:3 co-author - 17:3 co-teach - 12:21, 12:23, 13:1 coal - 100:6 Coale - 166:2, 166:16 Coast - 30:21 coauthor - 125:23 coauthored - 15:20 coffee - 29:18, 29:20, 30:17, 31:7, 31:13, 32:2, 32:4, 32:6, 32:17, 33:16, 34:23, 35:5, 35:6, 35:11, 35:13, 35:14, 35:16, 51:15, 51:22, 52:1, 70:4, 70:5, 70:8, 70:15, 70:19, 70:22, 71:1, 71:3, 71:4, 72:23, 73:5, 74:3, 74:8, 76:3 Coggin - 119:21 Coggon - 96:7, 96:8, 96:9, 99:4, 150:20 cognitive - 28:15 cohort - 31:18, 31:19, 31:22, 31:23, 33:4, 34:4, 34:12, 41:4, 46:2, 46:6, 46:17, 51:22, 56:17, 60:17, 98:9, 98:21, 122:16, 122:21, 122:23, 123:3, 123:11, 124:7, 125:15, 125:20, 136:6, 136:18, 136:23, 140:7, 142:2, 142:3, 142:6, 145:4 Cohort - 32:1 cohorts - 15:12, 56:19, 60:21, 125:14 collar - 63:5, 72:4, 72:7, 107:7, 133:19, 133:20 colleagues - 84:16 collect - 42:6, 48:22, 116:7, 136:12, 142:4 collected - 14:2, 75:17, 75:18, 86:19, 115:17 collecting - 118:23 colon - 13:23 color - 47:12, 133:13 colors - 134:4 Columbia - 58:13, 101:15, 101:23, 102:13, 102:20 combine - 94:11, 132:1, 132:7, 134:10 combined - 132:7 comments - 161:1 Commissioner 65:10, 65:15, 65:18 common - 30:17, 41:11, 120:10, 136:9, 136:15, 136:16, 136:19, 136:22 commonly - 9:14, 42:1 community - 61:3, 67:8 company - 10:12, 10:16, 10:17 comparable 81:14, 94:10 compare - 76:20, 89:18, 107:2, 107:4, 125:13, 127:6, 127:14, 133:15 compared - 52:2, 52:3, 52:8, 74:8, 4 103:1, 128:1, 133:9, 134:14, 143:6, 147:20, 157:1, 161:10 comparing - 82:8, 133:18, 161:11, 161:12 comparison 27:21, 125:8, 143:4, 149:20, 149:23 compendium 121:19 compilation - 49:11 compiled - 129:15 complete - 10:11, 15:13, 23:21, 165:14 completely - 54:2, 54:6, 157:23 completing - 10:10 completion - 8:4 component - 20:20, 20:21, 20:23 components - 15:3, 15:5, 148:20 composition 20:22 concentrations 140:19 concept - 27:12, 33:7, 69:22, 72:1 conceptual - 33:18 conceptually 28:14 concern - 30:1, 39:4, 65:14, 65:17 concerned - 140:5, 151:16 concerns - 22:9, 98:2 conclude - 33:10, 33:16, 35:19, 37:12, 60:15, 79:20, 80:1, 80:4, 84:23, 97:2, 97:22, 107:8, 111:5, 122:10, 141:12, 148:23, 155:8, 157:7 concluded - 151:22 conclusion - 4:9, 4:10, 4:13, 4:16, 4:17, 4:18, 4:23, 5:4, 5:5, 26:17, 30:5, 34:9, 36:6, 37:18, 41:3, 64:8, 95:16, 96:4, 108:14, 118:21, 122:7, 139:11, 144:11, 144:14, 144:16, 150:15, 152:19, 153:1, 153:9 conclusions - 3:18, 3:21, 4:2, 26:15, 49:8, 59:5, 59:7, 87:13, 108:21, 112:10, 153:13, 156:11, 162:9 Conditions - 161:8 conditions - 47:6, 100:17 conduct - 11:19, 32:9, 88:18, 103:7, 103:8 conducted - 48:6, 49:23, 75:15, 84:13, 141:13, 142:1, 146:18 conducting 130:10 conference 112:20, 165:11 confidence - 36:14, 36:17, 36:19, 36:20, 37:7, 37:10, 37:16, 50:15, 50:21, 50:22, 50:23, 51:10, 51:13, 52:11, 79:18, 87:14, 89:14, 89:17, 94:13, 95:4, 106:9, 108:4, 115:7, 126:17, 134:8, 134:11, 158:5 Confidence - 37:8, 37:9, 105:15 confident - 87:11 confirmed - 102:12 confounded - 97:19 confounder - 69:16, 69:17, 69:18, 69:20, 69:23, 71:7, 71:8, 72:9, 72:18, 72:22, 77:9, 137:7, 137:8 confounders 71:22, 72:13, 72:16, 75:1, 78:21, 106:15, 106:18, 106:19, 137:12 confounding - 72:8, 74:21, 79:7, 96:1 confronted - 61:12 confuse - 117:9 Connecticut 83:23, 84:5, 84:7, 84:11,86:1 connection - 18:12, 18:14 consecutively 39:8 consider - 64:7, 98:20, 98:22, 106:12, 106:21, 115:12, 115:14, 117:3, 160:1 consideration 73:15, 81:11, 97:20, 101:6, 109:7, 118:20 considered - 13:9, 25:12, 44:2, 72:10, 80:8, 81:18, 82:19, 87:20, 91:20, 96:6, 99:1, 101:10, 106:15, 106:19, 106:20, 113:9, 119:9, 121:23, 154:9 consisted - 124:12 consistency 47:21, 47:22, 48:20, 49:7, 51:8, 53:1, 58:15, 59:4, 59:6, 59:15, 60:2 consistent - 53:2, 57:11, 58:3, 59:21, 60:3, 124:10, 140:1, 143:22, 144:3, 151:21, 158:10 consistently - 71:19 constitute - 25:15 constructing 93:15 construction 99:21, 125:4 consulting - 10:15 consume - 35:16 consumption 32:3, 32:4, 33:16, 35:11, 70:15 contact - 88:17 contacted - 76:13, 126:11 contention - 56:22 Continued - 2:1 continues - 154:1 Continuing - 117:2 contract - 44:20 contracted - 54:4, 54:7 contractor - 21:13 contradicting - 162:21 contribute - 144:7 contributed 100:22 contribution - 90:5, 146:14, 148:8, 148:13, 150:5 Control - 39:1 control - 30:6, 31:22, 32:21, 34:13, 34:14, 34:15, 34:18, 34:21, 38:3, 39:11, 39:17, 39:21, 40:2, 46:3, 46:9, 46:18, 72:13, 72:15, 73:3, 73:4, 75:12, 75:21, 76:6, 76:18, 76:19, 77:2, 77:3, 82:1, 84:3, 89:7, 89:12, 93:5, 98:21, 101:22, 102:8, 102:23, 103:3, 103:16, 103:19, 105:5, 113:21, 113:22, 113:23, 114:2, 114:8, 114:12, 122:17, 136:12, 136:14, 139:3, 139:4, 139:7, 142:2, 142:5, 142:8, 142:9, 142:22, 152:2 controlled - 38:6, 73:6, 106:20, 109:13, 117:7, 120:22, 139:2, 139:17 Controls - 102:16 controls - 27:21, 35:10, 35:11, 35:13, 39:12, 40:13, 73:2, 74:2, 79:6, 79:15, 81:5, 85:1, 85:6, 85:14, 88:14, 88:16, 88:23, 89:6, 91:13, 91:14, 93:7, 94:11, 102:18, 102:19, 103:12, 104:5, 104:10, 107:21, 113:18, 114:19, 114:21, 120:23, 121:11, 121:12, 136:13, 151:14 convening - 24:15 convention - 37:10 convert - 140:20 Cooper - 2:9 cornerstone - 27:4, 27:5 correct - 9:17, 13:6, 14:6, 17:15, 21:8, 23:22, 32:21, 33:2, 34:18, 34:19, 35:22, 47:18, 48:9, 56:17, 57:3, 63:21, 64:5, 66:5, 71:13, 72:17, 77:20, 78:7, 82:22, 101:2, 117:14, 117:18, 121:18, 129:8, 129:17, 135:17, 140:23, 152:8, 163:8 Correct - 6:2, 12:8, 13:7, 14:14, 23:1, 25:3, 25:18, 47:19, 52:8, 57:4, 64:6, 66:6, 70:21, 77:21, 90:14, 97:1, 106:16, 107:16, 108:2, 163:12 correctly - 16:4, 91:3, 91:17, 152:6 Corroon - 2:5 cotton - 28:5, 28:7 Council - 126:10 count - 13:20, 32:7, 109:9, 109:10 counted - 31:5, 42:21, 64:11, 78:19, 142:6 countries - 5:18 country - 5:23, 6:11, 9:17, 76:17, 116:14, 118:9 County - 1:2, 166:10 couple - 26:8, 38:1, 53:4, 102:2, 126:8, 129:13 course - 10:2, 12:23, 13:1, 13:2, 14:9, 14:20, 18:6, 19:18, 20:22, 23:5, 27:12, 31:2, 42:3, 48:11, 51:8, 52:19, 54:21, 56:15, 57:8, 71:4, 76:22, 81:5, 83:5, 85:17, 86:9, 86:15, 87:10, 95:1, 97:13, 99:9, 102:21, 103:11, 106:3, 106:8, 107:10, 108:3, 108:13, 111:17, 116:15, 132:16, 141:6, 145:14, 145:16, 150:3, 151:10, 151:12, 151:17, 154:3, 154:19, 157:3, 163:23 courses - 12:21 Court - 1:1, 1:20, 3:1, 3:2, 3:3, 3:22, 5:9, 8:23, 11:15, 19:4, 36:22, 37:2, 47:16, 49:10, 49:18, 55:14, 55:16, 59:12, 59:14, 60:1, 64:9, 64:13, 64:17, 65:7, 65:9, 65:13, 65:21, 65:22, 66:7, 79:10, 81:20, 82:1, 82:3, 82:15, 82:17, 87:9, 94:21, 112:20, 113:2, 116:6, 120:7, 125:8, 126:2, 128:5, 129:21, 133:2, 140:21, 143:3, 145:21, 149:13, 153:3, 153:14, 158:7, 162:6, 162:14, 163:8, 163:13, 163:19, 164:11, 165:6, 165:8, 165:18, 166:1, 166:5, 166:6, 166:9, 166:17, 166:19, 166:21 court - 40:23, 151:7 Court's - 27:10, 118:20 Courtroom - 2:19 courtroom - 53:10 cream - 76:3 created - 131:15 criteria - 131:19, 145:3, 164:2 critical - 153:19 criticism - 78:3 critique - 130:5 cross - 45:17, 46:3, 46:9, 46:17, 46:22, 47:1, 47:3, 47:4, 47:7, 150:19, 151:4, 152:15, 165:14 Cross - 45:18, 45:19 cross-examination - 165:14 cross-sectional - 45:17, 46:3, 46:17, 46:22, 47:1, 47:3, 47:4, 150:19, 151:4, 152:15 Cross-sectional 45:18, 45:19 Crowe - 2:14 Crr - 166:16, 166:21 crude - 70:6, 105:7, 126:16 Crumplar - 1:14, 1:15, 64:20, 65:8, 65:11, 65:19 crystalline - 21:23 cultures - 29:1, 162:20 curious - 162:18 current - 12:14, 14:10 cursor - 49:15 curve - 143:19, 144:1 cut - 105:1, 109:19, 161:20 D Daimler - 3:4 Daimler-chrysler 3:4 Daimlerchrysler 2:3, 2:5 Danish - 117:8, 119:1 data - 60:15, 73:10, 75:16, 75:18, 88:6, 89:21, 96:20, 99:6, 108:14, 108:17, 116:7, 116:16, 116:18, 116:21, 118:2, 118:11, 118:12, 118:23, 119:3, 119:23, 121:15, 123:6, 126:6, 126:11, 128:5, 128:9, 128:17, 130:20, 130:22, 142:4, 143:22, 144:2, 144:20, 147:8, 147:11, 148:21, 148:22, 150:3, 150:4, 153:2, 153:8, 157:19, 158:9, 159:13, 159:16, 159:19, 161:21, 161:22, 162:1, 162:2, 164:12 database - 42:12, 116:9 date - 64:11 Daubert - 1:18 David - 2:11 daylight - 28:1 days - 28:8 dead - 66:8 deal - 14:1, 72:22, 74:22 Dealing - 161:17, 161:18 deals - 41:18 death - 42:3, 42:5, 42:12, 43:21, 43:22, 45:5, 45:6, 85:1, 86:9, 86:17, 96:22, 97:12, 97:14, 98:15, 100:17, 103:4, 118:3 Death - 42:3 deaths - 41:20, 42:8, 42:13, 42:16, 42:17, 42:18, 42:19, 42:22, 43:3, 43:5, 5 43:18, 83:3, 83:4, 83:6, 96:15, 99:7, 100:13, 100:15, 101:2, 115:18, 115:19 debate - 60:19, 61:2, 67:8 debruim - 2:11 decades - 17:23 deceased - 81:10 December - 16:3 decided - 6:20, 7:1, 131:15, 131:18 decisively - 61:10 decrease - 87:15 decreased - 143:11, 143:13 Defendant - 2:3, 2:5, 2:7, 2:9 Defense - 14:13 define - 29:4, 48:20, 58:18, 59:10, 68:2, 94:5, 97:4, 134:1, 134:17 defined - 42:2, 54:15, 83:9, 94:19, 97:5, 97:10, 114:18, 122:12, 124:9 definitely - 93:22, 94:14, 95:2, 101:18 definition - 22:8, 22:14, 26:7, 71:8, 74:21, 97:6, 105:1, 126:14, 126:15, 142:23 definitional - 69:15 definitions - 67:18, 69:14, 99:21, 130:19 degree - 4:3, 7:19, 7:22, 9:8, 11:8, 160:8 Delaware - 1:1, 1:21, 42:11, 166:6, 166:9, 166:12 delineation - 100:9 demonstrate 69:22 demonstrated 57:6, 71:18 Denmark - 118:7, 118:13 denominator's 69:1 dentist - 9:20, 9:21 Department - 3:16, 8:14, 14:12 departure - 111:16, 111:18, 111:19, 129:1 depicted - 17:5 depiction - 122:8 describe - 163:3 described - 66:16, 86:7, 91:1, 93:10, 131:8, 131:11 describes - 86:14 descriptions 106:6 descriptive - 151:14 design - 33:3, 33:4, 34:13, 34:14, 34:15, 41:11, 41:13, 43:12, 45:19, 45:20, 75:11, 83:17, 84:2, 98:19, 98:21, 101:20, 122:16, 136:12, 150:18 designs - 44:14, 45:17, 59:2, 142:9 detail - 24:23, 44:18, 119:15 details - 86:13 detectable - 96:3 detective - 46:6 determinants 22:10 Determinants 22:17 determinations 160:7 determine - 85:19 determined - 21:22 determining 24:17, 59:3 develop - 12:16, 24:12, 34:10, 41:23, 67:6, 100:20 developing - 56:3, 59:19, 61:1, 92:17 develops - 28:10, 45:9 diagnose - 55:3 diagnosed - 45:6, 45:10, 76:16, 102:13, 109:12, 118:4, 118:11, 118:15 diagnosis - 85:2, 152:10, 152:16 diameter - 21:20 die - 41:19, 42:1 dietary - 15:2 diets - 15:3 differ - 59:21, 135:23 difference - 12:3, 20:14, 33:10, 34:5, 45:22, 46:2, 69:11, 73:4, 95:13, 127:23, 149:11, 154:23, 155:8, 155:14, 155:15, 155:18, 156:3, 157:3 differences - 19:10, 21:7, 93:18, 154:8, 155:3, 156:3, 158:4, 162:8, 162:11 different - 11:18, 17:13, 19:2, 28:3, 31:21, 33:19, 35:8, 37:13, 37:22, 44:14, 46:16, 49:3, 49:4, 49:11, 53:14, 53:16, 53:17, 53:18, 53:23, 54:3, 54:4, 54:6, 54:19, 55:1, 58:8, 58:9, 58:10, 58:19, 58:20, 59:1, 59:2, 68:3, 69:6, 74:14, 81:8, 88:5, 94:6, 94:10, 95:10, 97:6, 99:21, 102:16, 103:1, 109:4, 117:10, 117:13, 117:15, 120:8, 130:1, 130:11, 131:17, 136:2, 136:4, 137:5, 139:14, 145:15, 153:17, 157:23, 158:2, 165:3, 165:4 differential - 82:13 differently - 41:5, 60:6 difficulty - 91:1 dig - 61:14, 114:15 digits - 145:18 dimension - 21:20 Diplomate - 166:2 direct - 127:18 Direct - 3:9 direction - 154:23 directly - 67:13, 69:9, 103:16, 104:2, 104:8, 104:10 directories - 86:3 disadvantage 76:23, 77:1 disadvantages 76:22 disagree - 22:3, 49:2, 135:3 disagreement 132:22 discipline - 6:16 disciplines - 9:8, 161:16, 162:12, 162:13, 163:17 discovered - 18:13 discrepant - 19:2 discuss - 17:8, 41:8, 44:4, 66:2, 71:11, 119:14, 137:17, 151:3, 160:19, 161:21, 165:20 discussed - 20:18, 37:4, 99:15, 101:13, 116:11, 119:21, 120:7, 123:18, 129:11, 140:8, 163:22, 165:15 discussing - 162:7, 162:10 discussion - 44:21, 78:9, 80:11, 117:12, 123:20, 161:1 disease - 3:19, 4:7, 5:2, 18:12, 20:17, 22:10, 22:16, 31:13, 31:14, 31:15, 33:19, 34:15, 38:14, 38:15, 38:20, 38:22, 38:23, 39:3, 46:20, 48:16, 53:21, 54:4, 54:6, 54:7, 55:20, 57:1, 59:19, 60:20, 64:4, 66:15, 69:20, 71:6, 71:13, 71:20, 71:21, 72:2, 75:2, 81:12, 89:2, 89:10, 90:9, 101:4, 105:5, 109:12, 128:7, 136:15, 136:17, 136:19, 136:22, 141:10, 144:7, 151:17, 156:8, 160:15 disease-free - 38:15 diseased - 107:23 diseases - 17:21, 22:19, 45:6, 54:3, 54:21, 100:17, 135:9, 143:20 disproportionate 100:14 disregarded 158:22 distinction - 120:6 distinctions - 69:10 distinguish - 54:15 distribution - 22:10, 22:15, 100:13 distributions 74:13 diversity - 58:16, 58:19 divide - 33:7 divided - 38:17, 41:2, 43:6 Doctor - 57:22, 73:23 doctor - 120:17 doctoral - 11:8 document - 15:15, 42:4, 42:5, 87:4 documented - 29:6, 29:7, 35:12 documents - 19:19, 24:12, 29:9 Domenic - 166:3, 166:19 done - 4:2, 14:15, 14:19, 14:20, 15:18, 46:5, 49:3, 49:4, 51:11, 51:12, 58:8, 70:6, 80:15, 81:22, 85:23, 86:1, 109:6, 113:13, 114:8, 115:15, 117:3, 123:15, 124:8, 127:5, 127:16, 129:16, 131:6, 165:6 dose - 21:20, 127:21, 128:9, 128:11, 128:12, 134:21, 135:12, 135:16, 140:20, 141:11, 143:17, 143:19, 143:23, 144:5, 156:22 doses - 150:13 double - 113:2, 145:18 double-check 113:2 down - 22:21, 36:21, 62:10, 74:1, 79:3, 103:21, 129:21, 146:6, 153:12 dozen - 164:17 Dr - 3:5, 3:11, 9:13, 9:16, 10:3, 10:5, 10:7, 17:3, 19:5, 19:6, 19:10, 44:13, 44:18, 53:8, 55:3, 55:18, 66:1, 69:11, 69:17, 74:23, 84:14, 84:15, 86:6, 87:22, 101:12, 101:14, 101:17, 109:22, 113:8, 113:12, 117:11, 120:13, 120:15, 122:15, 122:22, 125:14, 125:20, 128:4, 132:23, 143:15, 143:18, 144:18, 152:6, 152:7, 152:17, 152:22, 153:5, 158:9 dramatically - 57:13 draw - 36:6, 43:17, 49:8, 59:7, 87:13, 108:15, 143:2, 163:10 drew - 69:10, 143:17 drink - 32:6, 35:6, 35:13, 35:14, 70:8, 74:3, 74:8 drinkers - 29:18, 30:18, 31:7, 32:17, 35:5, 51:22, 52:2, 70:5, 71:3, 71:4 drinking - 29:20, 70:4 driven - 59:16 driving - 65:9 dropping - 108:10 dry - 28:10 due - 4:22, 42:13, 42:22, 154:8, 155:8 duly - 3:7 durability - 21:21 Duration - 19:13, 127:19, 156:20, 156:21 duration - 127:17, 142:16, 156:18, 158:1, 158:2 during - 91:2, 125:6, 141:2 E early - 19:5, 44:20, 70:3, 84:9 earth - 87:12 easier - 50:13, 59:5, 88:17, 99:12, 110:14, 117:23, 141:20 easily - 154:7 easy - 78:4, 147:22 eat - 15:5 economic - 142:11 editor - 9:15 effect - 70:22, 98:4, 149:14 effects - 21:4, 21:19 effort - 164:11 egg - 46:23 eight - 121:8, 121:10, 138:19 eighties - 51:12 Eighty - 13:14 Eighty-five - 13:14 Eilenhausen - 151:9 either - 60:7, 109:17, 130:20, 140:21, 145:22, 160:15 electrician - 123:23, 124:2 electricians - 99:21, 123:23 elementary - 27:13, 27:22, 28:13, 28:18, 48:1 elevated - 50:6, 58:4, 60:13, 62:18, 122:14, 124:1, 145:15, 145:17, 151:23, 156:10 elevation - 50:6, 59:9, 60:14 eligible - 6:19, 8:1, 8:3 eliminate - 127:12, 133:14 eliminated - 134:13 elsewhere - 72:6, 91:16, 124:5 emergency - 39:8, 39:14 Emory - 3:17, 11:19, 12:14, 13:19 employ - 84:2 employed - 3:15, 51:3, 93:19 employment 95:23, 142:16 end - 8:6, 31:5, 32:7, 34:4, 46:8, 54:23, 55:1, 91:13, 91:14, 107:6, 109:1, 135:13, 136:22, 143:16 End - 84:10 ended - 79:13, 104:1 engaged - 13:17, 15:22, 16:10, 62:6, 63:2, 63:9, 66:19, 93:22, 94:1, 94:15, 95:2, 105:21, 105:23, 106:5, 122:11, 126:21, 126:22, 6 127:7, 128:6, 133:16, 133:20 engineer - 11:6 engineering 10:22, 11:4 England - 2:17, 96:10, 96:12, 119:20 entail - 24:18 entails - 25:16 entered - 6:11 entertained - 29:13 entire - 85:6 entitled - 77:23 entry - 9:7 environment - 66:4, 71:12 environmental 11:9 Epa's - 21:12 epidemiologic 159:1, 160:9 epidemiological 9:5, 15:18, 18:10, 18:15, 23:12, 27:4, 30:16, 48:10, 59:16, 69:19, 75:8, 130:20, 162:9 Epidemiological 88:7 epidemiologist 3:13, 7:11, 7:13, 9:7, 20:5, 47:22, 53:19, 88:3, 130:12, 160:7 epidemiologists 9:17, 11:8, 16:11, 17:20, 27:2, 30:11, 36:4, 55:18, 62:1, 162:22 Epidemiology 3:17, 11:8, 13:1, 13:2, 14:5, 14:9, 84:10 epidemiology 9:10, 9:15, 9:22, 10:3, 10:7, 10:23, 11:16, 12:17, 13:6, 13:8, 13:10, 18:16, 22:4, 22:7, 22:8, 22:13, 23:2, 27:11, 48:8, 48:21, 56:5, 62:17, 122:8, 129:8, 135:23, 136:1, 146:9, 148:8, 149:3, 158:15, 158:22, 160:2, 160:20, 163:10, 163:14, 163:16, 163:21, 164:5, 164:7, 164:9, 165:5 episodes - 151:1 equal - 32:23, 35:17, 37:4, 40:20, 82:9, 100:13, 108:7, 131:16, 134:21 equally - 82:6 equals - 40:21, 41:3, 141:10, 148:1, 148:18, 148:19 Eric - 2:16 Erin - 2:12 error - 36:11, 36:13 errors - 36:11 especially - 102:3 Esq - 1:8, 1:10, 1:12, 1:14, 2:2, 2:4, 2:6, 2:8 essentially 133:18, 133:19 establish - 18:7 established - 18:14, 70:12, 84:11 establishing - 18:17, 135:21 estimate - 36:15, 36:16, 50:4, 50:14, 63:12, 63:18, 68:9, 68:21, 69:8, 90:17, 99:16, 108:5, 108:8, 108:9, 122:20, 126:16, 132:4 estimates - 57:17, 63:8, 67:12, 67:15, 90:19, 130:21, 132:2, 134:9 et - 110:5, 151:9 ethical - 48:17 etiological - 23:10, 23:11 evaluate - 26:9, 48:23, 56:11, 105:20, 127:3, 136:8 evaluated - 83:8, 87:18, 102:9, 151:19 evaluating - 24:16, 69:19 evaluation - 130:5 event - 45:12, 45:15, 46:8, 65:17, 150:20 events - 45:4, 45:5 Evidence - 13:3 evidence - 20:21, 24:16, 25:22, 26:1, 26:2, 26:4, 47:18, 52:14, 57:11, 58:18, 60:7, 61:9, 87:19, 95:20, 122:8, 122:10, 130:18, 150:4, 154:14, 160:1 exact - 36:12 exactly - 32:18, 41:3, 50:23, 56:13, 114:7, 121:9, 132:12, 163:4 Examination - 3:9 examination 165:14 examined - 3:7 example - 27:22, 28:23, 29:12, 32:11, 33:18, 34:22, 35:5, 39:5, 51:16, 53:22, 54:3, 69:23, 75:23, 79:1, 87:1, 89:18, 91:5, 106:20, 111:2, 111:8, 111:9, 131:8 examples - 61:6 exams - 6:13, 6:15 except - 111:23 exception - 23:18, 47:10, 127:18, 143:20 exceptionally 25:23 exceptions - 30:10 excess - 110:5, 148:2 executive - 21:16 exist - 26:8, 66:23, 162:8 existed - 30:20 exists - 163:10 exit - 6:5 expanding - 9:21 expands - 115:20 expect - 31:11, 34:7, 63:22, 86:21, 112:14, 127:21, 135:15, 144:4, 149:19, 150:7, 161:2 expectation 160:22 expected - 33:8, 43:3, 43:7, 52:9, 62:9, 97:15, 98:12, 99:7, 116:3, 116:23, 119:5, 149:11 experience - 91:18 experiment - 28:12, 29:1, 48:1, 48:3, 48:12 experimental - 26:3 expert - 16:2, 160:6 experts - 10:21, 21:5, 24:15, 65:12, 134:20, 135:7 explain - 24:22, 46:4, 47:21, 49:10, 51:18, 72:18, 103:21, 107:12, 149:15, 153:3, 153:13, 158:3, 162:8, 165:4 explained - 37:19, 122:2, 156:3 explanation 19:20, 20:2, 41:12, 155:23, 161:4, 162:11 explanations 37:20, 164:17 Exponent - 10:12, 10:13, 10:15, 10:19, 10:20, 11:2, 11:17, 11:21, 12:5, 12:9, 15:10 expose - 66:18 exposed - 15:11, 26:4, 32:17, 33:8, 37:5, 38:16, 56:2, 79:11, 80:6, 80:8, 85:15, 90:7, 94:9, 95:21, 109:8, 109:11, 121:8, 121:10, 121:11, 136:20, 141:2, 147:12, 147:21 exposure - 11:11, 17:18, 17:21, 18:22, 19:11, 19:13, 20:16, 24:18, 25:16, 32:2, 33:19, 38:14, 38:16, 38:19, 38:20, 38:21, 39:4, 43:14, 46:20, 55:19, 56:7, 56:8, 56:14, 56:16, 58:10, 60:20, 67:21, 67:22, 69:7, 69:20, 71:12, 75:4, 75:23, 76:2, 76:7, 76:10, 90:6, 94:6, 95:23, 98:3, 107:19, 108:1, 108:12, 110:18, 111:21, 124:10, 126:14, 126:15, 127:15, 127:19, 133:18, 134:15, 134:17, 134:22, 136:16, 136:17, 136:19, 137:8, 140:23, 141:5, 141:9, 141:16, 141:18, 143:10, 143:12, 144:2, 145:16, 145:22, 147:4, 147:5, 152:18, 155:21, 157:23, 158:1, 158:16, 158:19, 159:5, 160:14, 161:14 exposures - 14:22, 15:4, 20:9, 20:19, 25:16, 41:9, 48:15, 56:12, 69:2, 69:3, 71:17, 71:22, 72:2, 72:6, 72:10, 90:10, 90:16, 91:1, 91:2, 97:19, 105:2, 106:13, 106:22, 107:2, 107:3, 108:19, 109:3, 124:4, 124:5, 127:4, 127:10, 127:12, 127:21, 128:19, 129:2, 133:14, 134:1, 134:13, 134:15, 156:4, 158:18 express - 4:2 expressed - 64:8, 128:15, 161:11 extensive - 126:19 extensively - 55:21 extent - 59:22 extreme - 54:16, 108:9 extremely - 18:23, 32:20, 51:7, 52:11 eye - 36:8 eyeball - 61:19, 155:4 eyes - 47:12 F face - 74:4 facilities - 100:4 facing - 136:4 fact - 8:17, 31:10, 47:16, 58:6, 59:20, 68:7, 68:14, 68:20, 72:12, 75:1, 77:22, 109:23, 122:22, 132:15, 155:16 factor - 69:18, 69:20, 70:13, 72:9, 96:1, 137:19, 137:22, 147:18 Factors- 21:18 factors - 15:4, 22:17, 22:18, 23:13, 146:16, 147:18, 147:20 faculty - 3:16, 13:19 fail - 71:2 fair - 96:5, 97:10 familiar - 160:4, 162:1 families - 81:9, 88:18 far - 21:9, 92:3, 107:1, 108:12, 155:17 farmers - 100:6 Farris- 2:12 fashion - 38:5, 50:9, 131:13 fatal - 81:11 feasible - 48:17 feature - 75:20, 85:21,86:2 features - 54:12 federal - 12:10, 116:18 fellowship - 5:22 felt - 80:10, 127:1 few - 31:10, 83:6, 136:22 Few- 116:15 fiber - 19:17, 20:9, 20:20, 21:20, 21:21, 142:17, 158:16 fibers - 17:15, 18:12, 19:12, 20:8, 20:11, 20:12, 20:14, 21:2, 26:20, 94:9, 124:11, 128:17, 135:17, 140:19, 141:1, 144:6, 160:15, 161:17, 161:18, 162:19, 163:1, 163:2, 163:5, 164:18, 164:19, 165:2 field - 8:18, 9:5, 13:5, 13:8, 19:16, 20:2, 22:9, 22:21, 47:16, 56:5, 162:17, 162:23 fields - 9:6, 10:22, 159:2, 160:13 figure - 20:15, 99:11 figures - 99:22 fill - 28:4 final - 65:1, 65:5 Finally- 4:21 finally - 12:13, 67:23 findings - 51:9, 110:7, 111:1, 112:9, 123:14, 164:4 fine - 36:17, 46:1, 70:2 fingerprint - 54:13, 54:14 finish - 36:5 finite - 100:18 Finland- 58:12, 145:12 Finnish- 57:21, 145:20 first - 3:7, 6:11, 7:19, 10:9, 10:11, 12:17, 15:8, 15:9, 16:4, 17:2, 30:16, 31:23, 43:15, 50:8, 57:7, 64:12, 88:4, 97:3, 105:1, 107:9, 108:22, 110:10, 113:21, 126:13, 131:1, 133:6, 137:18, 139:18, 143:21, 146:18, 146:19, 159:8, 160:3 fit - 23:2, 147:11 fits - 147:8, 147:12 fitters - 99:20 five - 13:14, 40:14, 42:11, 63:10, 85:14, 112:17, 128:2, 143:10, 147:2, 147:8, 147:9, 147:14, 147:16, 148:5 fixed - 73:7 fixing - 124:1 flag - 135:14 flat - 128:12, 144:3 flip - 36:20 Floor- 1:21 flow - 31:18, 31:19 fluctuations - 144:4 focussed - 123:16 fold - 34:5, 74:7, 87:16, 147:8, 147:9 folks - 113:3 follow - 27:7, 30:23, 31:4, 32:6, 99:4, 117:8, 119:22, 124:21, 124:22, 125:3, 136:21, 142:4, 152:1, 158:2 follow-up - 99:4, 117:8, 119:22, 124:22, 125:3, 158:2 followed - 119:21, 124:23, 125:9, 125:18, 140:9 follows - 3:8, 28:15 followup - 31:5, 32:7 footnote - 31:20, 7 48:7 foregoing - 166:7 foremost - 137:18 forgot - 109:5 form - 160:9 formed - 114:20 formula - 132:15, 146:7 formulate - 159:8 formulating - 30:3 forth - 10:23, 16:12, 50:10, 79:3, 99:22, 118:17 forward - 124:8 four - 38:13, 38:18, 40:20, 41:2, 54:5, 54:10, 92:14, 92:23, 138:23, 142:18, 148:5, 148:6 four-year-old 54:10 fraction - 33:1, 33:4 frame - 46:18 France- 24:9 Francis- 2:16 Frank- 44:18, 122:15, 122:22, 143:15, 143:18, 152:6, 152:17, 152:22, 153:5 Frankfurt- 92:3 free - 38:15 frequency - 30:15, 73:7, 103:23, 134:7, 138:14 frequently - 50:13 friction - 20:8, 20:9, 41:9, 66:4, 94:9, 124:11, 144:6, 146:14, 158:16, 160:15, 161:17, 161:18 full - 23:20 fun - 8:23 functions - 21:19, 24:10, 24:11 fund - 118:13 fundamental 163:9 fundamentals 55:13 funded - 12:7, 14:11, 84:20, 129:17 funding - 11:20, 12:4, 12:6, 12:18 funnel - 107:10 future - 15:1, 164:5 G Gabay - 2:11 Ganci - 166:4, 166:21 garage - 66:22, 75:19, 79:14, 80:6, 90:21, 122:12, 122:23, 123:22, 124:6, 124:9, 125:5, 128:18, 134:2, 142:17, 148:13, 149:2, 149:3, 156:14, 158:11, 159:5 garages - 123:16, 140:12, 152:13, 152:19 gardeners - 111:3, 111:6, 111:9, 111:10 Garrett - 2:15 Gary - 2:13 gas - 99:20 gather - 65:13 Gechingen - 92:3 general - 17:8, 22:2, 39:9, 40:11, 42:1, 43:1, 52:2, 77:3, 77:4, 81:7, 97:16, 113:21, 113:23, 114:4, 116:4, 120:12, 130:8, 135:5, 138:3 Generally - 18:4 generally - 138:4 generate - 27:15 generates - 28:20 generic - 17:14 generically - 145:22 genetic - 15:4 Gentilotti - 1:13 geographical - 99:5 Germany - 92:3, 151:8 Gianaris - 2:8 gist - 154:20 Given - 74:23 given - 41:21, 46:19, 71:12, 75:1, 144:20, 157:12 glass - 28:4 Glenn - 2:15 God - 162:19 Goodman - 3:5, 3:6, 3:11, 44:13, 53:8, 55:18, 66:1, 74:23, 113:8, 128:4, 132:23, 143:15, 158:9 Gordis - 10:3, 10:5, 10:7 government - 8:11, 12:5, 12:10, 12:11, 116:18 Government - 15:16 grade - 19:13, 19:15, 19:17, 19:20, 20:11, 28:4, 164:18 Grade - 19:16, 20:10 grades - 20:2, 20:3, 20:7 graduation - 8:19 grants - 12:18 graph - 49:22, 50:16, 52:20, 99:10, 122:1, 139:1 Great - 119:20 greater - 153:21, 158:17 greatest - 95:9 green - 122:17, 133:13 Gribbin - 2:16 grinding - 156:2 ground - 165:2 group - 10:20, 11:4, 11:6, 11:9, 11:10, 11:13, 11:14, 16:2, 16:9, 16:11, 20:19, 25:15, 27:21, 30:6, 31:4, 32:5, 32:17, 32:18, 32:21, 38:17, 39:11, 54:15, 55:10, 57:7, 60:22, 62:23, 63:1, 64:3, 66:10, 66:22, 76:8, 80:8, 85:4, 85:14, 93:5, 99:23, 100:2, 100:5, 103:4, 138:11, 139:14, 142:10, 143:1, 146:19, 151:19, 151:23, 152:2, 157:1, 163:4 Group - 25:11 grouped - 99:12, 156:14 groups - 11:3, 32:9, 39:17, 67:5, 73:11, 83:7, 83:9, 93:7 grow - 28:2 guess - 16:13, 60:1 guidance - 24:12 gun - 46:13 Gustavsson - 123:9, 123:11, 123:13, 123:15, 124:5, 127:18, 140:4, 140:5, 140:7, 140:16, 140:23, 144:8 H habit - 138:14 habits - 138:12, 142:12 Hadley - 2:13 half - 16:15, 164:17, 165:11 halfway - 114:1 Hammar - 55:3 Hammond - 146:20 Hampton - 96:10 hand - 81:6, 81:7, 86:15, 86:18 handful - 100:10, 100:15, 115:9 hands - 7:22, 166:13 hands-on - 7:22 Hansen - 117:3, 117:6, 117:7, 117:9, 117:13, 117:15, 117:17, 117:20, 117:22, 119:1 happy - 158:3 hard - 150:18, 157:17 harmful - 48:15 Harvard - 131:9 head - 146:20, 162:22 Health - 3:17, 7:5, 7:8, 7:19, 7:20, 8:7, 8:14, 21:18, 24:8 health - 11:6, 11:13, 11:14, 22:11, 22:19, 92:2 health-related 22:11, 22:19 healthy - 35:9, 35:13, 38:20, 38:21, 43:16, 81:5, 81:9, 103:12, 107:22 hear - 134:19, 135:1 heard - 9:12, 9:13, 17:10, 21:13, 27:10, 29:3, 53:9, 83:22, 84:6, 91:22, 116:6, 135:2, 135:11, 152:6, 160:2 Hearing - 1:18 heavy - 29:17, 31:7, 79:11, 80:23, 100:22 held - 118:18 help - 29:5 helpful - 37:2 Hence - 43:5 Henry - 2:16 hereby - 166:6 herein - 166:10 Hessel - 17:4, 119:11, 119:13, 122:5, 126:1, 126:5, 149:13 heterogeneity - 132:5, 132:6, 132:9, 132:10, 132:15, 132:17, 132:19 Heterogeneity 132:20 heterogeneous 17:15, 19:1, 19:2, 20:19, 26:20, 55:10 heterogenic - 132:8 heterogenous 163:3 hierarchical 78:14, 78:15, 80:18 hierarchy - 79:4, 135:6 high - 45:9, 59:10, 59:11, 70:12, 78:16, 78:17, 91:15, 147:5 higher - 34:8, 52:9, 59:18, 98:12, 128:6, 149:19, 150:13, 155:21 highest - 20:11, 138:12, 142:20 highlights - 85:8 Hill - 164:1 histologically 102:12 histologicallyconfirmed - 102:12 histologies - 53:16 histology - 97:11 historical - 92:6 histories - 121:6 history - 66:3, 67:21, 67:22, 68:4, 78:13, 86:10, 91:6, 91:10, 127:14, 149:21, 149:22 hitters - 100:23 hmm - 84:19 hobby - 63:19 Hodgson - 99:2, 99:3, 99:9, 119:21 hole - 146:4 homogenous 142:11 honest - 162:19 honestly - 65:7 Honor - 3:4, 36:23, 55:11, 64:20, 94:17, 112:17, 153:11, 158:6, 161:19 Honorable - 1:6 hope - 39:8, 135:20 hoping - 15:1, 126:9 Hopkins - 7:4, 7:6, 7:7, 7:17, 8:5, 10:2, 10:8, 11:23 horizontal - 50:3 horse - 66:8 hospital - 5:20, 35:1, 39:6, 76:19, 113:17, 113:22, 114:1, 114:2, 114:8, 114:12, 120:5, 120:22 Hospital - 88:17 hospital-based 76:19, 120:5, 120:22 hospitals - 93:10, 113:16 Hospitals - 88:15, 89:1 hour - 165:12 Hrubec - 149:8 huge - 115:21 human - 18:14, 22:11, 23:12 Human - 8:14 humans - 24:21, 25:12, 25:13, 25:17, 25:22, 26:1, 26:4, 26:16, 160:15 hundred - 10:21, 13:11, 13:13, 13:15, 39:6, 39:7, 40:1, 40:6, 42:16, 136:21 hygiene - 127:19, 140:17 Hygiene - 16:20 hygienist - 88:2 hypotheses - 75:22, 109:4, 149:17 hypothesis - 27:15, 27:17, 27:20, 28:1, 28:2, 28:21, 28:23, 29:13, 29:19, 30:3, 30:12, 30:14, 41:23, 48:16, 49:3, 49:5, 51:19, 59:16, 59:17, 59:20, 68:9, 68:21, 76:9, 78:5, 78:7, 85:9, 90:1, 90:4, 92:10, 92:13, 93:5, 105:17, 114:14, 128:15, 128:22, 130:15, 130:16, 130:23, 131:4, 131:5, 134:1, 135:20, 140:5, 141:9, 143:23, 144:9, 146:7, 146:8, 158:11, 159:8, 159:10, 160:17, 163:20 Hypothesis - 33:21 hypothesize 164:16 hypothesized 70:22 hypothetical 29:11, 31:3, 34:6, 35:17, 38:9 hypothetically 67:9 I I' - 60:9 Iarc - 24:3, 24:4, 24:5, 24:8, 24:9, 24:11, 24:14, 25:1, 25:7 ice - 76:3 iceberg - 82:7 icebergs - 82:9 idea - 29:18, 65:7 ideal - 77:2 identical - 28:6, 28:7, 120:1 identifiable 107:18 identified - 31:1, 35:10, 44:4, 44:9, 53:5, 110:19, 111:22, 113:15, 114:4, 122:21, 152:10 identify - 22:12, 31:17, 32:5, 35:3, 39:6, 44:7, 69:3, 131:1 identifying - 23:13, 32:1 Iii - 1:6, 1:12 illegitimate - 29:23 illness - 34:21 illustrated - 33:14 immediately - 31:8, 66:17 imperfect - 115:20, 127:20 8 important - 20:5, 20:15, 20:21, 20:23, 23:12, 26:11, 27:6, 27:10, 27:18, 38:10, 39:17, 39:22, 40:4, 46:22, 47:8, 47:18, 49:6, 50:5, 53:19, 60:12, 72:21, 75:20, 77:8, 78:22, 80:10, 80:12, 88:11, 90:20, 94:3, 95:7, 95:16, 95:23, 97:6, 98:4, 103:5, 103:6, 108:5, 124:14, 124:15, 124:18, 128:21, 130:13, 133:2, 137:7, 140:15, 141:4, 141:8, 145:3, 154:3, 154:15, 154:19, 155:11, 155:13, 155:19, 156:5, 156:11, 160:12 imprecise - 134:9 Inc - 2:7 incidence - 41:15, 69:5, 69:11, 69:13, 118:11, 135:8 include - 131:4 included - 37:11, 99:5, 121:19, 130:22 includes - 7:18, 10:21, 11:7, 11:10, 16:11, 37:17, 52:12, 124:18 including - 115:2 incomplete - 82:6, 82:7, 86:16 inconceivable 156:8 inconsistencies 156:15 inconsistency 53:1 inconsistent 57:23, 140:1 increase - 52:6, 56:3, 60:4, 61:12, 62:5, 62:6, 66:19, 66:20, 74:7, 87:16, 135:8, 144:5, 145:8, 147:9, 148:20, 149:9, 150:22, 152:18, 154:14, 156:6, 156:7, 163:15 increased - 4:7, 4:11, 4:14, 4:19, 5:1, 5:6, 12:20, 18:17, 24:18, 33:15, 35:20, 53:6, 56:2, 57:12, 58:5, 60:7, 60:23, 66:13, 66:21, 70:5, 80:5, 92:17, 95:21, 111:3, 111:5, 111:10, 111:11, 111:13, 141:17, 143:11, 143:12, 144:12, 144:21, 150:15, 152:12, 160:22, 163:11 increases - 29:20, 52:18, 158:12 increasing - 144:5 indeed - 110:21 index - 142:15 indicate - 128:5, 128:9, 128:17, 128:20 indicates - 148:21, 148:22 indicating - 143:17, 153:20 indices - 142:17 indistinguishable 95:12, 122:14 individual - 36:9, 64:13, 90:1, 130:17, 131:14 individually - 130:6 individuals - 4:14, 5:1, 35:9, 85:15, 156:13 indulge - 66:7 industrial - 88:2, 127:19, 140:17 industry - 12:6, 19:23 influence - 21:18 information - 14:2, 20:4, 42:6, 43:20, 43:21, 48:23, 75:23, 76:1, 82:6, 86:17, 86:19, 88:9, 103:17, 104:11, 115:22, 123:18, 147:3 informative - 98:20, 102:2, 102:3, 102:4, 133:5, 154:19 informed - 102:6 injuries - 8:16 innovative - 113:18 inoculates - 56:23 inquire - 69:1 insist - 133:7 installation - 61:6, 90:20, 91:4, 91:12, 93:23, 94:2, 94:15, 95:2, 106:2, 106:5, 108:18, 121:10, 122:13, 126:14, 126:23, 127:4, 127:13 instance - 28:3, 37:16, 40:10, 45:7, 52:4, 57:15, 57:18, 128:3 instances - 35:23 instead - 41:17, 105:19, 117:23, 155:5 Institute - 14:12, 88:1,89:21, 126:10 insulation - 72:3, 91:19, 165:3 insulator - 58:7, 78:16, 78:18, 78:22, 79:2 Insulators - 125:17, 125:18 insulators - 57:7, 57:14, 59:9, 59:18, 60:18, 78:19, 79:12, 91:7, 107:5, 136:7, 146:19, 149:21, 149:22 intensity - 138:13 interact - 146:16 interaction - 15:2, 128:23, 129:3 interest - 32:2, 34:22, 38:22, 39:3, 41:23, 118:19, 132:23, 161:20, 161:23 interested - 20:7, 61:13, 157:20, 159:11 interesting - 78:12, 81:2, 86:2, 93:16, 114:11, 115:16, 115:20, 123:21 Interestingly 28:14 International - 24:5 internship - 5:19 interpretation - 163:11 interpreting - 47:9 interrupting - 48:18 interval - 36:17, 36:19, 37:7, 37:9, 37:10, 37:11, 37:16, 37:17, 37:22, 50:15, 50:17, 50:21, 51:11, 51:13, 52:11, 79:18, 87:14, 89:15, 89:17, 94:13, 95:4, 96:19, 105:15, 108:5, 115:21, 126:17, 134:11 intervals - 36:14, 57:17, 106:9, 115:8, 134:8, 155:5, 155:6, 155:7, 155:10, 155:17, 158:5 interview - 35:4, 81:3, 81:4, 85:22, 85:23, 86:1, 103:15, 103:20, 104:4, 114:6, 114:7 interviewed - 104:8, 104:10 interviews - 81:22, 87:6, 88:19, 103:7, 103:9, 103:13, 103:19, 104:2, 117:17 introducing 103:16 introductory - 10:2 Introductory 12:23, 14:4 intuitive - 34:14 investigate - 27:2 investigates - 46:16 investigation 23:4, 23:5 involved - 4:10, 4:18, 5:5, 86:19, 91:12, 112:1, 156:1, 158:12, 158:17 irrefutably - 18:5 irrelevant - 131:4 issue - 17:22, 46:16, 46:22, 46:23, 58:15, 72:7, 77:8, 83:14, 114:10, 119:14, 122:7, 134:7, 149:6, 153:17, 159:6 issues - 25:9 itself - 22:9, 49:14, 81:12, 136:11, 136:23 J Jacobs- 1:15, 65:3 Jamaica - 8:8 James- 2:4 jar - 28:4, 28:6 Jarvholm- 124:14, 124:17 Jarvhom- 124:16 Jensen- 117:9, 117:16, 117:23, 118:22 job - 10:9, 10:11, 56:23, 85:19, 106:6, 127:17, 132:8, 139:18 jobs - 127:7 John- 1:8, 7:7 Johns- 7:3, 7:5, 7:16, 8:5 Johnson- 2:6, 2:7, 2:15 joined - 13:19, 15:10 Joseph- 1:6, 1:12, 2:11 journal - 16:19 Jr- 2:2, 2:4 Judge- 98:19 judgments - 110:7, 112:5 jumped - 57:9, 57:10 June- 16:20, 16:21, 16:22 jury - 64:23, 65:5 juxtapose - 51:14 K Kai - 2:14 Kaplan - 2:13 Kathleen - 2:13 keep - 54:22, 86:3, 112:18 keepers - 100:2 keeping - 123:7 keeps - 84:7 kept - 118:23 key - 131:2 kin - 81:4, 81:13, 81:22, 88:18, 103:9, 103:13, 103:18, 104:3, 104:5, 117:17 kind - 44:23, 45:1, 45:2, 96:8, 98:17, 99:3, 105:23, 118:16, 135:16, 141:1 kinds - 31:21, 164:20 King - 1:21 knowledge - 10:22, 133:17, 159:3 known - 37:13, 44:8, 45:7, 50:6, 53:21, 67:20, 68:20, 70:10, 78:15, 127:14, 134:14, 149:18, 157:11 knows - 45:8 Kohlburg - 2:12 Kron - 2:4 Kuery - 2:16 L lab - 164:7, 165:5 labeled - 39:4 laboratory - 18:13 lack - 53:2 Laden- 131:8, 131:9, 131:11, 144:18 land - 69:8 Langer- 160:2, 160:3, 161:6, 162:15, 164:15 large - 10:15, 54:6, 56:3, 60:15, 63:1, 63:22, 64:3, 66:11, 77:5, 77:7, 83:2, 83:6, 103:15, 111:19, 136:6 largely - 158:5 larger - 63:21, 89:15, 89:19, 99:5, 135:16, 138:1 last - 13:20, 41:7, 44:17, 96:16, 110:17 Last- 15:20, 69:15 lasting - 45:13, 45:15 latencies - 157:21 latency - 109:7, 153:6, 153:10, 156:20, 156:22, 157:22, 158:4 lawyers - 107:6 lay - 132:20 lead - 15:7 leading - 9:16 leap - 112:10 least - 20:10, 82:8, 103:8, 103:14, 109:7, 135:12, 150:2, 150:4, 160:16 leave - 64:20, 79:9, 79:16, 88:19 Leblanc- 1:11 lecturers - 100:3 led - 29:12 left - 5:23, 74:16, 163:16 legitimately - 29:18 Lemen- 69:11, 69:17, 109:22, 117:11 lends - 136:11, 136:23 length - 19:17, 20:20, 21:20 Leon- 10:3 less - 26:1,34:14, 43:12, 61:8, 97:18, 98:13, 98:20, 99:7, 108:7, 128:1, 128:13, 134:22, 134:23, 137:8, 141:17, 154:9, 156:19, 164:6 leukemia - 53:22, 53:23, 54:4, 54:9 level - 11:1, 28:13, 66:21, 100:5, 127:21, 141:5, 143:1 levels - 142:16, 142:17, 154:4 librarians - 107:5 life - 37:6, 49:23, 51:21, 66:14, 66:17 lifestyle - 142:12 lifetime - 91:2 light - 28:2 likely - 34:10, 41:6, 72:5 limit - 76:1 limitation - 44:2, 77:9, 77:10, 77:14, 87:10, 97:14 limitations - 37:20, 43:11, 44:3, 77:11, 77:19, 77:23, 78:5, 101:7 limited - 43:20, 76:7 line - 52:23, 100:9, 143:17, 144:3 linear - 143:19, 143:23, 144:7 lines - 153:20 lingering - 162:3 lining - 90:20, 127:3 linings - 80:7, 98:3 link - 118:7, 118:10, 118:12 linked - 110:19, 111:23 listed - 114:22, 125:23 Listed- 114:23 lists - 50:8 Literally- 124:15 literature - 28:20, 29:13, 126:4, 126:8, 137:15, 139:11, 144:15, 152:11, 158:23, 159:2, 159:7, 159:9 Lithuania- 5:13, 5:20, 6:17 9 Litigation- 1:3 litigation - 19:19 lives - 45:13 living - 86:4, 102:20, 105:22 Lie- 2:9 Lip- 1:11, 2:2, 2:5 load - 12:20 Lobboilog- 151:6, 151:8 located - 24:9 logic - 31:19 logical - 46:15, 64:17 logistic - 127:9, 129:1 logs - 86:3 long-lasting 45:13, 45:15 look - 15:2, 20:6, 23:6, 23:8, 46:12, 47:11, 49:23, 54:11, 60:17, 61:5, 61:9, 71:2, 76:3, 76:8, 87:13, 103:1, 106:23, 126:19, 127:2, 127:12, 127:17, 128:11, 128:23, 133:15, 133:23, 134:13, 155:6, 158:4, 160:21, 164:3, 164:13 looked - 45:3, 83:7, 85:2, 104:19, 116:19, 149:5, 149:10, 153:16 looking - 14:1, 33:1, 47:14, 48:14, 58:9, 58:20, 60:2, 60:3, 91:9, 96:14, 96:21, 101:21, 146:15, 149:7, 155:5, 156:17 looks - 122:9, 153:17, 165:19 Lorimer- 151:13, 151:18 Los- 88:13 low - 28:13, 51:7 lower - 34:9, 50:16, 98:12, 154:1, 154:12, 155:19 lowest - 100:5, 142:21, 164:18 lunch - 112:23, 165:22 luncheon - 165:13 lung - 4:14, 4:19, 13:22, 15:21, 18:6, 18:9, 21:21, 22:22, 33:19, 33:20, 33:22, 34:10, 38:4, 39:6, 39:12, 40:7, 40:19, 40:21, 41:6, 49:23, 50:1, 51:5, 70:12, 81:13, 93:11, 93:12, 93:13, 118:15, 122:6, 123:16, 123:17, 123:18, 134:23, 135:15, 135:17, 135:22, 135:23, 136:3, 137:1, 137:2, 137:4, 137:11, 137:15, 137:19, 137:22, 138:7, 139:2, 139:13, 140:5, 141:5, 141:17, 144:11, 144:22, 145:5, 145:8, 145:14, 145:23, 149:4, 150:7, 156:7, 161:12 Lung- 145:16 lungs - 67:3, 76:21, 77:6, 81:12, 157:9 luxury - 12:22 lymphoblastic - 54:3, 54:9 Lynne- 166:2, 166:16 Lyon- 2:15, 24:9 M main - 87:10 maintenance 100:1 major - 11:3, 13:2, 44:2 majority - 23:23, 97:7, 107:7 Malignant - 89:9 managers - 100:3, 100:4, 101:3 manifests - 81:12 manner - 26:13, 27:17, 78:15, 144:8 manual - 72:4, 72:5 Marcus - 151:10, 151:11 Margaret - 2:17 Mark - 2:14 marriage - 164:7 marrow - 54:1 marry - 165:5 Masters - 7:20, 11:1 Masters-level - 11:1 match - 72:23, 73:9, 103:22, 104:1, 104:7, 113:23, 118:6 matched - 73:2, 79:6, 79:7, 89:6, 103:18 matching - 82:3 materials - 18:23 math - 32:19, 33:13, 38:5, 38:10, 147:14 mathematical 132:21, 149:1 mathematically 35:21, 35:23 matrix - 23:3 matter - 25:8, 28:3, 31:14, 31:15, 64:21, 157:15 Mcdonald - 57:18, 58:11, 75:8, 75:11, 80:13, 81:20, 85:22, 103:2 Mcdonalds - 78:11 Mcelvenny 119:17, 119:20 Mcguirewoods 2:2 Mclaughlin - 50:10, 50:14 Md - 3:6 mean - 18:17, 20:3, 25:19, 32:10, 45:12, 47:7, 47:22, 48:20, 49:7, 56:13, 60:1, 60:11, 61:11, 61:12, 62:9, 62:11, 62:12, 62:15, 64:23, 66:23, 80:1, 80:19, 81:15, 82:4, 85:21, 90:14, 93:23, 96:14, 96:21, 98:22, 100:19, 106:18, 109:15, 111:17, 115:18, 127:20, 130:11, 133:9, 135:18, 143:18 meaning - 19:2, 20:23, 26:20, 32:17, 50:6, 60:3, 105:8, 109:8, 147:8 meaningful - 83:16, 99:16, 99:17 means - 19:14, 19:20, 36:11, 37:4, 44:13, 48:22, 53:17, 60:23, 111:15, 154:12 measles - 8:8 measure - 43:8, 43:10, 57:3, 67:13, 69:6, 99:17, 141:4, 142:15, 156:22 measurements 140:18, 141:1, 141:3 measures - 69:9, 127:19 mechanic - 92:16, 93:21, 107:13, 107:17, 108:11, 125:7, 129:3, 136:16 mechanics - 3:19, 4:6, 15:22, 61:22, 62:1, 62:9, 62:21, 63:3, 63:19, 66:22, 67:5, 72:4, 75:5, 75:6, 75:19, 83:9, 90:21, 91:4, 92:14, 92:15, 93:19, 93:21, 94:7, 94:12, 95:11, 95:12, 95:20, 96:2, 98:1, 98:9, 99:8, 100:6, 104:12, 105:2, 105:4, 105:6, 105:10, 105:14, 105:20, 106:4, 107:1, 107:2, 107:4, 107:23, 109:2, 110:6, 114:17, 115:2, 115:23, 116:2, 121:23, 122:12, 122:13, 122:23, 123:1, 124:9, 124:19, 124:23, 126:19, 126:21, 127:6, 133:19, 134:14, 138:10, 138:17, 139:13, 140:9, 140:11, 144:12, 146:9, 146:10, 148:7, 150:15, 151:20, 152:12, 159:23, 161:14 Mechanics - 105:9 mechanism - 23:21, 26:5 mechanisms 53:11 mechanistic - 28:23 Medical - 13:3 medical - 4:3, 5:14, 5:15, 5:16, 5:17, 6:8, 6:16, 7:10, 9:5 medicine - 7:2, 7:9, 7:18, 8:1, 8:3, 8:18, 8:20, 10:22 Medicine - 11:22, 13:3 meeting - 16:3 Melissa - 2:14 members - 76:13 men - 90:15, 100:2, 100:4, 101:3 mention - 80:11, 98:6, 109:5 mentioned - 43:15, 44:18, 84:9, 89:20, 145:22, 161:5 mentioning - 152:3 Mesothelioma 136:17, 145:17 mesothelioma 4:7, 4:12, 15:21, 18:6, 22:22, 55:4, 55:6, 55:7, 55:20, 56:3, 59:9, 61:1, 62:2, 62:6, 62:10, 62:13, 62:22, 64:10, 66:3, 66:15, 67:1, 67:6, 67:11, 67:19, 67:21, 68:15, 69:5, 75:5, 76:15, 81:10, 83:11, 83:23, 89:9, 90:6, 92:15, 92:17, 95:21, 96:1, 97:5, 97:9, 98:4, 100:17, 100:18, 101:23, 102:13, 107:15, 110:6, 110:20, 111:23, 117:3, 119:7, 120:4, 120:7, 120:11, 120:22, 121:8, 121:23, 122:11, 123:22, 124:11, 125:3, 125:21, 130:18, 134:23, 135:6, 135:10, 136:1, 136:5, 136:10, 137:5, 137:6, 137:14, 138:2, 139:9, 141:6, 144:21, 145:6, 145:13, 145:19, 150:6, 151:3, 151:18, 156:7 Mesotheliomas 66:16 mesotheliomas 67:9, 68:1, 84:4, 90:12, 97:8, 113:15, 116:23, 123:19 message - 65:2 meta - 77:15, 129:10, 129:23, 130:3, 130:8, 132:1, 133:3, 134:5, 134:6, 134:9, 137:13 meta-analysis 77:15 metaanalysis 15:11,44:6 metastases - 77:6 metastasizing 76:21 method - 27:8, 27:11, 27:12, 27:19, 28:13, 47:23, 57:3 methodological 14:21 methods - 25:1, 49:3, 51:3 Michael - 2:13, 2:15, 3:5, 3:6 middle - 19:12, 50:19, 126:1 Miersen - 117:4, 117:6, 117:7, 117:21, 119:2 might - 31:12, 33:13, 40:5, 45:10, 49:18, 51:19, 61:14, 63:14, 63:19, 64:14, 68:16, 80:9, 82:12, 84:22, 90:12, 117:12, 127:1, 130:10, 159:16 Milham - 58:13, 61:11, 83:1, 115:10, 116:10, 116:11 millers - 159:21 million - 63:4, 63:10, 118:9 Mills - 52:5 mind - 57:8, 92:12, 151:21, 160:18 mineral - 21:22 mineralogy - 160:6 minerals - 163:4 miners - 100:6, 159:21 mining - 111:23, 112:3 minus - 148:1, 148:2, 148:4, 148:17, 155:9 minute - 18:21, 20:18, 49:13, 63:13, 98:10 minutes - 64:18, 102:2, 113:4 Miranda - 2:13 mishmash - 73:10 miss - 131:3, 145:10 missing - 32:3 mixed - 127:12 mixture - 25:11, 25:23, 26:4 model - 123:7, 127:8, 127:9, 129:2, 129:6, 147:7, 147:11, 147:13, 147:16, 147:17, 148:19 modern - 15:3, 28:18 molds - 147:23 molecular - 54:12, 164:6 moment - 67:1, 122:6, 137:17 monograph - 25:1 Monographs - 24:4 monographs - 25:7 month - 8:6 morning - 3:1, 3:11, 3:12, 64:15 morphologic 20:23 mortality - 41:10, 41:14, 41:19, 43:5, 43:10, 83:3, 83:8, 96:11, 98:4, 116:7, 116:9, 150:22 Most - 11:1, 23:17, 34:17, 78:8, 156:5, 160:21 most - 7:15, 9:14, 14:23, 23:12, 24:15, 26:7, 27:18, 40:4, 46:22, 48:14, 72:8, 85:9, 98:22, 119:19, 127:17, 133:12, 134:22, 135:13, 136:11, 141:3, 144:3, 155:19, 156:5, 160:3 mostly - 50:16 motor - 5:6, 15:22, 15:23, 62:7, 63:2, 63:3, 63:10, 90:21, 93:21, 94:7, 98:1, 98:9, 100:5, 104:12, 105:5, 107:13, 114:17, 122:12, 124:19, 126:20, 130:19, 133:18, 134:2, 134:17, 139:13, 161:14 motor-vehicle - 5:6, 15:22, 15:23, 62:7, 63:2, 63:10 Mount - 44:19, 151:12 move - 6:20, 7:1, 28:18, 34:1, 37:1, 10 55:12, 65:12, 88:21, 88:22, 99:23, 100:2, 108:8, 118:8, 133:1, 164:1, 164:9 moved - 5:21, 65:2, 65:6 movie - 46:5, 46:9, 46:19 moving - 124:8 muddled - 117:12 multidisciplinary 10:20, 16:11 multiple - 24:10, 28:12, 91:2, 112:15, 112:16, 131:7, 136:13, 137:11, 137:12, 152:3 multiplicative 145:23, 146:8, 147:15, 147:18, 148:18 municipal - 140:12 murder - 46:8, 46:12 must - 30:5, 31:18, 136:19, 160:20, 162:17 N narrow - 22:21, 54:10 narrowly - 54:15 National - 14:12, 87:23, 89:21, 126:10 naturally - 75:3 nature - 11:12 Nci - 88:3, 120:21, 126:6 Neal - 2:15 necessarily - 9:10, 20:1, 43:8, 45:11, 99:16, 133:4, 133:7 necessary - 70:2 need - 18:18, 27:7, 28:2, 34:16, 36:5, 40:23, 45:23, 72:10, 100:8, 141:20, 141:21 needs - 44:1,80:12, 87:17, 101:6, 154:16, 155:6 negative - 33:15, 36:2, 100:10, 100:12, 109:17 nested - 142:1 Nested - 142:8 never - 19:22, 30:9, 30:10, 32:22, 66:23, 79:2, 86:6, 105:6, 105:8, 107:4, 114:8, 114:22, 124:1, 127:16, 139:5, 150:1, 157:17 nevertheless 104:5 new - 112:11 New - 1:2, 44:19, 88:13, 166:9 newborn - 67:1, 67:2 newborns - 66:16 newly - 45:6, 118:4 Next - 81:22, 113:10 next - 24:2, 24:3, 30:15, 31:17, 35:7, 60:8, 61:5, 64:23, 74:10, 81:4, 81:13, 81:18, 82:19, 82:20, 83:20, 83:21, 87:20, 88:18, 89:22, 89:23, 91:20, 95:15, 96:6, 99:1, 101:9, 101:11, 103:9, 103:13, 103:18, 104:2, 104:5, 113:8, 117:17, 131:23, 138:21, 143:6, 144:19, 154:15, 157:21, 161:3 next-of-kin - 81:4, 88:18, 103:13, 103:18 Next-of-kin - 81:22 nice - 89:12 Nicholson - 44:18, 152:1, 152:7, 152:20, 153:5, 158:10 nine - 116:3, 148:6 Niosh - 44:20, 116:6, 116:7, 116:8, 116:12, 116:18, 158:10 nitpick - 110:11 Nobody - 45:8 nobody - 46:15, 80:1 non - 34:3, 34:11, 67:1, 67:23, 71:4, 82:13, 90:7, 94:12, 95:1, 97:9, 105:15, 107:23, 111:14, 143:19, 149:1 non-association 111:14 non-coffee - 71:4 non-differential 82:13 non-diseased 107:23 non-exposed - 90:7 non-mathematical 149:1 non-mesothelioma - 97:9 non-occupational 67:23 non-significant 94:12, 95:1, 105:15 non-smokers 34:3, 34:11 non-threshold 143:19 non-zero - 67:1 nonasbestos 111:23, 112:3 nonconfounders 137:6 nondrinkers - 52:3 none - 46:21, 123:5, 129:3 None - 32:23 nonexperimental 48:5 nonexposed 32:18, 38:17, 121:11 nonsignificant 52:16, 117:1 nonsmoker 149:10 nonsmokers - 40:8, 73:11, 74:12, 74:17, 138:18, 149:6 Nonsmokers 74:19 nonsmoking - 73:3, 73:4 nonspecific - 156:9 normal - 111:16 north - 113:14 notable - 97:23 note - 51:4 noted - 2:21, 77:11, 77:16, 77:17 nothing - 83:12, 83:13, 158:1 notice - 65:15, 65:16 notion - 59:15 number - 6:13, 30:5, 32:18, 32:20, 33:9, 34:6, 40:12, 41:20, 43:1, 43:4, 44:15, 51:23, 52:8, 52:9, 52:17, 63:9, 63:21, 63:22, 63:23, 68:19, 73:18, 79:15, 85:17, 89:10, 89:11, 89:12, 89:13, 89:16, 95:1, 97:4, 97:15, 100:19, 116:22, 119:5, 121:7, 121:9, 121:16, 131:19, 138:23, 155:6, 157:12 numbers - 31:2, 32:23, 50:12, 52:10, 58:1, 74:6, 74:12, 74:13, 99:10, 104:17, 106:3, 106:8, 109:13, 109:18, 109:22, 110:7, 112:1, 115:3, 115:5, 121:3, 121:4, 121:7, 141:19, 153:19, 157:14 nuts - 55:12 nuts-and-bolts 55:12 O o'clock - 165:10 observation 27:14, 27:20, 29:6, 115:7, 125:12 observational 27:19, 48:5, 48:8 observations - 29:7 observe - 27:23, 36:13, 62:13, 68:8, 71:1 observed - 32:12, 33:8, 43:4, 43:6, 52:7, 52:8, 68:10, 68:11, 68:22, 74:20, 77:15, 92:14, 97:3, 97:15, 116:3, 116:4, 116:23, 119:5, 154:8 observes - 29:14, 29:16 observing - 33:20 obtain - 8:17, 11:20, 12:18 obviously - 22:23, 84:14 Occupation - 63:15 occupation - 4:22, 42:4, 43:23, 56:6, 56:10, 66:13, 78:16, 86:5, 98:14, 100:1, 105:3, 111:21, 157:7, 160:21 occupation's 105:4 occupational 15:11, 43:14, 56:5, 63:1, 63:11, 63:13, 63:16, 64:3, 66:10, 67:20, 67:22, 67:23, 68:4, 75:23, 76:8, 78:13, 83:7, 85:3, 86:10, 90:16, 92:2, 107:18, 108:1, 116:9, 121:6, 127:15, 128:19, 129:8, 138:11, 142:10, 164:10 Occupational 16:20 occupationally 87:3 occupations - 57:5, 62:23, 72:7, 80:23, 90:2, 91:16, 99:18, 100:11, 100:15, 100:19, 102:10, 107:7, 110:18, 114:23, 115:1, 118:18, 138:21 occur - 29:19, 66:11, 67:9, 67:20, 67:21, 68:1, 75:2 Occurred - 42:20 occurred - 33:9, 66:5, 157:5 occurs - 53:14, 132:9, 164:23 October - 1:18, 2:18, 166:13 odd - 79:17 odds - 40:17, 40:18, 40:19, 40:20, 40:22, 74:5, 85:12, 85:13, 90:19, 94:11, 94:15, 102:9, 105:14, 106:8, 121:13 offered - 155:23 office - 11:23, 100:4, 101:3 Office- 166:11 Official - 1:20, 166:5, 166:17, 166:19, 166:21 often - 29:2, 56:6, 160:23 Oi- 2:7 old - 54:10, 98:16, 139:14 older - 70:14 Olsen - 117:9, 117:16, 117:22, 117:23, 118:22 Once- 109:18 once - 7:22, 18:16, 54:11, 107:11, 130:23 One - 23:6, 34:2, 41:7, 42:17, 68:22, 76:2, 81:11, 88:11, 93:8, 93:20, 94:6, 115:16, 123:22, 127:2, 130:4, 132:19, 145:3, 156:17, 157:10 one - 9:14, 9:16, 10:1, 12:22, 15:9, 17:7, 19:5, 20:20, 21:15, 23:8, 24:10, 24:11, 25:11, 25:15, 28:7, 28:8, 28:9, 28:10, 29:15, 31:10, 32:5, 32:12, 32:13, 33:10, 33:11, 34:7, 35:9, 36:16, 37:14, 37:15, 39:13, 40:20, 40:21, 41:3, 43:19, 46:18, 50:2, 50:4, 50:5, 50:7, 50:11, 51:12, 53:7, 53:20, 54:5, 55:6, 56:8, 56:20, 58:3, 58:4, 59:12, 60:10, 61:5, 64:12, 67:18, 69:10, 72:18, 74:4, 74:16, 75:18, 76:2, 76:3, 76:9, 76:17, 78:14, 81:1, 81:6, 82:5, 85:7, 85:14, 86:15, 86:22, 87:13, 88:1, 89:22, 92:16, 93:10, 94:5, 94:10, 95:13, 96:16, 97:17, 101:6, 101:9, 101:10, 103:5, 104:19, 104:20, 107:9, 107:12, 107:15, 108:3, 108:8, 108:19, 108:22, 109:10, 111:8, 112:16, 113:10, 113:14, 119:10, 120:20, 122:6, 122:9, 122:14, 124:2, 124:13, 125:3, 125:10, 125:11, 125:12, 127:3, 128:11, 129:20, 131:7, 131:12, 132:11, 132:13, 133:6, 133:9, 134:5, 134:10, 134:12, 134:16, 134:18, 135:11, 135:14, 139:8, 139:14, 139:16, 139:22, 140:3, 142:8, 142:23, 143:5, 144:19, 144:22, 145:11, 145:18, 148:1, 148:2, 148:4, 148:6, 148:14, 148:15, 148:17, 148:19, 148:20, 148:23, 149:16, 151:2, 151:5, 151:8, 151:10, 151:15, 154:16, 155:6, 155:7, 156:10, 157:13, 158:3, 159:12, 159:13, 163:17, 163:22 ones - 42:21, 48:5, 114:23, 145:10, 165:2 ongoing - 14:10, 18:1 onset - 45:7, 45:14, 157:22 opening - 65:5 operate - 159:20 opinion - 4:5, 4:8, 4:9, 29:23, 43:12, 68:6, 68:7, 87:9, 133:2, 133:11, 135:10, 135:21, 152:11, 157:15, 160:10 opinions - 134:20 opportunity - 127:1 opposed - 43:18, 45:4, 134:2, 134:17, 150:20 opposite - 154:23 options - 165:20 order - 82:20 Organization - 8:7, 24:9 orient - 49:13 originally - 5:12, 64:22, 161:21 Ossiander - 57:20, 58:12, 58:13, 61:11, 115:10, 116:10 otherwise - 148:21, 148:22, 149:10, 162:4 ought - 144:7 outcome - 47:13 outcomes - 23:22 outlier - 61:11 11 outliers - 53:7 outside - 160:1 overall - 52:18, 135:6, 144:5 overhead - 73:23 overlap - 155:5, 155:7, 158:5 overlapping 155:10, 155:17 overload - 129:4 oversimplified 32:19 overview - 55:23 overwhelmingly 100:14 P p-value - 154:7 P-value - 154:7, 154:9 pack - 50:2, 139:6 Pads - 161:8 page - 129:19 pancreatic - 29:15, 29:16, 29:20, 30:17, 31:2, 33:17, 34:22, 35:3, 35:9, 52:1, 70:4, 70:10, 70:21, 71:2, 71:4, 73:1, 76:2 Pancreatic - 70:8 panel - 21:12 panels - 24:15 paper - 15:6, 15:7, 15:8, 15:14, 17:3, 20:1, 44:9, 77:15, 77:22, 78:4, 92:13, 92:21, 93:10, 109:1, 110:13, 110:17, 114:9, 114:15, 119:2, 129:10, 132:18, 144:19, 146:18, 152:2, 160:2, 160:3, 160:4 papers - 15:20, 88:5, 89:21, 152:10 pare - 136:19 parenchymal 153:21 part - 8:12, 13:2, 21:14, 24:8, 26:8, 39:22, 69:8, 77:22, 78:8, 87:18, 126:1 participants 155:20, 155:21 participate - 8:10 particular - 14:15, 24:17, 31:18, 32:11, 37:15, 38:23, 41:20, 45:3, 55:20, 72:8, 80:21, 85:3, 87:2, 89:5, 105:13, 114:10, 116:4, 123:8, 139:23, 144:2, 145:19, 156:1, 158:12, 159:17 particularly - 15:3, 20:7, 62:23, 70:14, 80:12, 136:14, 136:15, 140:15 parts - 22:13, 58:9 pass - 6:13, 6:14 past - 38:19, 40:4, 40:5, 64:19 pathologist - 55:9 pathology - 76:14 patients - 29:9, 29:17, 39:7, 71:5, 88:17, 93:11 Patricia - 166:4, 166:21 Patrick - 17:4 pattern - 157:2, 157:4 Pause - 141:14, 141:22, 153:15 pay - 118:17 Pc - 1:9 pedantic - 37:1 pediatric - 14:7 pediatrician - 6:1, 6:17, 10:4, 10:5, 10:6 Pediatrics - 13:4 pediatrics - 5:19, 5:22, 6:19, 6:22, 6:23, 8:22, 9:2 peer - 21:12, 77:19 peer-review - 77:19 pension - 118:13 people - 4:6, 9:4, 10:21, 11:10, 13:4, 15:5, 15:22, 16:10, 29:17, 31:4, 31:12, 31:14, 31:15, 32:1, 32:6, 32:20, 33:21, 35:4, 35:8, 35:16, 37:5, 38:19, 38:20, 38:21, 40:6, 40:9, 40:13, 40:14, 41:5, 41:6, 41:19, 42:6, 43:9, 43:16, 43:18, 45:5, 49:8, 52:7, 54:11, 56:2, 56:11, 57:13, 58:8, 58:21, 59:2, 62:6, 62:13, 63:1, 63:4, 63:7, 63:9, 63:14, 63:19, 64:2, 68:4, 68:14, 69:2, 69:7, 70:19, 71:2, 71:5, 72:6, 74:3, 74:8, 76:20, 77:5, 78:1, 78:19, 79:13, 81:8, 81:9, 85:22, 86:3, 86:21, 88:23, 90:9, 90:14, 91:4, 93:12, 93:18, 93:19, 94:8, 95:2, 98:22, 99:13, 100:20, 102:19, 103:8, 104:8, 104:9, 105:20, 107:3, 107:22, 109:10, 112:23, 114:4, 114:21, 121:5, 122:11, 122:12, 127:6, 127:12, 127:13, 127:22, 128:1, 128:6, 128:13, 133:15, 134:15, 136:6, 136:20, 138:15, 138:19, 138:22, 140:11, 140:13, 141:16, 141:17, 142:4, 142:7, 142:9, 146:23, 147:3, 147:4, 147:20, 149:20, 149:23, 150:1, 151:16, 152:13, 153:18, 153:20, 153:23, 155:1, 155:2, 155:14, 155:15, 156:18, 157:5, 157:8, 160:21, 161:14, 163:6, 164:16 People - 36:14, 38:14, 38:18, 39:10, 77:5, 88:19, 107:6, 118:8, 127:5, 149:20 Pepple - 2:7 per - 32:12, 32:13, 32:14, 32:16, 50:2, 138:13, 138:17, 138:19, 140:19 percent - 13:11, 13:13, 13:14, 35:6, 35:8, 36:14, 37:9, 37:21, 42:14, 43:2, 43:4, 50:23, 67:16, 68:5, 90:15, 91:3, 91:5, 91:6, 91:18, 104:1, 104:2, 104:4, 126:20, 126:22, 138:15, 138:22, 153:21, 153:23, 155:1, 155:2, 157:15 percentage - 67:8, 68:13, 68:20, 68:23, 90:12, 91:15, 155:20, 155:21, 156:12 perfect - 81:14, 102:21 perfectly - 131:12, 136:23 performed - 44:5, 70:6, 91:18, 129:7 Perhaps - 9:18 perhaps - 106:6, 157:10, 162:10, 165:5 period - 41:21, 66:17, 75:17, 86:12, 96:22 Peritoneal - 120:6 peritoneal - 83:11, 120:4, 120:10 Perry - 2:14 persistence - 21:21 Person - 118:5 person - 32:16, 45:10, 45:13, 58:7, 66:13, 67:2, 78:15, 79:1, 86:8, 98:15, 103:7, 107:13, 109:9, 114:7, 125:2, 149:9 person's - 91:2, 149:4 personally - 10:1 persons - 4:10, 4:18, 5:5, 125:9 perspective - 60:12 pertained - 115:22 pertaining - 104:11 pertinent - 85:9, 114:16 Petersen - 116:11 Peterson - 82:23 Ph.d - 11:1 Ph.d.s - 11:9 phase - 123:20 phone - 86:5 phrase - 60:11 phrased - 135:20 physical - 21:1 physician - 6:12, 29:7, 29:9, 29:14 Physician - 7:13 physicians - 11:7, 62:12 physics - 11:5 Pi - 41:16 piano - 78:18, 78:23 pick - 93:11, 148:19, 159:9 picture - 52:23 piece - 93:12 pioneers - 19:5 pipefitters - 61:18 Pir - 41:17, 117:23, 118:1, 119:6 place - 96:11, 101:23 placed - 25:23 places - 49:4 Plaintiff - 1:9, 1:11, 1:13, 1:15 plants - 27:23, 28:1 plausibility - 164:3, 164:13 play - 34:23 plays - 36:12 pleura - 120:8 pleural - 83:10, 97:6, 97:7, 97:9, 102:13, 161:11, 161:12 Pllc - 2:7 plotted - 122:1 plug - 121:7 plumbers - 61:18, 99:19 plus - 142:19, 143:7, 147:16, 148:1, 148:3, 148:4, 148:17, 155:9, 157:16 Pm - 165:23 Pmr - 41:9, 41:14, 41:17, 43:5, 83:2, 83:18, 96:10, 97:16, 98:6, 98:8, 99:4, 99:14, 99:16, 100:5, 101:5, 115:15, 115:19, 116:5, 117:1, 118:1, 120:1, 122:18, 150:21 Pmrs - 150:23 point - 28:15, 31:4, 45:9, 46:19, 51:8, 64:18, 69:21, 70:10, 82:13, 88:10, 88:20, 93:3, 111:18, 118:22, 119:6, 125:6, 135:13, 147:7, 154:11, 158:3, 162:4, 164:17, 165:20 pointed - 135:7 pointer - 49:16 points - 17:11 polio - 8:8 political - 6:6 pooled - 132:1 population - 39:9, 40:11, 42:1, 42:9, 43:2, 43:9, 43:16, 43:17, 52:2, 75:3, 77:3, 77:4, 77:7, 81:7, 84:3, 88:12, 88:20, 93:8, 97:16, 101:22, 102:8, 103:3, 113:16, 113:21, 113:23, 114:1, 114:5, 114:12, 116:4, 118:9, 118:14, 164:8 population-based 84:3, 88:12, 101:22, 113:16 populations 22:11, 77:2 pose - 115:1 posed - 60:6, 162:6 positive - 33:14, 38:6, 70:3, 100:9, 100:11, 111:2 possible - 17:11, 39:20, 72:11, 100:8, 142:8, 157:10, 157:18 possibly - 128:16 post - 100:1 poster - 16:5, 16:6 Potential - 161:7 potential - 20:19, 23:8, 23:13, 24:12, 48:15, 138:6 potentially - 59:2, 80:9, 137:2, 156:17 Potter - 2:5 powerful - 72:8, 78:21, 80:22 practice - 7:23, 29:16, 164:6 practiced - 5:19, 11:16 preamble - 25:7 preceded - 46:21, 47:13 precedes - 86:8 precise - 134:11 precisely - 110:10, 124:5 precision - 52:18 predict - 23:22 Predominantly 31:21 premise - 38:13 prepared - 3:21 presence - 28:1, 49:8, 75:1, 98:2, 163:22 Present - 2:11, 2:20 present - 16:1, 21:7, 47:11, 52:23, 59:10, 116:21, 122:1, 129:6, 132:18 presentation 16:14, 139:11 presented - 16:5, 88:7, 110:1, 115:3, 121:15, 158:10 pressure - 45:8, 45:10, 45:11, 45:12 pretty - 54:10, 57:23, 61:19, 77:12, 78:4, 86:14, 126:18, 156:9 prevalence - 138:14 prevalent - 138:22 Preventive - 11:22 preventive - 7:2, 7:9, 7:18, 7:23, 8:3, 8:18, 8:20 previous - 91:6, 134:20, 138:16 previously - 21:5, 110:19, 111:22, 116:10 Price - 2:4 primarily - 12:18, 123:16, 151:17 printed - 73:22 priori - 115:1, 115:2, 145:3 Privately - 12:7 probability - 4:3, 51:4, 108:6, 108:9, 154:8 problem - 47:8, 81:4, 81:5, 103:11, 104:3, 106:11, 121:2, 136:5, 156:23, 157:12, 159:18, 160:12 problematic 98:17, 98:18 proceeding - 26:12, 84:12 proceedings 166:8 process - 77:19, 130:7, 130:9 product - 41:9 production - 99:22, 138:20 products - 66:4, 71:12, 146:14 Professional - 12 166:4 professionally - 63:9 professor - 12:15 proffer - 162:1 prognosis - 54:17 program - 5:18, 7:2, 7:18, 8:5, 12:16 projects - 12:19, 13:18 promised - 149:12 prone - 57:1 proportion - 41:14, 67:23, 69:2, 90:9, 103:15, 104:9, 150:21, 154:13 proportionate 41:10, 41:15, 43:5, 43:10, 83:3, 83:8, 96:11 proportions 35:16, 41:18, 41:19, 68:3, 104:8 prospectively 46:16, 142:4 prostate - 13:22, 14:6 Prostate - 14:7, 136:23 protections - 100:1 protective - 79:21 Prothonotary 166:11 provide - 99:10, 123:6, 142:18 provides - 20:2, 24:18 Public - 3:17, 7:5, 7:8, 7:19, 7:20 publication - 15:8, 20:10, 50:9, 77:19, 88:4 publications 15:10 Publicly - 10:17 publish - 92:7 published - 16:16, 16:18, 17:1, 82:21, 88:5, 89:23, 92:6, 92:20, 114:9, 118:1, 119:1, 119:8, 119:11, 119:18, 150:14 pull - 42:15 pumps - 140:18 pure - 73:15 purple - 122:17 pursuant - 44:20 pushing - 112:19 put - 21:13, 28:5, 28:7, 30:9, 32:9, 34:4, 35:15, 52:20, 86:11, 110:14, 115:5, 131:13, 153:12, 158:17, 159:4, 165:12 Put - 73:23 puts - 66:13, 68:19, 92:16 puzzling - 160:20 Q qualifications 65:1 qualify - 6:15 qualitative - 130:4 quality - 131:20, 133:23 quantified - 36:13 questionable 159:22 questioning - 90:22 questionnaire 120:16, 120:20, 121:1 Questions - 158:6 questions - 26:11, 55:13, 60:8, 105:22, 126:9, 129:13, 161:23, 162:3, 162:5, 165:8 quick - 8:21, 27:22, 55:22, 155:4 quickly - 29:4, 33:12, 113:11, 120:13, 128:5, 129:15, 148:7 quite - 11:20, 29:14, 54:12, 83:6, 114:8, 136:22 quote - 19:4, 80:3, 112:11, 124:5 R raised - 30:1, 92:10, 93:17, 98:2 ran - 50:17 random - 36:10, 36:13, 39:13 Random- 36:11 randomly - 52:21, 113:21 range - 67:12, 67:15 rank - 78:17, 99:18, 138:23 ranked - 99:14, 138:16 ranks - 138:11 Rare- 89:10 rare - 32:20, 64:4, 75:2, 120:10, 120:12, 120:23, 136:5, 136:8, 136:11, 136:15, 136:17, 136:19 rarely - 61:17, 76:6 rate - 69:4, 69:12, 69:14 rather - 23:18, 119:20, 151:20 ratio - 33:1, 33:10, 40:17, 40:18, 40:21, 40:22, 41:2, 43:6, 74:5, 79:17, 83:3, 85:12, 85:13, 90:19, 94:12, 96:11, 105:14, 106:8, 121:13 ratios - 83:8, 94:16, 102:9 raw - 162:2 ray - 45:7, 151:20, 153:18, 156:9, 156:13 Rdr- 166:16 Re- 1:3 reach - 6:14, 30:5 reached - 3:18, 4:13, 4:16, 4:23 reaching - 64:8 read - 152:2 reader - 99:12 ready - 55:15, 59:7 Ready- 55:16 real - 37:6, 49:22, 49:23, 51:15, 51:21, 155:8, 155:18 real-life - 51:21 real-world - 51:15 really - 10:20, 11:18, 27:5, 45:8, 58:22, 61:19, 80:15, 80:22, 85:18, 87:11, 88:8, 95:12, 106:18, 109:19, 114:15, 118:8, 123:17, 124:14, 136:7, 154:20 Realtime- 166:3, 166:5 Reardon- 2:14 reason - 18:8, 22:3, 61:15, 72:15, 83:1, 88:16, 93:13, 124:18, 135:18, 147:14, 163:13, 163:14 reasonable - 4:2, 131:12, 159:14 reasons - 39:14, 151:16 recalculate 157:11, 157:18, 157:19 receiving - 13:4 recent - 26:7, 119:19, 160:3 recently - 160:19 Recess- 113:6 recess - 64:18, 65:20, 113:4, 165:13, 165:21, 165:22 reconcile - 62:19 reconstruct - 86:9 record - 13:16, 45:1, 48:7, 86:15, 92:6, 117:12, 139:10, 166:11 recorded - 124:19 records - 118:10 recruit - 30:18 recruiting - 142:7 red - 135:14 Reduction- 161:6 Reese- 2:12 reference - 114:19, 114:20, 143:1 references - 9:13, 109:23 refugee - 6:6 regard - 71:10 regarding - 128:15 regardless - 85:2, 128:13, 148:18 regards - 59:21 region - 96:23, 99:5 regional - 24:14 Registered- 166:2, 166:4 registries - 14:3, 87:7, 102:20, 116:12 Registry- 84:5, 84:11 registry - 35:2, 35:3, 84:6, 84:7, 88:11, 88:13, 113:17, 118:2, 118:5, 118:14 registry-based 88:11, 113:17 regression - 127:9, 129:1 regrettably - 162:23 regular - 34:2, 34:7, 129:1, 150:18 regularly - 35:6 regulators - 16:12 relate - 15:21 related - 3:19, 8:16, 9:10, 11:5, 14:19, 16:23, 18:5, 18:9, 20:10, 22:11, 22:19, 42:17, 64:21, 85:5, 85:11, 90:12, 127:7, 133:21, 151:16 relates - 14:21, 108:16, 162:15 relationship 17:20, 128:10, 138:6 relative - 43:8, 43:13, 50:1, 50:3, 52:1, 52:15, 57:17, 57:19, 57:20, 57:21, 58:3, 59:18, 59:22, 73:14, 74:18, 74:19, 79:12, 85:12, 90:7, 90:19, 100:18, 102:9, 109:21, 115:6, 119:6, 123:6, 130:21, 130:22, 131:21, 134:5, 139:12, 142:23, 143:8, 145:5, 145:8, 145:11, 145:19, 147:2, 147:19, 147:21, 147:23, 148:1, 148:2, 148:3, 148:10, 148:12, 148:17 Relative- 70:11 relatively - 109:21 relevance - 159:21, 160:8 relevant - 26:5, 48:23, 131:1, 131:2, 159:10, 159:19, 160:11, 160:14, 161:6, 161:9, 161:13, 161:15, 164:14 reliable - 43:12 relied - 117:18 relying - 86:22 remained - 126:9 remains - 166:10 remember - 16:4, 44:21, 108:6, 151:1 removal - 109:3 removed - 89:2, 93:13 removing - 79:8, 79:11, 79:13, 80:22 repair - 4:11, 4:19, 5:6, 15:23, 61:7, 61:23, 62:7, 63:2, 63:10, 85:4, 85:11, 90:21, 91:4, 91:12, 93:18, 93:23, 94:2, 94:15, 95:3, 106:2, 106:5, 108:18, 115:23, 116:2, 121:10, 122:13, 124:9, 126:14, 126:21, 127:4, 127:13, 130:19, 133:16, 152:13 repairman - 83:10 repairs - 95:22, 134:3 repeat - 91:8 repeated - 48:2, 48:3, 48:4 replaced - 164:22 report - 21:14, 29:4, 30:13, 76:15, 92:18, 106:4, 121:4, 121:16, 121:17, 144:8, 151:18, 152:20, 154:4, 154:7 reported - 111:4, 111:5, 112:12, 118:4, 145:4, 145:7, 146:22, 151:15, 151:20, 152:7, 154:6, 164:12, 166:8 Reporter- 166:1, 166:2, 166:3, 166:4, 166:5, 166:17, 166:19, 166:21 reporter - 41:1, 151:7 Reporters- 1:20, 166:5 reports - 11:19, 29:2, 29:6, 29:9, 130:20 represent - 100:12 representative 39:9, 40:11, 77:3, 77:4, 77:7, 81:6 reproduced - 28:12 reps - 100:3, 101:2 require - 54:18, 135:16, 150:12 required - 54:17 requires - 48:1, 135:11 Research- 13:3, 17:23, 24:6, 88:7 research - 10:15, 12:7, 12:16, 12:19, 13:18, 14:10, 14:15, 15:18, 16:2, 16:16, 16:23, 19:6, 21:10, 23:3, 23:10, 23:11, 24:14, 27:4, 27:6, 27:16, 28:15, 28:19, 30:2, 36:5, 48:6, 53:1, 129:16, 146:20 researcher - 29:8, 37:12, 68:19 researchers - 18:3, 49:5, 51:3, 58:20, 75:4, 91:23, 108:15, 159:15 researching - 21:7 resection - 93:11, 103:3 reservations - 78:2 residence - 86:4 residency - 6:18, 7:3 resident - 7:9, 12:1 residents - 11:22 respect - 49:14, 53:9, 54:16, 60:18, 93:16, 99:14, 105:18, 114:13, 135:22, 137:21, 138:12, 138:13, 139:23, 142:11, 146:17, 159:22 respiratory - 89:2 respond - 54:18 response - 122:15, 122:22, 127:21, 143:18 responsibilities 12:14 responsible - 12:22 responsive 128:10, 128:11, 128:12, 141:11, 143:19, 144:1 result - 36:9, 36:12, 36:18, 37:6, 37:18, 37:22, 51:1, 73:16, 85:9, 87:14, 94:23, 109:15, 125:11, 125:12, 131:22, 133:3, 135:17, 145:13 resulted - 125:10 results - 16:1, 16:5, 34:5, 35:15, 37:13, 49:12, 80:11, 85:5, 94:9, 95:20, 99:11, 108:6, 108:10, 108:23, 109:1, 114:16, 114:17, 13 121:22, 122:9, 146:9, 154:11, 162:21 Results- 84:10, 120:1 reverse - 34:12, 136:18 review - 21:12, 44:5, 77:19, 129:7, 130:9, 130:10, 130:14, 137:14, 159:7, 160:5, 160:10, 160:19 reviewed - 159:1, 162:2 reviewing - 19:18, 126:8 reviews - 144:15 risk - 3:19, 4:7, 4:11, 4:14, 4:19, 5:1, 5:6, 11:11, 15:21, 16:10, 18:17, 23:13, 24:19, 29:20, 33:15, 35:20, 37:5, 43:9, 43:13, 50:1, 50:4, 50:6, 52:1, 52:6, 52:15, 53:6, 56:2, 56:4, 57:11, 57:17, 57:19, 57:20, 57:21, 58:3, 58:4, 59:9, 59:18, 59:22, 60:7, 60:13, 60:23, 61:12, 62:6, 62:18, 66:13, 66:14, 66:15, 66:17, 66:19, 66:20, 66:22, 67:1, 70:5, 70:11, 70:12, 73:14, 74:7, 74:18, 74:19, 78:16, 79:12, 80:5, 80:9, 85:13, 87:15, 87:16, 90:7, 90:8, 90:19, 91:15, 92:17, 93:18, 95:21, 96:2, 98:1, 111:3, 111:5, 111:10, 111:12, 111:13, 115:2, 115:6, 119:6, 122:10, 123:6, 124:1, 127:3, 127:23, 128:7, 130:21, 130:22, 132:2, 134:5, 136:8, 137:18, 137:22, 138:1, 139:12, 141:10, 141:17, 142:23, 143:8, 143:11, 143:13, 144:12, 144:21, 145:5, 145:8, 145:11, 145:19, 146:16, 147:2, 147:9, 147:19, 147:21, 147:22, 147:23, 148:1, 148:2, 148:3, 148:9, 148:10, 148:12, 148:17, 148:20, 149:4, 149:7, 149:9, 150:6, 150:7, 150:8, 150:16, 152:12, 152:18, 156:6, 156:7, 158:12, 160:15, 160:23, 163:11, 163:15 Risk- 16:7, 16:9 risks - 19:1, 19:3, 102:9, 149:19 Rodelsperger91:21, 91:23, 92:2, 95:8 Roggli- 161:10, 162:15, 164:16 role - 36:12 roll - 40:9 rolling - 39:7 room - 39:8, 39:14, 147:5, 150:12 rotating - 113:3 rotation - 8:13, 12:1 Rothman- 9:13, 9:15, 9:16 roughly - 50:15, 51:1, 52:12, 125:2, 145:10 round - 34:7 rounded - 31:3 rounding - 136:20 row - 133:6, 133:13, 155:20 rows - 133:4, 133:5 Rpr- 166:19 Rr- 134:5 rubric - 26:16 rule - 23:18 run - 29:3, 92:2, 142:7 running - 165:19 rural - 5:20 S sake - 125:8 sales - 100:2, 101:2 sample - 77:5, 87:11, 89:15, 89:19, 113:20, 134:7 sampling - 142:7 Samuel - 2:2 saw - 101:1, 104:1 scale - 57:16 scattered - 52:21 scenarios - 110:18, 111:22 scheduling - 65:9 schematic - 70:7 scheme - 99:18 Schmidt - 2:7 school - 5:15, 5:17, 7:21, 27:13, 27:23, 28:13, 28:18, 46:14, 48:1 School - 3:17, 7:5, 7:8 science - 11:6, 22:9, 30:4, 164:8, 165:5 sciences - 10:23 scientific - 4:3, 6:9, 16:2, 27:8, 27:11, 27:18, 28:13, 47:23, 61:3, 158:23 Scientific - 27:12 scientists - 21:6 score - 131:20 scoring - 133:11 Scott - 2:12 scratch - 162:22 scratching - 24:1 screen - 15:6, 17:6, 21:11, 29:5, 49:17, 80:4, 82:21, 83:1, 83:2, 83:12, 92:4, 115:4, 121:20 Seattle - 30:22 second - 7:2, 7:21, 16:15, 26:8, 43:19, 67:17, 124:2, 129:19, 133:13, 141:20, 153:4, 157:10 section - 77:11, 77:22, 78:9 sectional - 45:17, 45:18, 45:19, 46:3, 46:17, 46:22, 47:1, 47:3, 47:4, 47:7, 150:19, 151:4, 152:15 sections - 161:1 security - 99:23 see - 30:1, 46:12, 49:1, 50:16, 52:4, 64:5, 72:17, 93:5, 112:14, 113:3, 113:5, 116:16, 118:7, 134:6, 135:14, 150:6, 150:7, 155:13, 160:23 seeing - 136:22, 151:1 Seelaus - 2:14 Seerr - 84:9, 84:10 selected - 88:14, 93:9, 102:19, 113:20, 113:22, 145:2 selection - 39:21, 64:23, 65:1, 65:5, 114:8 selects - 116:12 Selikoff - 19:5, 19:6, 19:10, 125:14, 125:15, 125:20 sense - 22:6, 112:22 sensitive - 135:13 sent - 126:12 separately - 149:7, 149:10 separating - 149:6 September - 102:15 series - 8:15, 29:7, 92:7 service - 85:11, 100:1 Services - 8:14 services - 85:5 Servicing - 161:8 Session - 1:18 set - 27:16, 73:10, 82:1, 89:22, 90:4, 92:5, 101:20, 119:3, 119:23, 122:9, 126:6, 126:11, 130:3, 130:16, 131:18, 133:7, 133:8, 140:17, 140:18, 150:3, 159:16, 159:17, 160:21 sets - 94:11 setting - 102:7 settings - 158:19, 159:5 seven - 13:18, 13:20, 116:2, 116:4 seventies - 51:12, 151:19 several - 10:21, 12:21, 17:23, 63:4, 65:16, 67:18, 75:21, 76:8, 93:20, 132:2, 136:21, 149:17, 151:5, 153:17 Several - 110:18, 111:21 sex - 114:6 shade - 75:5 share - 3:21 ship - 91:17, 126:21, 149:23, 150:1, 150:2 Shipyard - 71:15 shipyard - 60:17, 60:18, 60:22, 72:3, 78:18, 79:1, 79:2, 91:18, 107:5, 126:22 shipyards - 91:5 Short - 65:20 short - 163:1, 163:2, 164:17, 165:19 shortcomings 161:10, 161:11 shortest - 20:11 Shortest - 20:12 show - 33:12, 34:12, 36:20, 38:5, 50:9, 50:13, 54:20, 59:19, 62:17, 62:21, 69:23, 70:1, 70:18, 83:12, 85:5, 94:4, 104:14, 108:17, 108:23, 110:16, 131:14, 135:19, 146:8, 149:13 showed - 70:7, 71:11, 71:13, 87:5, 131:14 showing - 49:22, 60:3, 82:9, 157:14 shown - 138:11, 152:18 shows - 29:11, 52:5, 61:11, 89:14, 129:16, 132:19, 139:1 sick - 38:18, 38:21 side - 32:9, 34:5, 35:15, 74:16, 74:17, 145:12 sidetracked - 39:16 Siemiatycki - 50:10 signal - 135:8 significance 94:20, 132:3, 154:4 significant - 30:6, 62:5, 94:12, 95:1, 97:17, 98:1, 105:15, 111:17, 111:19, 112:4, 127:23, 129:5, 143:9, 143:10, 143:12, 145:9, 154:10, 155:3, 155:16, 158:5 significantly 97:18, 109:14 similar - 31:17, 32:5, 33:18, 35:8, 125:13, 125:15, 125:20, 132:11, 132:12, 149:23, 153:22, 159:16 Similarly - 157:16 Simmons - 2:9 simple - 105:6, 105:7, 112:7, 124:7, 149:1 simplicity - 31:3 simply - 36:11, 109:1 Sinai - 44:19, 151:12 Singewald - 2:12 single - 12:22, 76:7 sister - 161:16, 163:17 Sister - 162:13 sit - 8:19, 52:14 site - 83:3, 97:11, 120:8 sites - 53:14, 76:21, 103:2 situation - 51:15, 80:18 situations - 47:14 six - 5:18, 8:6, 85:15, 87:16, 112:18, 118:9, 121:10 six-fold - 87:16 six-month - 8:6 six-vear - 5:18 size - 87:11, 89:15, 125:14, 125:16, 134:7 skewed - 81:15, 86:22 slide - 16:5, 16:13, 24:2, 24:3, 29:5, 29:11, 49:11, 49:14, 73:22, 109:22, 121:19, 122:1, 129:14, 129:15, 138:16, 138:21, 139:18 slides - 71:11, 129:21 Slights - 1:6 Slow - 103:21 slowly - 12:20 small - 5:20, 52:10, 64:2, 85:18, 87:10, 89:10, 96:2, 109:18, 109:20, 109:21, 110:7, 112:1, 134:8, 150:3 smaller - 95:3, 106:4, 106:8, 138:1, 138:4 Smaller - 138:3 smoke - 33:21, 40:3, 40:13, 40:14, 41:5, 71:3, 138:15, 138:18, 138:22 smoker - 34:7, 149:2, 149:9 Smokers - 74:17 smokers - 33:20, 34:2, 34:3, 34:10, 34:11, 40:6, 73:11, 74:12, 74:17, 138:18, 147:1, 149:6 Smoking - 147:2, 148:10, 148:12 smoking - 33:22, 38:4, 40:19, 40:20, 50:1, 51:5, 70:11, 70:13, 70:23, 71:3, 71:5, 73:2, 73:3, 73:7, 73:15, 74:15, 137:19, 138:5, 138:12, 139:2, 139:3, 139:4, 139:7, 139:12, 139:17, 145:21, 146:2, 146:13, 146:17, 146:23, 147:10, 148:8 Smr - 33:5, 123:6 snapshot - 46:18, 47:15 so-called - 76:19, 147:13 social - 142:11 societies - 76:14 Society - 14:11, 16:7, 16:9, 88:7, 146:20 Somers - 2:4 sometimes 112:15, 147:11, 157:18 somewhat - 132:12 somewhere - 51:1, 67:15 Sorry - 68:18 sorry - 41:2, 42:19, 48:18, 121:9, 153:1 sort - 17:14, 28:14, 44:7, 48:9, 78:8, 78:20, 82:3, 90:11, 92:7, 94:6, 101:20, 113:12, 113:18, 114:1, 131:13, 157:20, 161:9, 164:7, 14 164:8, 165:15 sorts - 13:16, 55:23, 140:13 sound - 57:23 sounds - 57:22 sources - 20:9, 28:22, 90:6, 136:13 south - 113:14 South - 96:10 space - 50:17, 88:9 Spain - 113:13, 114:18 speaking - 36:4, 51:2, 138:4 specialist - 13:9 specially - 20:15 specialties - 13:5 specific - 54:22, 68:19, 90:22, 94:8, 112:9, 131:5, 137:10, 153:22 specifically - 15:21, 19:9, 23:3, 90:19, 105:21, 120:5, 121:3, 160:14, 160:17 speculative - 110:8, 112:6 Spell - 151:7 spending - 102:1 spent - 7:8 Spillane - 1:8 Spirites - 57:19, 58:11 Spirta - 87:23 Spirtas - 87:21, 87:22, 89:23, 90:17, 90:23, 106:3, 120:15, 120:20, 121:2, 126:11, 126:16 split - 74:11, 127:5 spontaneous 67:10, 67:19 spouses - 75:6 sprayers - 145:12 sprouts - 28:10 stable - 88:21 staff - 113:2 stage - 7:21 stages - 7:19 stand - 37:23, 41:7, 44:12, 71:9, 79:9, 165:21 stands - 24:5 Starbucks - 30:19 start - 46:7, 46:12, 99:19, 111:20, 147:10 started - 12:15, 13:19, 17:23, 88:1, 99:19, 164:6 starts - 66:17 State - 1:1, 58:14, 84:7, 166:6, 166:9 state - 12:10, 45:13, 83:6, 97:23, 115:15, 116:11, 130:15, 130:23, 150:19 state-wide - 116:11 statement - 22:2, 65:5, 109:23, 110:2, 112:8, 112:9, 135:5, 138:3 statements - 21:15 States - 6:3, 6:10, 6:12, 8:11, 58:12, 63:9, 75:16, 118:7 states - 22:11, 22:20, 45:6, 47:5, 116:12, 116:13, 116:14, 116:19, 116:20 statistical - 51:4, 94:19, 100:7, 132:3, 154:11 Statistically - 36:4 statistically - 36:7, 51:2, 53:5, 64:5, 127:23, 155:3, 155:7, 155:16 status - 6:7, 74:15, 142:11 stay - 66:20, 88:22 stayed - 109:16, 128:12 step - 31:17, 35:7, 74:10, 131:23, 159:8 steps - 31:19 Steven - 2:6 still - 12:2, 18:1, 28:10, 46:10, 62:21, 83:13, 87:23, 99:11, 108:8 Stockholm - 123:15, 140:12 stop - 6:21, 29:22, 81:20, 84:13, 93:14, 110:22 story - 46:6, 61:8 stove - 100:2 straight - 26:13, 92:6 stratified - 73:19, 74:22 Stratified - 73:20, 73:21 stratify - 74:10 Street - 1:21 strengths - 80:15, 80:17, 81:17, 131:16, 131:17 stroke - 41:23, 42:1, 42:14, 42:17, 42:23, 43:6 stroke-related 42:17 strong - 26:3, 71:13 strongly - 57:6 structural - 21:1 structure - 11:2, 21:23 structured - 11:3, 27:17 structures - 11:5 students - 46:4 studied - 17:20, 51:19, 55:18, 57:8, 62:1, 113:17 studies - 8:15, 11:19, 11:20, 13:17, 14:1, 18:10, 18:15, 23:12, 28:23, 31:21, 31:22, 32:23, 34:17, 34:18, 34:21, 38:3, 43:20, 44:4, 44:15, 45:15, 46:22, 47:3, 48:5, 48:8, 49:1, 49:12, 49:22, 50:4, 50:8, 51:3, 51:10, 51:11, 51:21, 51:22, 52:1, 52:15, 52:18, 52:19, 58:8, 58:17, 58:19, 58:20, 59:1, 59:6, 59:16, 59:19, 59:20, 61:9, 64:7, 64:10, 64:14, 66:1, 69:16, 70:3, 70:4, 71:10, 71:23, 72:12, 75:21, 76:7, 77:18, 85:21, 98:18, 113:12, 115:9, 117:2, 118:20, 119:8, 119:19, 121:22, 122:1, 122:4, 122:17, 122:18, 122:19, 122:21, 122:23, 123:3, 124:7, 124:8, 124:10, 127:17, 130:17, 131:4, 131:14, 131:15, 131:20, 131:23, 132:4, 132:8, 132:11, 133:6, 133:8, 133:10, 133:14, 134:1, 134:8, 134:13, 135:19, 136:3, 136:6, 136:14, 136:18, 138:10, 139:1, 139:6, 139:8, 139:14, 139:17, 139:20, 139:23, 145:2, 145:4, 145:7, 145:15, 145:17, 145:20, 149:5, 150:14, 150:19, 151:4, 151:13, 152:14, 152:15, 159:2, 159:5, 159:9, 159:10, 159:21, 159:22, 160:9, 160:12, 160:16, 160:18, 161:3, 161:5, 161:15, 162:7, 162:9, 163:1, 164:7, 164:8, 164:12, 164:14 Studies - 49:2, 132:10 study - 8:10, 8:13, 15:15, 17:5, 30:17, 31:19, 31:23, 32:1, 32:21, 33:3, 33:4, 33:5, 33:13, 36:2, 37:6, 37:14, 37:15, 38:6, 38:17, 41:4, 41:8, 41:10, 41:12, 41:15, 41:22, 43:11, 43:22, 44:14, 44:18, 45:1, 45:2, 45:3, 45:16, 46:2, 46:3, 46:6, 46:9, 46:17, 47:1, 47:4, 47:6, 48:10, 52:4, 52:13, 57:18, 57:19, 57:20, 57:21, 61:14, 62:5, 69:19, 72:23, 75:9, 75:11, 75:12, 75:15, 75:19, 76:2, 76:18, 76:19, 77:2, 77:14, 80:13, 80:16, 80:17, 80:20, 81:4, 81:18, 81:21, 82:19, 82:20, 82:23, 83:2, 83:13, 83:20, 83:21, 83:23, 84:2, 84:4, 84:12, 84:14, 84:20, 85:7, 87:8, 87:12, 87:17, 87:20, 88:1, 88:4, 88:11, 88:20, 89:3, 89:5, 89:12, 89:23, 90:18, 91:20, 92:5, 92:7, 92:8, 93:15, 95:8, 95:9, 95:15, 96:6, 96:8, 96:10, 96:11, 97:10, 97:13, 97:14, 98:16, 98:19, 99:1, 99:3, 101:5, 101:9, 101:11, 101:20, 101:22, 102:1, 102:4, 102:6, 102:9, 102:16, 104:19, 112:16, 113:8, 113:13, 113:17, 115:9, 115:12, 115:15, 115:16, 117:3, 117:6, 117:7, 117:8, 117:10, 117:13, 117:15, 117:21, 118:1, 119:10, 119:17, 120:4, 120:5, 120:21, 120:22, 122:16, 123:8, 123:11, 123:13, 124:17, 125:23, 126:18, 126:19, 128:4, 130:5, 130:14, 131:1, 131:7, 131:9, 131:20, 133:9, 134:10, 136:23, 137:3, 138:6, 139:16, 139:22, 140:7, 141:2, 142:6, 142:8, 142:22, 144:8, 149:7, 149:8, 149:13, 150:18, 151:10, 151:13, 151:14, 157:19, 161:6, 161:8, 161:9, 161:13, 164:10, 164:12 subject - 87:2, 89:20 subjects - 89:3 submit - 118:19, 128:4, 144:19 submitted - 8:23, 131:10, 139:10 subset - 94:14, 95:3, 105:20 substance - 24:17 substances - 24:13, 55:12 subtract - 148:15 succinct - 11:16, 26:13 succinctly - 138:9 sufficient - 25:22, 26:2 suggest - 149:3, 152:11 Suite - 1:21 suits - 136:11 sum - 149:19 summarize - 121:22 summary - 18:2, 21:16 summation - 158:9 Superior - 1:1, 1:20, 166:6, 166:8 supervisors 138:20 supplemental 118:13 support - 144:8, 144:21, 150:14 supposed - 69:12 surface - 21:23, 24:1 surprised - 111:1 surprising - 40:5, 66:8, 67:4, 75:4, 144:20 surrogate - 56:7, 127:20 Surveillance - 84:10 surveillance - 116:9 survival - 23:7, 54:19 survive - 112:23 Svfs - 21:19 Sweden - 151:10 Swedish - 123:15, 124:19 switched - 108:21 Switzerland - 151:5 sworn - 3:7 synergistic 145:23, 149:14 synergistically 146:3 synergy - 128:16, 128:17, 128:22, 129:1, 148:23 synonyms - 48:9 system - 88:19, 88:22, 162:12 systematic - 18:10, 44:5, 129:7, 130:6, 130:9, 130:10, 130:13, 131:13, 137:14, 144:15, 159:7, 160:10 T table - 38:13, 46:20, 74:4, 74:11, 79:17, 85:13, 115:5, 154:15, 155:11, 156:11, 156:23, 157:21 Table- 153:16, 153:17, 154:17, 155:11, 155:13 tables - 38:11, 74:12, 89:16, 104:15, 106:7, 131:14, 132:19, 153:13, 156:16 tabulate - 50:12 Tarry- 2:2, 3:2, 3:4, 3:10, 36:21, 36:23, 37:3, 49:17, 49:20, 55:11, 55:15, 55:17, 59:13, 60:16, 64:16, 65:21, 65:23, 79:9, 79:19, 82:11, 82:16, 82:18, 94:17, 94:19, 94:22, 95:6, 112:17, 112:22, 113:7, 141:15, 143:14, 153:11, 154:5, 158:6, 158:8, 161:19 taught - 14:5 taxes - 118:17 teach - 12:21, 12:23, 13:1, 13:2, 14:8 teachers - 41:23, 42:15, 42:20, 42:21, 43:6, 66:11, 67:5, 100:3, 101:3 teaches - 10:7 teaching - 11:21, 11:22, 12:20, 14:4 technical - 129:4 technically - 112:11 technique - 73:13 Ted- 2:8 temporality - 46:23, 47:9 Ten- 147:16 ten - 28:8, 30:23, 31:1, 31:6, 32:6, 32:8, 32:11, 32:12, 32:13, 34:3, 34:5, 34:10, 50:15, 50:20, 64:18, 64:19, 119:2, 119:4, 119:5, 127:22, 128:1, 128:2, 128:13, 128:14, 147:4, 147:10, 147:14, 148:6, 148:10, 148:11, 148:12, 148:17, 148:18, 148:19 15 ten-fold - 34:5 tend - 33:21, 157:8 tends - 49:1 tenfold - 87:15 term - 17:14, 18:23, 19:17, 23:6, 29:3, 53:11, 53:13, 69:16, 110:12, 128:23, 132:20, 132:21 terms - 11:16, 55:7, 56:6, 90:7, 90:8, 121:6, 127:23, 129:3, 133:22, 138:14, 141:9 Teschke- 58:12, 101:9, 101:11, 101:12, 101:14, 102:4, 110:5, 113:9, 129:20 test - 27:16, 28:2, 30:11, 49:2, 49:5, 61:19, 75:21, 78:6, 90:4, 93:5, 101:21, 102:7, 105:17, 109:4, 132:16, 132:21, 141:8, 149:17, 159:10, 160:17, 163:19 tested - 27:20, 28:21, 128:22, 130:17, 141:11, 149:14 testified - 3:7, 21:6, 44:17, 117:17, 152:6 testimony - 9:12, 17:9, 53:9, 134:19 testing - 48:15, 59:17, 76:9, 131:5 Teta- 83:21, 84:3, 84:14, 84:15, 85:21, 86:6, 87:8, 89:18 textbook - 22:8, 71:8 textbooks - 47:10 themselves - 21:2, 69:16 theory - 144:6 Therefore - 37:12, 43:2, 112:5 therefore - 19:1, 19:3, 76:6, 86:21, 92:16, 136:6, 137:12, 151:22 they've - 71:18, 80:14, 106:19 They've- 106:19 thinking - 81:15, 111:20, 165:10 third - 28:4, 40:21, 41:3, 69:20, 72:9, 119:22, 127:16, 133:22, 133:23 Thomas- 1:14 thousand - 32:13, 32:14, 32:16, 34:2, 42:16, 42:17 thousands - 115:18 three - 5:20, 6:18, 8:15, 11:3, 29:15, 57:20, 60:9, 88:5, 92:20, 99:13, 104:13, 114:18, 132:13, 133:4, 133:5, 135:9, 135:11, 145:10, 157:14 threshold - 143:19, 143:23, 144:7 thrive - 27:23 throughout - 75:15, 116:14 throw - 78:3 Thursday- 1:18 tie - 64:14 tier - 132:11, 132:12, 132:13, 133:6, 133:9, 133:22, 134:5, 134:12 tiers - 132:9, 133:5 tighter - 89:16, 107:10, 107:11 tiny - 85:17 tip - 82:7, 82:10 tissue - 53:18, 161:11, 161:12, 162:20 title - 56:23 today - 3:22, 13:20, 163:22 together - 21:13, 131:14, 140:1, 147:19, 147:20 tomorrow - 118:16 tongue - 13:23 took - 10:9, 31:4, 78:13, 96:11, 101:23, 165:12 Took - 120:20 tool - 23:13 top - 28:5, 147:9 topic - 17:1, 34:18, 133:1 total - 41:20, 42:8, 42:13, 42:18, 68:3, 68:23, 79:5, 79:15, 85:14, 116:22, 121:12, 149:19 Total - 42:19 totality - 48:23, 61:9, 87:18 touch - 49:17 touched - 18:21, 47:20 tough - 65:13 town - 86:3, 86:8, 87:7, 92:2 towns - 86:11 toxicity - 21:18 toxicologists 11:10, 16:12 traded - 10:17 traditional - 127:11, 163:21 train - 65:10 trained - 10:1 training - 5:9, 5:14, 5:16, 6:9, 6:14, 6:16, 7:16, 7:23, 8:6, 8:12, 8:22, 9:5, 9:9, 10:10, 10:11, 11:1, 13:4 transcript - 166:7 Transcript- 1:18 transport - 100:3 trauma - 39:7, 39:10, 39:14, 40:14 treatment - 23:7, 54:17 treats - 29:9 tree - 75:6 trial - 64:22 trials - 48:11, 48:16 triangle - 70:16, 72:1 trick - 150:17 trickiest - 69:7 tries - 163:14 triggered - 151:18 triple - 145:18 trivial - 60:14 trouble - 47:9 true - 12:12, 17:19, 97:12, 114:1, 114:2, 123:1, 123:2, 136:18, 152:16 trustworthiness 131:21, 131:22 try - 39:17, 65:3, 86:9 trying - 20:15, 65:12, 78:6 Tumor- 84:5 tumor - 88:12, 120:23, 135:8 tumors - 55:10 tuner - 78:19, 78:23 turn - 162:4 turns - 35:12 Twenty- 40:14 Twenty-five- 40:14 Twice - 52:7 twice - 95:10, 109:6 two - 7:4, 7:8, 7:18, 12:15, 12:17, 15:20, 22:13, 29:15, 31:21, 32:9, 34:4, 35:15, 38:11, 38:13, 40:22, 41:2, 42:4, 42:14, 43:2, 43:4, 45:4, 46:20, 53:7, 62:19, 62:20, 72:22, 73:5, 74:4, 74:7, 74:11, 78:1, 79:17, 82:9, 83:5, 83:16, 88:10, 89:16, 89:20, 93:7, 94:6, 96:19, 102:14, 102:15, 104:15, 105:21, 106:7, 108:21, 113:13, 118:23, 119:8, 119:11, 119:19, 122:19, 123:19, 130:1, 130:4, 131:2, 132:12, 143:21, 145:2, 146:15, 147:18, 151:12, 153:19, 156:16, 161:5, 165:10 Two- 26:11, 68:22, 160:18 two-by-two - 38:11, 38:13, 46:20, 74:4, 74:11, 79:17, 83:16, 89:16, 104:15, 106:7 two-fold - 74:7 two-year - 96:19 twofold - 52:6, 52:9 type - 21:22, 22:22, 28:20, 33:13, 41:8, 105:17, 105:18, 144:2 types - 13:21, 14:15, 17:13, 18:22, 43:19, 53:17, 55:1, 56:11, 104:13, 105:21, 128:16, 160:8 typical - 50:9, 53:7 Typically- 41:18, 61:17 typically - 14:8, 45:12, 45:15, 48:6, 78:10, 127:5 U Uh-hmm - 84:19 ultimately - 21:19, 22:18, 47:17 umber - 56:1 unable - 69:3, 97:19 unanswered - 126:9 uncertainty - 36:10 unclear - 147:2 undefined - 103:3 under - 125:16, 137:3 undergo - 6:15, 164:20, 164:21 underlying - 43:16, 43:17 understood - 34:17, 65:14, 68:17 undertaken 162:14 underwent - 93:11 unequivocal 61:20, 145:7 unequivocally 60:8 unexplained 68:14, 68:15 unexposed - 37:5, 69:5, 114:20, 142:22 Unfortunately 162:16 unified - 56:11 unit - 39:10, 40:14 United - 6:3, 6:10, 6:12, 8:11, 58:11, 63:8, 75:16, 118:7 University - 3:17, 7:6, 7:7, 12:14, 84:18, 96:9, 101:14, 101:15, 101:18 university - 12:4 unless - 136:6, 148:21, 148:22 unlike - 46:17, 55:3, 85:22, 116:10, 118:7, 119:4 unlikely - 37:19, 37:21 unmatched - 73:10 unquote - 112:11 unreasonable 40:12 unresolved - 47:1 unusual - 9:11 Up - 129:20 up - 18:11, 21:11, 28:18, 38:1, 44:19, 57:16, 62:21, 67:17, 74:12, 74:23, 79:13, 99:4, 101:20, 104:1, 104:7, 107:6, 110:15, 113:14, 117:8, 119:7, 119:21, 119:22, 121:19, 122:16, 123:3, 124:22, 125:3, 127:5, 136:22, 140:17, 143:1, 147:9, 149:12, 152:1, 153:11, 153:13, 157:20, 158:2, 158:21, 165:15 upcoming - 64:22 upper - 126:1 useful - 47:8, 90:23, 136:15, 156:17 usefulness - 47:3 utilize - 27:21 V Va - 88:15, 88:16, 88:19, 88:22, 88:23 vaccination - 8:9 vaccine - 8:16 vaccine-related 8:16 vaccines - 8:16 vague - 94:7 valuable - 126:15 value - 74:4, 154:7, 154:9 variability - 54:16 variant - 58:2 variety - 9:6, 17:15, 19:13, 19:14, 28:22, 39:21, 55:8, 56:16, 85:3, 102:10 Various - 20:23 various - 7:23, 10:21, 11:4, 17:21, 20:3, 20:6, 21:6, 24:12, 69:13, 83:7, 83:8, 99:17, 130:19, 154:1 vary - 17:18, 19:3, 55:2 varying - 58:1 vehicle - 5:6, 15:22, 15:23, 62:7, 63:2, 63:3, 63:10, 104:12, 105:2, 105:4, 105:5, 105:13, 105:19, 106:4, 107:1, 107:2, 107:4, 107:13, 107:17, 107:23, 108:11, 109:2, 110:6, 122:13, 126:20, 127:6, 130:19, 133:19, 134:2, 134:18, 139:13, 161:14 vehicles - 114:17, 124:1 verbalize - 40:23 Verechia - 166:3, 166:19 versed - 11:11 versus - 54:5, 90:7, 93:19, 138:15, 154:21, 157:3 vertical - 50:7 Veteran - 88:15 view - 28:15, 88:20, 154:11 virtually - 13:11, 13:14, 106:10 Virtually - 13:13, 94:18, 94:23 visa - 6:5 visits - 39:14 visually - 50:12 volume - 134:21, 135:12, 141:9 volunteers - 30:18 voter - 102:20, 102:22 voters - 102:19, 103:13 votes - 102:22 W Waddell - 1:10, 1:11, 165:17 Wait - 63:13 wait - 52:14, 82:14 waiting - 109:8 wall - 16:6 Washington 58:13, 101:16, 101:18, 115:16 watch - 46:7 watched - 46:9 watching - 46:5 water - 28:5, 28:6, 45:23 ways - 39:13, 72:22, 73:18, 81:1, 94:6, 95:11, 131:7, 133:10, 146:15, 153:17 weakness - 131:17 weaknesses 81:17, 131:18 Wednesday - 64:23, 65:4 week - 65:6 weeks - 65:16 weight - 47:17, 158:17, 159:4 weighting - 24:16 Welch - 120:3, 120:4, 120:13, 120:15, 120:19 well-accepted 131:19 well-defined - 42:2 well-versed - 11:11 West - 30:21 Whales - 96:12 whatnot - 142:12 whatsoever 122:10 whereas - 12:5 whichever - 135:16 White - 65:10, 65:15, 65:18 white - 72:7, 107:7, 133:19 white-collar - 72:7, 107:7 who-done-it - 46:5 whole - 43:23, 76:10, 76:11, 88:21, 137:9 wide - 9:6, 17:15, 52:11, 106:9, 108:4, 116:11, 134:8 widely - 19:3, 57:23 wider - 95:4 wife - 103:20 William - 2:12 Wilmington - 1:21, 42:10, 166:12 windowsill - 28:8 Withdraw - 158:21 witness - 44:17, 162:1, 162:4 Witness - 49:19, 59:23, 60:5, 79:11, 81:23, 82:2, 82:4, 94:18, 94:23, 141:23, 143:4, 153:16, 162:13, 162:16, 163:12, 163:18, 163:21, 164:14, 165:7, 166:13 Woitowitz - 91:21, 91:22, 92:1, 95:7, 103:2 women - 90:16 word - 17:14, 100:10, 110:12, 131:22, 152:4, 152:9 words - 18:11, 59:17, 64:1, 73:15, 74:6, 86:20, 93:12, 114:3, 131:2, 133:15, 148:20, 156:18 worker - 78:18, 79:2, 123:22, 125:4 workers - 60:17, 60:18, 60:23, 63:5, 72:4, 72:5, 79:14, 80:6, 99:22, 99:23, 107:5, 116:2, 122:13, 127:13, 133:19, 133:20, 134:2, 147:12, 152:13 workforce - 123:19 works - 33:13, 34:13, 38:5, 87:23, 88:3, 147:14, 147:17 world - 51:15, 58:9 World - 8:7, 24:8 worry - 137:9 worse - 157:9 worth - 102:1, 115:18, 116:16 wrap - 74:23, 119:7, 149:12, 153:11, 165:15 write - 11:19 writing - 12:18 written - 50:20 wrote - 92:13 X x-ray - 45:7, 151:20, 153:18, 156:9, 156:13 Y Yale - 84:17, 84:18 yard - 63:14 year - 5:18, 7:21, 8:2, 8:19, 15:20, 30:23, 32:14, 54:10, 75:18, 76:17, 86:11, 96:19, 119:8 years - 5:20, 6:18, 7:4, 7:9, 12:15, 12:17, 26:8, 30:23, 31:1, 32:7, 32:13, 32:16, 42:11, 51:23, 70:14, 75:17, 76:16, 83:5, 84:12, 92:20, 98:16, 99:6, 102:14, 102:15, 109:7, 109:9, 109:11, 115:17, 116:15, 116:16, 116:19, 116:20, 118:23, 119:3, 119:4, 124:21, 124:22, 124:23, 125:2, 125:9, 125:18, 127:22, 128:1, 128:2, 128:13, 128:14, 139:6, 156:14, 156:19, 156:20, 157:22 Yesterday - 117:11 yesterday - 44:17, 45:21, 69:17, 77:18, 93:17, 109:22, 122:15, 122:22, 140:22, 143:15, 165:16 York - 44:19, 88:13 yourself - 66:18, 76:1 Z zero - 63:2, 66:23, 67:1, 83:10, 83:11, 142:18, 142:21, 143:4 zooming - 54:22 16