OSHA: Proposed Standard For Indoor Air Quality: ETS Hearings, January 17, 1995

OSHA: Proposed Standard For Indoor Air Quality: ETS Hearings, January 17, 1995




January 17, 1995

Department of Labor

Washington, D.C.

The above-entitled matter came on for hearing, pursuant to notice, at 9:45 a.m.


Administrative Law Judge



R.J. Reynolds
Christopher R. E. Coggins 11378
Michael W. Ogden 11381
Paul R. Nelson 11420
Stephen B. Sears 11445
Michael W. Ogden 11471
Christopher R. E. Coggins 11486
Hoy R. Bohanon, Jr. 11512


Ms. Sherman 11543



228 11419 11419

229 11419 11419

230 11444 11444

231 11444 11444

232 11468 11468

233 11468 11468

234 11486 11486

235 11486 11486

236 11511 11511

237 11542 11542

238 11542 11542

9:45 a.m.

JUDGE VITTONE: Good morning, ladies and gentlemen. We resume our hearings into the proposed rule by the Occupational Safety and Health Administration on indoor air quality.

For the record, this is day 51 of these hearings.

We have on this morning a panel from the R.J. Reynolds Tobacco Company that will be testifying today and will be also here for the next several days, Tuesday, Wednesday and Thursday of this week.

I would like to begin by introducing the chairman of the RJR panel, Dr. Coggins.

DR. COGGINS: Thank you, Judge Vittone.

Good morning, ladies and gentlemen. I am Chris Coggins. I hold a Ph.D. from the University of Edinburgh, Scotland, and I am a board certified toxicologist.

JUDGE VITTONE: Excuse me a second. I'm sorry. I forgot something here. I have a preliminary matter that somebody wanted to raise with me.

Mr. McNeely, would you come forward, please?

MR. McNEELY: Actually, Mr. Herman will be making --

JUDGE VITTONE: Okay. Mr. Herman.

Why don't you come on around here, Mr. Herman?

MR. HERMAN: Judge Vittone, I am not at all certain of the procedure that will be followed this morning.

JUDGE VITTONE: I will be glad to explain it to you, then. What we plan to do here today, and I guess I should advise everyone here, sometimes I tend to forget that not everybody understands how we operate, what we will be doing over the next several days is simply this: the R.J. Reynolds panel will be making a presentation today, individuals members of the panel will be called at various times and make a presentation on various subject matter, on topics, I'm not sure exactly what each person will be talking about.

We will be taking them in seriatim. Dr. Coggins is the chairman, as I understand, of the panel and then he will call upon individual members of this panel to speak on various topics.

That will probably consume all of this morning and at least a portion of this afternoon's proceeding. After they have completed their direct testimony, using the slides and everything, then they will be made available for examination by the participants in the proceeding.

I will then turn to Ms. Sherman, who heads up the OSHA staff. Ms. Sherman and her staff will be able to examine the panel about their testimony. I do not know how long that will be but I am sure it it's gong to be lengthy examination.

After she has completed her direct examination, then I will make the panel available for examination by other participants in this proceeding who have complied with the rules with respect to becoming actual parties or participants to the proceeding and that will probably take us through some time on Thursday.

Then at that point after all of the participants have completed their examination, if there is any brief clarification that the panel members want to make, I will give them an opportunity to make a statement at that point and then we will end with them.

It is the same pattern we have pretty much followed in this proceeding for every other witness or panel of witnesses that have testified in this proceeding so far, including the NIOSH panel, the American Medical Association, the unions that have testified, anybody else.

So that's basically our regular process, so that's it.

MR. HERMAN: May I ask if any of these gentlemen have engagements that will cause them to leave at any time before we have an opportunity to fully speak with them about the issues they have raised?

JUDGE VITTONE: Well, I think we've gone over this but I'll give Mr. Grossman an opportunity to respond to that.

Mr. Grossman, since you haven't been back for a while, why don't you tell everybody who you are?

MR. GROSSMAN: I will. Your Honor, my name is Ted Grossman. I am here on behalf of R.J. Reynolds.

The panel, all of the members of the panel, have reserved today and tomorrow and Thursday for these hearings as the schedule calls for. Obviously they haven't scheduled an indefinite appearance and neither has any other organization.

JUDGE VITTONE: So they will be here.

Now, just so we are clear, the OSHA panel has also scheduled for Thursday two other individuals which we will bring on at the completion of the RJR panel, Mr. Allan Hedge and Mr. Gene Davidson. I don't know anything about those gentlemen but they will be here on Thursday but they will be here after the RJR panel is completed.

MR. GROSSMAN: Yes, Your Honor. We scheduled them in accordance with the schedule, has OSHA has.


Mr. Herman, does that answer your question, sir?

MR. HERMAN: Yes, it does. And I understand that thankfully we will have a full opportunity because the entire panel will be here for the duration.

MR. GROSSMAN: For the duration of Reynolds' testimony. Yes.

MR. HERMAN: And cross.


MR. HERMAN: Your Honor, I have only one comment.


MR. HERMAN: And I appreciate your allowing me to take this time just to clarify what the procedure was.

We have an objection in principal to allowing RJR to appear en masse like this. It really thwarts the ability, our ability, to examine each of the gentlemen individually and it allows RJR to take the stage for a very long period of time without breaking the continuity and allows them in effect to have a continuing commercial for the benefit of the cigarette industry.

We understand your ruling and we appreciate it and our concern is that while unintentional certainly the way that this process is being allowed falls square into the concerns that those of us who represent the other side of this case are most concerned about and that is a full and complete airing of the issues without the opportunity, though we respect your ruling, I want to state it for the record, without the opportunity of being able to question each of these gentlemen after their testimony is concluded. It's going to adversely affect our ability to get a full and complete airing of our side of the issue.

And having said that, I appreciate the time you have allowed me.

JUDGE VITTONE: You are welcome, sir.

Mr. Grossman?

MR. GROSSMAN: Thank you, Your Honor. If I may respond briefly, we've been across this bridge many times before, both as to Reynolds and as to others, and, as Your Honor has already said, every other panel, the AMA, OSHA, NIOSH, has testified in precisely this way. The only difference, I might add, is that Reynolds has identified the authors of the submissions that it has before you.

Mr. McNeely raised exactly this point on December 7th and at page 8686 of the transcript --

JUDGE VITTONE: Mr. Grossman, I don't want to cut you off, I'll give you an opportunity to respond but I do not intend to change my ruling in this matter.

MR. GROSSMAN: I was just going to say at that page you said that there would be no revision and no reconsideration of your ruling.

JUDGE VITTONE: My ruling stands. I will permit the RJR panel to proceed, present its entire direct testimony and then be made available for examination.

Thank you, Mr. Herman, for your comments. I understand your position and it's the same position that the OSHA staff took but it's a position I decided not to adopt and decided to proceed in this manner and this is the way we will go forward.

Thank you, Mr. Herman.

Thank you, Mr. Grossman.

Anything else before we begin?

(No audible response)

JUDGE VITTONE: Seeing no other hands raised, Dr. Coggins, you can resume what you were saying before.

DR. COGGINS: Thank you, Judge Vittone.

Good morning, ladies and gentlemen. I am Chris Coggins. I hold a Ph.D. from the University of Edinburgh in Scotland and I am a board certified toxicologist. I am a Principal Scientist in the Research and Development Department of the R.J. Reynolds Tobacco Company. I have performed and published toxicological evaluations with tobacco smoke since I received my doctorate in 1976.

My colleagues and I are here today to present data that will illustrate a number of serious shortcomings in OSHA's proposed rule as it relates to environmental tobacco smoke, ETS.

We hope that our information will help OSHA develop a better, more appropriate approach to ETS in the workplace.

As you will hear, Reynolds Tobacco believes that the proposed rule requires substantial modification. In particular, we do not believe there is any justification for treating ETS in a different manner from other indoor air components.

When we have completed our presentations, we will answer questions concerning them. I will be the moderator for the panel and I will direct the questions to the most appropriate panel member.

Now, let me introduce our panel.

Michael Ogden, who received his Ph.D. in analytical chemistry from Virginia Tech in 1985, will speak first. Dr. Ogden will present data on the actual amounts of ETS that are found in a number of different environments, including workplaces and homes. Dr. Ogden will show that non-smoker exposures to ETS are typically very much smaller than those cited in the proposed rule.

Next, Paul Nelson, who obtained his Ph.D. in analytical chemistry from Georgia Tech in 1987, will talk about the severe limitations in the use of nicotine and cotinine to precisely predict ETS exposure.

Then, Stephen Sears, on my left, who obtained his Ph.D. in theoretical chemistry from UNC at Chapel Hill in 1980, will review some of the problems in the OSHA risk assessments, problems that in the main have not yet been raised at these hearings.

Dr. Sears' presentation was co-written by Mr. Thomas Steichen, who obtained his Master's degree in statistics from the University of Kentucky in 1978. Mr. Steichen will also be on our panel to answer questions.

After that, Dr. Ogden will return to present data showing that misclassification affects the results in epidemiologic studies concerning ETS.

I will then analyze some animal inhalation studies. I will show that the studies used in the proposed rule are inappropriate and that more relevant studies were not given the attention they deserve.

Finally, Mr. Hoy Bohanon, who obtained his Bachelor's degree in mechanical engineering from Georgia Tech in 1974, will discuss engineering solutions as alternatives to a national smoking ban. Mr. Bohanon is a professional engineer.

Our presentations will demonstrate why from a scientific and technical point of view there is no justification for attempting to impose a zero exposure standard for ETS in the workplace and why proper ventilation under the general IAQ standard is the least costly, least intrusive and best way to deal with workplace ETS issues.

Now, Dr. Ogden will present data on the actual amounts of ETS that are found in a number of different environments, including workplaces and homes.

Dr. Ogden will be using the slide projector.

JUDGE VITTONE: Okay. Each of you has slides, as I understand it, which you will be using.

DR. COGGINS: That's correct.

JUDGE VITTONE: I have been provided with copies of your presentation as well as copies of the slides.

With the completion of each witness I will identify by exhibit number his presentation and slides for the record.

It is also my understanding that there are copies of your statements and slides available in the back, outside the hall here?

DR. COGGINS: That's correct.

JUDGE VITTONE: Okay. So they are on the table back there for anybody who may want to see them.

DR. COGGINS: Correct.

JUDGE VITTONE: Thank you very much.

We will proceed with Dr. Ogden if you are ready, sir.

Are you going to be using the slides right away, Doctor?


JUDGE VITTONE: Let me move out of your way.


DR. OGDEN: Thank you, Dr. Coggins, Your Honor.

Good morning. I am Dr. Michael Ogden. I hold a Bachelor of Science degrees in both chemistry and applied mathematics and a Ph.D. in analytical chemistry.

Since 1985, I have been employed by R.J. Reynolds Tobacco Company in the Research and Development Department. For the past nine years, my research has been almost exclusively focused on studying environmental tobacco smoke by developing and applying methods for assessing ETS exposure.

My testimony will address the treatment of exposure to ETS in the Notice of Proposed Rulemaking and comments made by OSHA consultants.

In slide 1, I pose five questions. They are:

(1) How should we measure exposure to ETS?

(2) What is the correct definition of exposure?

(3) How much ETS is actually in the workplace?

(4) How does ETS exposure compare for living with smokers and working with smokers?

(5) How many workers are actually exposed in their workplace?

Significant new data relevant to each of these questions and, more specifically, to OSHA's assertions are also summarized and introduced at the appropriate places.

Before OSHA can regulate any workplace exposure, including ETS, it must demonstrate with substantial evidence that a significant risk of material impairment to health is present in the workplace.

In addition, OSHA has to know the answer to two very basic questions and these are how much ETS is in the air and how many workers are exposed. The NPR does not adequately answer either of these questions. As a result, the proposed workplace ETS rule is based on inadequate information.

In my testimony here today, I will demonstrate the following points as outlined in slide 2.

(1) Methodology for measuring exposure. OSHA has not measured ETS exposure in the workplace and has not considered the most up-to-date and relevant information on how to go about measuring the exposure.

(2) The definition of exposure. In much of the NPR, OSHA is apparently using an incorrect definition of exposure.

(3) The concentration of ETS in the workplace. OSHA doesn't know how much ETS is in the typical workplace. As I will demonstrate, typical exposure levels are up to 250 times lower than assumed by OSHA.

(4) Equating home and workplace exposures. In lieu of actually measuring exposure, OSHA attempts to equate workplace exposure for people who work with smokers with home exposure for people who live with smokers. Research shows that instead of these exposures being equal, the workplace exposures are actually five to 10 times lower than home exposures.

(5) The number of workers exposed. OSHA doesn't know how many workers are actually exposed to ETS at work. They have significantly overestimated the number of workers exposed.

The important point is this: Without the best available answers to any of these questions, there is only one logical outcome: OSHA's conclusions regarding ETS in the NPR are not meaningful.

I'll now turn to the first question I posed: How do we actually measure exposure to ETS? Quite simply, we can measure people's exposure to chemicals in the air that are attributable to ETS. These compounds are called markers. This is slide 3.

In May 1993, I presented to OSHA a summary of the commonly used ETS exposure markers. In that presentation, I concluded that two markers, 3-ethenylpyridine and solanesol, were the best available markers for the vapor phase and the particulate phase, respectively, of ETS.

Nicotine is a commonly used marker for ETS; however, it is inferior to both 3-ethenylpyridine and solanesol, particularly at low concentrations.

Recently, these conclusions regarding nicotine and 3-ethenylpyridine have been questioned by Dr. Katherine Hammond at these hearings. I would like to address her criticisms here but before I do I would like to present some additional background information that is relevant to support my qualifications to make these conclusions.

I developed and published the method for measuring nicotine in ETS that is referred to as the XAD-4 nicotine method. As outlined here in slide 4, this method is currently the most thoroughly tested, the most rigorously validated, and the most widely used method in the world.

I currently serve as Association of Official Analytical Chemists Associate Referee for studying nicotine in environmental tobacco smoke. This title is used to designate a technical expert charged with supervising method validation and providing statistical analysis of the results and writing the technical protocol.

This XAD-4 nicotine method has undergone two successful international collaborative studies. As shown in slide 5, it is an approved or official method of the following organizations: The Association of Official Analytical Chemists, the U.S. Environmental Protection Agency, and the American Society for Testing and Materials. Also, this method is currently a draft international standard within ISO, the International Standards Organization.

It is currently in use by researchers in at least eight countries. The analysis is also available in at least four commercial laboratories in the U.S. and one in Canada. This same method is also used for the determination of 3-ethenylpyridine in ETS.

Testifying on behalf of OSHA, Dr. Hammond correctly concluded that one of the most important attributes of a marker is that its concentration should increase with the source strength and reflect the concentration of the complex mixture, here, ETS. However, she incorrectly concluded that 3-ethenylpyridine fails this most fundamental test for a marker.

In her Figure 1, which I have reproduced here in slide 6, Dr. Hammond shows results from one experiment done by another researcher in which both nicotine and RSP increase more or less linearly with the number of cigarettes smoked. However, the data for 3-ethenylpyridine do not appear to increase in the same fashion. Not only is this result for 3-ethenylpyridine implausible, it is incorrect.

Shown here in slide 7 are the ETS vapor phase marker results from a similar experiment conducted at R.J. Reynolds. These data demonstrate two points.

First, 3-ethenylpyridine increases linearly with the number of cigarettes smoked. Second, 3-ethenylpyridine also tracks exactly the vapor phase of ETS, as measured by carbon monoxide and flame ionization detector response, which is an indication of the total volatile organic compounds present in ETS.

As you can see, although nicotine increases in a similar fashion, it actually overestimates the vapor phase of ETS. This overestimation becomes more predominant with increasing levels of smoke.

Myosmine also exhibits the same trend as nicotine, although to a lesser extent.

Dr. Hammond levies an additional criticism against 3-ethenylpyridine. She claims that 3-ethenylpyridine is not as sensitive a marker as nicotine in detecting ETS. This also is not true.

In 1992, we published limits of detection for both nicotine and 3-ethenylpyridine showing that the that 3-ethenylpyridine is actually twice as easy to detect as nicotine.

Slide 8 shows the ETS particulate phase marker results from the same experiment at RJR. As for the vapor phase components, all markers increase linearly with increasing concentration. However, all particulate phase markers track each other very well. This is to be expected in this controlled experiment where all RSP comes from ETS.

In addressing how to go about measuring ETS exposure, the NPR concludes that the use of an internal measure of individual exposure such as body fluid cotinine is preferable to actually measuring external exposure. I disagree with this statement.

While an internal measure may be well suited to some experiments in the workplace, it is certainly not true with regard to ETS.

First of all, of the potential biomarkers currently available for estimating exposure to ETS, cotinine is the only one that is even marginally useful.

Of all biomarkers that have been proposed, cotinine is the most tobacco-specific. Also, cotinine has a half-life of approximately 17 hours, although it certainly can vary over a much wider range. This was discussed in more detail previously by Dr. Neil Benowitz.

This half-life is the amount of time it takes for half of any nicotine inhaled to be eliminated from the body as cotinine. This means that at best cotinine provides an integrated estimate of exposure over the preceding one to three days.

However, in the context of measuring workplace ETS exposure, cotinine is not to be preferred over air monitoring. There are several reasons for this, however, I will explain only one of them here in slide 9.

A significant problem with the use of cotinine for workplace exposure assessment is this same relatively long half-life that I just mentioned. By relatively long half-life, I now mean relative to the continuous amount of time spent at work.

The continuous amount of time spent in the workplace for most workers is only about eight hours. This continuous amount of time would even be less for workers who may leave their workspace to run errands, go out to eat, et cetera.

In order to use cotinine or any other biomarker with a similar half-life to infer anything about exposure at work requires the implicit assumption that all out-of-workplace exposures are the same for everyone.

This is, of course, an illogical assumption. An individual's body fluid levels of cotinine cannot distinguish between nicotine inhaled at work, at lunch, at home or anywhere else.

I would like to cite one specific example from my own research that illustrates how relying solely on cotinine levels would have resulted in a serious error concerning potential workplace exposures.

This study, outlined in slide 10, was conducted in Columbus, Ohio in 1991. The results were presented to OSHA in May 1993, and were published later that year.

Among all non-smoking subjects who were exposed to ETS at home, a statistically significant difference in cotinine levels was found between those who worked outside the home and those who did not work outside the home. In short, the subjects who were spousal-exposed to ETS at home and who also worked outside the home had higher cotinine levels than subjects who were spousally exposed at home and who did not work outside the home.

If I had relied solely on cotinine data, I would have attributed this increased cotinine to ETS exposure at work. However, I would have been wrong.

Air monitoring in the homes revealed higher concentrations of nicotine and 3-ethenylpyridine in the home of the working subject. Thus, the difference in cotinine was truly attributable to a difference in home exposure and was not due to a workplace effect.

Personal monitoring of workers going about their daily activities in their workplace is a much better way to determine actual exposure.

I would like to offer this specific advice to OSHA: If you want to know what exposure is in the workplace, measure it.

With readily available materials and methods, a number of informative constituents of ETS can be measured, including 3-ethenylpyridine, solanesol and even nicotine.

From the quote in slide 11, you can see that Meridian Research reached the same conclusion regarding ETS exposure in the workplace. This was in a 1988 report which was commissioned by OSHA.

Moving to my next point in slide 12, now let's examine the proper definition of exposure and contrast that to what was actually used in the NPR.

Since it is critical in evaluating the merits of this entire section of the NPR, the correct definition of exposure must first be given.

Exposure can correctly be defined simply as being equal to concentration times time. This equation defining exposure is explicit in that exposure in any given environment is equally dependent upon both the time spent in that environment and the concentration level of the contaminant in that environment. Without assessment of both variables, exposure cannot be determined.

Let me illustrate with a simple analogy. Remember that exposure is equally dependent on two variables, time and concentration. Similarly, the distance you drive in your car is dependent on two variables, the amount of time you spend in your car and the average speed of the car.

Imagine if you will trying to estimate how far I drove yesterday if all I tell you is that I spent an hour and a half in my car.

If I were at home in North Carolina yesterday in the middle of the afternoon, I could have driven all the way from Winston-Salem to Durham, a distance of about 75 miles. If I were here in Washington yesterday during rush hour, I might not have made it the four miles or so to the Key Bridge.

The treatment of ETS exposure among the non-smoking, working U.S. population in the NPR is superficial and extremely problematic. In fact, OSHA consistently uses an incorrect definition of exposure in the NPR in sections which focus only on duration.

OSHA employs comparisons of reported durations of ETS exposure between homes and workplaces where smoking occurs.

The data sources on which OSHA relies, such as the CAP survey, only yield estimates of potential exposure. The vast majority of information cited in this section of the NPR contains no measurements of workplace exposure whatsoever, an omission that most seriously limits the development of a workplace risk assessment.

Compounding the problem of using an incorrect definition, OSHA also has improperly analyzed many of the studies on which it does rely.

For example, in the NPR's treatment of the California Activity Pattern, or CAP, survey, this contains a number of errors and misinterpretations. I'll cite just one example from the CAP survey.

The NPR claims the CAP study shows that 51 percent of male and 38 percent of female non-smokers reported ETS exposure at work and further claims that this verifies the high percentage of non-smokers who are exposed to ETS while at work.

These figures actually represent the percentages of exposure time reported to occur at work for non-smokers who reported exposure to ETS at any location. These are not the percentages of the working populations exposed, as stated in the NPR.

Additional shortcomings of the CAP survey and OSHA's interpretation of it in the NPR are detailed in my written submission.

Similar problems occur in two other studies cited predominantly in the NPR, the studies of Cummings et al. and Emmons et al. In these two studies, the authors actually incorporated an analytical measure of ETS exposure in addition to subjective responses regarding potential exposure duration.

The NPR failed to cite the analytical data from either study and chose instead to rely entirely on potential exposure duration as an inappropriate surrogate for exposure. Additional details on these two studies will be provided later.

As I stated earlier, OSHA is obligated to determine how much smoke is actually in the workplace air before they can move to regulate it.

Moving now to my third point, let's see how OSHA went about determining this critical information.

In an attempt to establish the concentration of ETS in the workplace, OSHA has relied on a limited number of outdated, non-representative and extreme data sets. Proper consideration of recent, representative personal monitoring studies demonstrates that the NPR has overestimated typical worker exposures at least 10 to 100 times.

In the NPR, four studies are cited to support the conclusion that the average RSP level during smoking in smoking buildings was 262 micrograms per cubic meter while in non-smoking buildings the RSP levels averaged 36 micrograms per cubic meter. This is slide 13.

Of the four citations given in support, one is obviously an incorrect citation and cannot be verified. The other three include a report by First and two from Repace and Lowery. These three studies describe RSP measurements made with a portable piezobalance over a decade ago.

There are four reasons why this is important. These are outlined in slide 14.

First, without getting into too much technical detail about this measurement device, suffice it to say that the portable piezobalance is most often used for short-term measurements. This is because it requires substantial maintenance between measurements to keep it operating adequately.

Accordingly, the levels reported are in general short-term, peak concentrations, not long-term, time averaged concentrations, the latter of which is needed to truly characterize exposure. For example, Repace and Lower report most sampling times are only 10 to 20 minutes.

Second, the suitability of this type of measurement device has been questioned by First, one of the authors cited for the measurements. In his opinion, First says, and I quote, "This apparatus was not designed for this type of service and lacks the sensitivity and precision needed to sense the small incremental concentrations attributable to tobacco smoking in public places."

Third, the portable piezobalance is meant to measure area concentrations, not personal exposures of people going about their normal activities.

And, fourth, it is readily apparent that typical smoking behavior in public in the 1990s is much different from smoking behavior when these measurements were taken in the late 1970s and early 1980s.

While measurements of the magnitude reported in these studies are possible, they are far from being typical, at least in the 1990s. These extreme values need to be viewed in the context in which they were generated and also compared to modern day, realistic RSP concentrations in workplaces with and without smokers.

In contrast to these area monitoring values cited in the NPR, I presented in my written submission the personal monitoring results from two recent, representative, population-based surveys of workplace RSP levels. These data are summarized here in slide 15.

One of these is a 24-hour personal monitoring study which we conducted in Mt. Laurel, New Jersey in 1992. A detailed report of this study is included as an appendix to my written submission. The other, a recently completed nationwide 24-hour personal monitoring study being conducted by Oak Ridge National Laboratory is also described in a report by Jenkins et al. to the docket.

In both of these studies, total RSP is measured in both smoking and non-smoking workplaces. In addition, and more importantly, both studies also include direct measurements of ETS RSP or the amount of RSP attributable to ETS based on solanesol determination. The exact details of the method used, sometimes referred to as Sol-PM, are described in my written submission and in the literature.

The results of these two studies are shown in slide 15, along with the values cited in the NPR.

These results show the total RSP levels in the average smoking workplace are at least five times lower than stated in the NPR. More important, and more relevant, the largest and most recent data set of Jenkins et al. also shows that ETS RSP levels are over 200 times lower than assumed in the NPR.

In summarizing workplace data for RSP, the NPR concludes that RSP is elevated 10 to 100 times during smoking. There are some extreme data in the literature that can be used to support this contention. However, as slide 16 correctly summarizes, recent data from truly representative studies show that total RSP is elevated only by factors of two to three, not 10 to 100, and, of course, only a fraction of this RSP is attributable to ETS.

OSHA's summary statement at best represents short-term, peak concentrations in selected environments measured more than 10 years ago and does not reflect the current status of long-term exposures in workplaces in the U.S. that permit smoking.

These typical increases in RSP levels of only two to three times in workplaces that permit smoking are also supported by two additional studies which are cited in the NPR. These data are summarized in slide 17.

In a study by Spengler et al., personal RSP exposure levels differed by less than a factor of three, 34 versus 13 micrograms per cubic meter between exposed and non-exposed groups.

In a study by Sexton et al. for workplace exposures, reported times in excess of two hours per day of ETS exposure resulted in a personal RSP level of 39 micrograms per cubic meter, compared with 30 and 34 micrograms per cubic meter for those reporting up to two hours per day and no workplace exposure respectively.

While the first of these studies suggests RSP increases due to smoking of less than a factor of three, the latter study shows virtually no increase in RSP exposures due to working with smokers.

In spite of these findings, the NPR concludes that the data cited are sufficient to support OSHA's risk assessment. OSHA summarizes the limited ETS exposure data presented in the NPR with statements regarding nicotine and ETS RSP and these are reproduced here in slide 18.

For nicotine, the NPR claims that the "ETS-nicotine exposures of the average worker appear to be of the order of 5 to 10 micrograms per cubic meter ..., and for the most exposed workers, [are] 50-100 micrograms per cubic meter."

"For EST-RSP, exposures are about tenfold that of the nicotine levels."

The summary conclusions drawn from the database assembled in the NPR are compared in my written submission to the largest, most representative, most recent and most relevant database on workplace exposures ever assembled, that is, the study by Oak Ridge National Laboratory.

Slide 19 compares the OSHA cited value for nicotine exposures with nicotine exposure levels actually measured by Oak Ridge. As seen, the NPR has overestimated average worker exposure levels 25 times and has overestimated most exposed worker exposure levels approximately 10 times.

Shown here in slide 20 is a comparison of the OSHA-cited value for ETS RSP with exposure levels actually measured recently by Oak Ridge.

As I said earlier, from the Oak Ridge study, ETS RSP can be determined for each workplace based on solanesol, the best available tobacco-specific indicator of the ETS contribution to RSP.

Based on these data, OSHA has overestimated average worker exposure levels approximately 250 times and has overestimated the most exposed worker exposure level approximately 15 times.

In summary, as demonstrated here, the NPR does not adequately address worker exposure to ETS and, as a result, OSHA has incorrect information regarding how much ETS American workers typically encounter in the workplace.

My fourth point in this section on ETS exposure addresses OSHA's comparison of workplace exposures for people who work with smokers to home exposures for people who live with smokers.

The NPR attempts to establish that ETS concentrations in these two environments are equivalent. An obvious question to ask is why would such a link be important?

The answer is quite simple. The attempt to equate exposures between home and workplace situations is prerequisite for attempting to substitute spousal smoking epidemiology for workplace exposure epidemiology.

Consistently and incorrectly the NPR asserts that ETS exposures are equivalent for living with and working with smokers. In this section of the NPR, OSHA appears to be arguing that ETS concentrations are equivalent for living with and working with smokers. Even if concentrations were equivalent in both venues, which as I will show momentarily they aren't, true exposures remains significantly different.

Slide 21 recalls the correct definition of exposure, that is, exposure equals concentration times time. It is straightforward to see that if the concentrations are equivalent in two environments, the ratio of exposures becomes equal to the ratio of the time spent in the two environments.

So the relevant issue now becomes how much time does a typical worker spend at work and at home? Don't forget to include days off, weekends, holidays and vacation days.

Most workers spend more waking hours at home than they do at work. Plus, they have sleeping time which hopefully also occurs at home. In a year, most people will spend far more time at home than they do at work. For example, a year consists of 8760 hours. The typical work year is only 2000 hours.

So how does this affect OSHA's proposition?

Well, even if you were to assume equal ETS concentrations at home and work, long-term exposures due to living with smokers will be far greater than long-term exposures due to working with smokers. However, that's a worst case scenario.

Let's now answer the question of how home ETS levels compare with workplace ETS levels, assuming of course there are smokers present in both places.

In this type of data analysis, it's not important to list the actual air concentrations or exposures in the home and workplace environments. What really matters, and it's a simpler way to look at the data, is the ratio between the two scenarios.

True exposure ratios between living and working with smokers are available from several recent population based exposure studies. These data have been collected by using either long-term personal monitoring, as was done in the survey we conducted in Ohio in 1991, or by using separate personal monitoring in the workplace and at home for one full 24-hour period, as was done in the survey we conducted in New Jersey in 1992 and in the recently completed study conducted by Oak Ridge National Laboratory.

The relative weekly exposure estimates obtained in Columbus, Ohio can be calculated for three different ETS markers and are shown here in slide 22. These include the week-long personal monitoring for nicotine and 3-ethenylpyridine and also salivary cotinine measured four times during the one-week period.

The exposure ratios between non-smokers who were exposed only at home and non-smoker who were exposed only at work are 11 for 3-ethenylpyridine, 10 for nicotine and 6 for cotinine. Remember, these are exposure ratios. They indicate that non-smokers who worked with smokers experienced 6 to 11 times lower exposure than non-smokers who live with smokers. These data were submitted to OSHA in May 1993 and were published later that year.

Weekly estimate of home compared to work exposures from the New Jersey study are described in detail in my written submission. These data show average exposures are 2.9 times lower for working with smokers based on 30 different volatile organic compounds.

Additional data from that study showed 3.6 times lower average exposure for working with smokers based on five different particulate phase markers.

Based on three vapor phase ETS markers and four different indicators of the particulate phase of ETS, Jenkins et al. reached similar conclusions from the Oak Ridge study. Quoting from Jenkins in slide 23, "Although participants perceived their greatest exposures to ETS to occur in the workplace, in fact exposure to ETS when living with a smoker is about a factor of five greater than that received in a smoking workplace."

As summarized in slide 24, all of these data I have just presented which in total were gathered over the last three years in 14 different U.S. cities paint a very consistent picture of exposure due to working with smokers versus living with smokers. Using over 30 different markers of ETS exposure, these studies show that workplace exposures due to working with smokers are three to 11 times lower than home exposures due to living with smokers.

Taken together, these data indicate that true exposure to ETS between living with smokers and working with smokers is minimally three times lower and is most likely five to 10 times lower in the workplace.

Although these two studies were not cited by OSHA, let's move on to studies that were included in the proposed rule.

In two of the predominant studies cited in the NPR, the researchers asked subjects to guess how much ETS they were exposed to. Then they attempted to verify exposures by measuring cotinine. However, the NPR does not cite the cotinine data from either study. A proper consideration of these data actually refutes OSHA's claim regarding the magnitude of workplace exposure relative to home exposure.

Let's now consider what these two studies cited by OSHA actually show.

In the study by Cummings et al. and the data re-analysis requested by OSHA, 77 percent of workers reported ETS exposure at work. In the NPR, OSHA states, "This further analysis indicates that the workplace is a significant source of ETS exposure for non-smoking employed people."

Please recall the correct definition of exposure and realize that in the re-analysis there is no consideration for duration of exposure or ETS concentration. This reveals that the assertion put forth in the NPR is incorrect regarding what the Cummings et al. study actually shows.

Simply based on the percentage estimates, the only valid conclusion is that the workplace can be a significant source of potential exposure among the studied population.

To begin to evaluate objectively whether the workplace is a significant source of true exposure, an analytical measurement of exposure is needed.

In the re-analysis of the Cummings data requested by OSHA, urinary cotinine concentrations are provided according to the four exposure scenarios of all combinations of exposure at home and work. These data are reproduced in slide 25.

Clearly the effect of home exposure on cotinine levels is predominant. In fact, people reporting exposure to ETS only at work and not at home had the lowest cotinine levels of any group in the study, even lower than people reporting no exposure either at home or at work.

Cummings provides a peculiar explanation in an attempt to minimize the results of this re-analysis. This is slide 26. He states, "This analysis of cotinine values suggests that home exposure is more important than work exposure. However, this result is misleading since many of the subjects included in our study took time off from work to attend our clinic. Thus, cotinine values would, of course, be influenced more by home and public location exposures, not workplace exposure."

Such an explanation is hardly plausible, since cotinine has a fairly long half-life of about 17 hours and presumably most subjects would have been away from work for only one to two hours before providing the urine sample.

Assuming a simple model of first-order decay with a cotinine half-life of 17 hours shows that cotinine concentrations would decrease less than 8 percent after two hours away from the source of ETS exposure. Moreover, any decrease in cotinine concentrations incurred by coming to the clinic is equally applicable to any source of ETS exposure, whether it is at home, at work or in public places.

Slide 27 contrasts two very different views of what this study actually shows. The NPR states that the data of Cummings et al "present results to show significant workplace exposures to ETS."

In fact, Cummings' re-analysis of the cotinine data shows one thing quite clearly. That is, and this is very important, essentially there is no discernable effect of workplace exposure.

As additional evidence, I would like to quote from the second study cited in the NPR, that of Emmons et al. This passage on slide 28, however, was not cited in the NPR. "When those with and those without a smoker in the household were examined separately, we found that subjects who lived with a smoker received more exposure in the home than in the workplace."

In further describing the study of Emmons et al., OSHA claims that this study substantiates the magnitude of workplace exposures. The NPR also states, "For example, Emmons et al. found that the majority of ETS exposure occurred in the workplace." As I said before, OSHA appears to be using an incorrect definition of exposure.

As a consequence, the NPR reaches an inappropriate conclusion. What Emmons et al. actually show, like CAPS and like Cummings et al. is that the reported duration of ETS exposure in the workplace is higher than in the home.

Remember your guess as to how far I drove yesterday in an hour and a half?

So what does a proper consideration of the Emmons et al. study show?

Well, it actually raises two important issues. The first important issue is outlined here on slide 29. Emmons et al. report two components of potential exposure. That is, exposure duration in minutes per day and exposure intensity rated as near versus far. While the subjects rated approximately three times the potential exposure duration at work than at home, their ranking of intensity was approximately two times greater in the home.

Clearly true exposure is a combination of both duration and intensity or concentration. It is impossible to tell from the subjective data the net effect of these two differences.

Second, Emmons et al. did make an analytical measure of ETS exposure. However, this is not mentioned in the NPR. In a subsequent report on the same study, Emmons et al. give cotinine data.

Slide 30 reveals the original authors' conclusions regarding their own study. According to the authors, and again I quote, "Volunteers who lived with but did not work smokers had significantly higher cotinine concentrations than volunteers who were exposed to smokers only in the workplace. In addition, volunteers who had regular ETS exposure at home and at work had significantly higher cotinine ... than volunteers whose primary exposure was at work."

Regarding the same data set in another report, Emmons et al. state, "Subjects who lived with a smoker received more exposure in the home than in the workplace."

The cotinine results from Emmons et al. are reproduced in my written submission, along with the results from Cummings et al. and the more recent studies I described earlier.

The similarity of cotinine pattern among these five studies is striking. As emphasized on slide 31, the overwhelming conclusion from these data is this: ETS exposure at home for those living with smokers is a substantially larger contributor to elevated cotinine levels than ETS exposures at work for those working with smokers.

What happens when these data are compiled with new data from nationally representative monitoring studies that rely on multiple markers, including better markers, of ETS exposure?

The resulting data set portrays a very consistent pattern of ETS exposure that demonstrates working with smokers results in five to 10 times less ETS exposure than does living with smokers. This conclusion differs substantially from the position as stated in the NPR that exposures in the two venues are equivalent.

OSHA also states in the NPR the database on nicotine concentrations shows significantly higher average exposures in workplaces than in residences.

Before moving on to a discussion of what the nicotine data show, it's appropriate to pause for just a moment and remember why so much attention is being focused on comparing home exposures with workplace exposures.

Quoting from the NPR on slide 32, "Thus risk estimates based on residential exposures are expected to accurately reflect occupational risks in most workplaces and possibly underestimate the risks in some workplaces."

Dr. Sears will address this topic further in his presentation.

Again, more recent, more relevant and more representative data which used nicotine concentration as an index of exposure in fact showed just the opposite.

Slide 33 reminds us that in the Oak Ridge study Jenkins et al. report nicotine exposures for smoke exposure away from work and smoke exposure at work venues. The ratio of the medians shows differential exposure to be a factor of 6.8 higher away from work.

This compares to the exposure ratio of 10.3 based on nicotine monitoring previously reported to OSHA from the study we conducted in Columbus, Ohio.

Thus, using an additional marker, nicotine, as the index of ETS exposure shows similar findings.

Working with smokers results in seven to 10 times lower exposure than living with smokers.

I would like to turn now to my final point on exposure and that is how many workers are actually exposed to ETS in the workplace?

Slide 34 shows that to estimate the prevalence of ETS exposure among the U.S. non-smoking workforce in the NPR OSHA attempts to discredit the largest and most representative survey done to date, that is, the NHIS survey. Instead, the NPR places far too much significance on a much smaller and clearly inappropriate study.

Proper consideration of these data reveals that OSHA has actually used the overall best estimate of population exposure as the lower bound in the risk assessment and suggests that the upper bound used in the NPR is biased high by at least a factor of two.

OSHA has not established a lower bound for the risk assessment. Let me briefly explain why this is true.

The NHIS, or the National Health Interview Survey, was conducted by the Centers for Disease Control and is the largest and most representative survey of its type done to date.

This survey found that 18.81 percent of the American workforce reports ETS exposure at work. The NPR states that this 18.81 percent "may be an underestimate of frequency of exposure in the workplace because it is based solely on self-reported information and the question was not very specific in defining immediate work area."

On its face, the information from NHIS would appear to be the definitive database for OSHA to use in characterizing the prevalence of occupational exposure to ETS.

On slide 35, let's look at the advantage and disadvantages of this study and try to figure out what, if anything, is wrong with it.

Obvious advantages of NHIS include:

(1) It's a very representative study. In fact, it appears to be the exact population OSHA is trying to characterize.

(2) It is a very large survey, with over 7000 non-smoking respondents.

(3) A very recent survey conducted in 1991.

(4) It is the same survey used in the NPR to estimate the percentage of non-smoking workers in the U.S. that would be covered by this proposed standard.

OSHA levies two criticisms against the survey. The first criticism is that NHIS is based solely on self-reported information. How does one expect to find out how many workers in the U.S. are potentially ETS exposed if not by self-report? This has to be the starting point for obtaining the necessary information.

As is obvious from a careful scrutiny of the other two data sets OSHA has relied predominantly upon, that is, the studies of Cummings et al. and Emmons et al., which I have already described briefly, self-reported exposures actually tend to overestimate true exposures. On this first point, the NHIS study results would seem to be biased high, if anything, rather than low.

The second criticism in the Notice of Proposed Rulemaking is that the NHIS study was not very specific in defining immediate work area. The survey asked participants whether during the past two weeks anyone had smoked in their immediate work area. It would appear that this definition of exposure would actually include a significant number of respondents who are only incidentally or infrequently exposed.

Based on the question asked, exposure due to someone walking through one respondent's work area with a lit cigarette once during a two-week period gets the same weighting in this survey as the exposure for another respondent who might share an enclosed office with a smoker every day.

Also, it seems likely that most workers have a reasonably good concept of their immediate work area. To OSHA's second criticism, the NHIS study results would again seem to be biased high, if anything, and not low.

It appears that OSHA has exactly what it needs in this single study, a midpoint estimate of the prevalence of occupational exposure.

After evaluation of its strengths and weaknesses, this still appears to be the definitive database for OSHA to use in characterizing the prevalence of occupational exposure to ETS.

Adding further support to the prevalence of exposure information in this study is the similar prevalence of ETS exposure among the non-smoking workforce that we found in two studies which we conducted in 1992.

In one, a nationwide, random digit dialing telephone survey of non-smoking females, we found that 16 percent of working females reported exposure to ETS at work.

In the second, a mall intercept study, which was conducted in nine U.S. cities and included collection of saliva for cotinine determination, we found that 18 percent of confirmed non-smoking females reported exposure to ETS from their co-workers in their immediate work area.

These data, in combination with the NHIS data, demonstrate that approximately 16 to 19 percent of the U.S. workforce appear to be ETS exposed in their place of work.

Another source considered in the NPR for defining non-smoker ETS exposure prevalence in the workplace is the work published by Cummings et al. As shown on slide 36, OSHA states, "A recent re-analysis of the data file showed that among the non-smoking, currently employed subjects, 48.67 percent ... reported exposure to ETS at work and not at home." This is the number that OSHA uses as the estimate of the upper limit of the number of non-smokers exposed to ETS at work.

Let's look at the advantages and disadvantages of this study on slide 37. The only obvious advantage of this study is that it employed one analytical estimator of ETS exposure. That is, cotinine.

The obvious disadvantages of the Cummings et al. study are several, some of them are major. They include the following:

(1) It is a very non-representative study, deriving all study participants from a cancer screening clinic in Buffalo, New York by inviting them to participate in a study on ETS.

(2) It is a relatively small study, incorporated 339 participants.

(3) It is a relatively old study, conducted in 1986.

(4) According to Cummings et al. it "over represented females and whites and under represented persons below 40 years of age."

(5) It also was not very specific in defining exposure.

This is clearly a non-representative study, both in the selection criteria used for contacting potential subjects in a cancer clinic and the way subjects were invited to participate, by informing them it was an ETS study.

OSHA selects from the 78 percent who reported ETS exposure the 48.67 percent who reportedly were exposed at work and not at home.

The mean urinary cotinine level determined for the 48.67 percent of the study group is actually the lowest of all the groups, even lower than the group reporting no exposure either at home or at work.

So in essence, as I've outlined here on slide 38, OSHA uses as the upper limit for prevalence of occupational ETS exposure a number that is obtained from a group of study participants that on average had no detectable ETS exposure above background.

The NPR also refers to Emmons et al. as justification for the upper limit on the number of workers exposed.

OSHA states, and I quote, "Seventy-six percent of the subjects reported being regularly exposed to ETS in the workplace. The percentage of subjects reporting exposure at work is similar to that found by Cummings et al."

The issue of study representativeness, that is, being representative of the U.S. population at large, is critical in a determination of the maximum number of potentially exposed non-smokers in the U.S. workforce. It is true that the percentages of the study populations reportedly exposed at work are similar between Cummings, 78 percent, and Emmons, 76 percent. However, the self-selective nature of recruited subjects was also similar in both studies.

Cummings et al. invited attendees at a cancer screening clinic to "participate in a study on ETS." Likewise, the subjects in Emmons et al. "were volunteers who responded to advertisements for a study on passive smoking." Also, the Emmons et al. subjects were selected to yield a wide range of ETS exposures, not necessarily to be representative. In fact, 106 of the 186 subjects represent just nine different work sites.

In no way can the reported prevalence or magnitude of ETS exposure among these study participants be generalized to the entire non-smoking U.S. workforce.

Given the nature of recruitment in both studies, it is surprising that even larger percentages did not report ETS exposure.

Think about this. Why would people volunteer to participate in a study on driving habits if the didn't drive?

Now think about this. Why would people volunteer to participate in a study on ETS exposure if they did not believe they were ETS exposed?

The fact that these two studies are not representative is no secret. In their paper, Cummings et al. plainly state that, "Given the self-selected nature of the study population and potentially limited generalizability of results ...."

Likewise, Emmons et al. describe their study population as "a motivated sample of convenience"

Again, proper consideration of these data reveals that the NPR has actually used the overall best estimate of population exposure as the lower bound in the risk assessment and suggests that the upper bound used by OSHA, that is, the 48 percent, is too high by at least a factor of two.

By under-emphasizing the most representative study and over-emphasizing an inappropriate study, the NPR has not established an accurate representation of the prevalence of ETS exposure among the non-smoking U.S. workforce.

Slide 39 concludes that the NPR's assessment of the non-smoking working population's exposure to ETS while at work is in accurate. Reliance on outdated, non-representative, extreme and limited data sets has led to substantial overestimation of true ETS exposure in typical U.S. workplaces.

As a result, OSHA concludes that non-smoker exposures due to living with smokers are equivalent to non-smoker exposures due to working with smokers.

True differences in exposure between these two populations are shown to be three to 10 times lower for occupational exposures in more recent studies that have been submitted to the docket.

In the proposed rule, the absolute magnitude of workplace ETS levels has been inferred from an extremely limited and biased data set derived mostly from short-term, peak concentrations measured by area monitoring over a decade ago.

Typical workplace levels appear to be 25 to 250 times lower than assumed by OSHA. By simply relying on the studies cited and data sets in the NPR, OSHA has significantly overestimated both the magnitude and prevalence of the workforce population's exposure to ETS.

In short, OSHA's proposed workplace ETS rule is based on misinformation. OSHA has not measured ETS exposure, OSHA doesn't know how much ETS is in the typical workplace and OSHA doesn't know how man workers are exposed to ETS at work. Therefore, OSHA's proposed workplace ETS regulations are unjustified.

Thank you.

JUDGE VITTONE: Thank you, Dr. Ogden.

Dr. Ogden's printed statement will be identified for the record as Exhibit 228. His slides will be identified as Exhibit 229.

(The documents referred to were marked for identification as Exhibit Nos. 228 & 229 and were received in evidence.)

MS. SHERMAN: Your Honor, would this be a good time for break?

JUDGE VITTONE: We'll take a five minute recess. And then we'll come right back. Please. I'm going to ask you to be precise on the time.

Off the record.

JUDGE VITTONE: Dr. Coggins, would you introduce the next speaker, please?

DR. COGGINS: Thank you, Judge.

Now Dr. Nelson will talk about the limitations in using nicotine and cotinine to precisely predict ETS exposure.

Dr. Nelson.

JUDGE VITTONE: All right. Dr. Nelson.

DR. NELSON: Thank you.

Good morning. I am Dr. Paul Nelson. I'm an analytical chemist with R.J. Reynolds tobacco company. For the past seven years I've been studying the chemistry and fate of environmental tobacco smoke, or ETS.

Specifically, my research has focused on three areas: measuring a quantity and behavior of ETS constituents; testing methods for ETS reduction and removal from indoor spaces; and third, helping to determine human responses to ETS.

(Slide Presentation)

Today I will discuss some problems with the use of nicotine as a quantitative marker for ETS levels and the use of nicotine and cotinine as quantitative biomarkers for ETS exposure.

In the proposed rule, OSHA claims that nicotine and its metabolized cotinine are good quantitative biomarkers for ETS. OSHA states that "nicotine and cotinine accurately predict the quantity of ETS exposure that has taken place over a recent time span." This claim is not justified by the existing body of evidence.

Why? There are two basic problems with giving nicotine and cotinine as quantitative biomarkers for ETS exposure.

First, nicotine does a poor job of predicting exposure to ETS constituents in the field, or "real world." Second, nicotine and cotinine concentrations in biological fluids do not accurately quantify past exposures to ETS. That is to say, the amount of nicotine and cotinine in your body right now cannot tell me when you were exposed to ETS, where you were exposed, or how much ETS exposure you had. As a result, nicotine and cotinine cannot be used to quantify ETS exposure. And they cannot provide a reliable basis for quantitative risk assessment.

Nonetheless, nicotine and cotinine concentrations can provide useful information in many circumstances. Especially in distinguishing between smokers and non-smokers.

Slide 2 illustrates the fact that nicotine exposure of smokers is much greater than the nicotine exposure of non-smokers. In general, regular smokers of more than a few cigarettes per day are easily distinguished from even the most heavily exposed non-smokers on the basis of nicotine or cotinine measurement in plasma, urine or saliva.

In addition, when looking at groups of individuals, cotinine measurements can distinguish groups who are not exposed to ETS from groups who are moderately or heavily exposed.

Nicotine is one of the easiest markers to measure in the field. It is often used to characterize the amount of cigarette smoke in the air. In conjunction with data from other ETS markers nicotine data can help provide a general indication of ETS levels. Furthermore, in controlled exposure studies nicotine and cotinine in body fluids can be used to determine the uptake of nicotine from the air. However, they do not necessarily indicate uptake of any other ETS constituents.

Before I demonstrate why nicotine and cotinine cannot be used as quantitative markers, or biomarkers, for ETS exposure, I should define what I mean by quantitative. For a quantitative relationship to exist, a numerical relationship must exist. In the case of ETS, a quantitative marker should predict the concentrations of other ETS constituents, both accurately -- that is, give the right number -- and precisely, which means reproducibly.

For example, the odometer in a car provides a quantitative indication of the number of miles traveled, regardless of the automobile used or the route followed. Terms such as "a lot," "far," or "not much" mean different things to different people and cannot be numerically related to the distance a car has travelled. In a similar fashion, a number of terms frequently used to assess ETS exposure, such as "some," "a little," and "a lot" have no numerical basis and therefore cannot be used to quantify ETS exposure..

So why can't we use nicotine and cotinine to quantify ETS exposure? First, I'll talk about the use of nicotine for quantifying atmospheric levels of ETS. Then I'll discuss the problems with using nicotine and cotinine as quantitative biomarkers.

Nicotine fails to satisfy commonly accepted criteria for an ETS exposure marker. These criteria, which have been listed by the National Research Council, are shown on Slide 3. And they simply follow common sense.

First, the marker should be specific to ETS. Other sources of the marker should not be strong enough to substantially affect the marker concentration.

Second, the marker should be easy to detect.

Third, all cigarettes should have similar amounts of the marker.

And, finally, the ratio between the marker and the compound of interest should be constant, regardless of the brand smoked, the sampling location, ventilation rates or furnishings of an indoor space.

The first criteria, that a marker be unique to ETS is probably met in areas where smoking usually takes place. However, nicotine has reversible adsorption properties. That is, after ETS generation, nicotine easily sticks to surfaces such as walls, furniture and clothing. It is later re-emitted from these surfaces in the absence of other ETS constituents. So in areas where smoking takes place only occasionally, nicotine can sometimes be detected in the absence of other ETS constituents.

For example, let's say I had a party at my house last night where several people smoked. Even though no one is smoking there now, there's a good chance I'll be able to detect nicotine in the air today. So I could be exposed to nicotine without being exposed to other ETS components.

This characteristic appears unique to nicotine because no other ETS constituent is known to exhibit the same degree of adsorption or re-emission over a long time period.

The presence of nicotine in the absence of ETS has been demonstrated by a number of studies, including chamber and field studies where unexpected background nicotine measurements have been recorded in the absence of smoking. In addition, many researchers have demonstrated that nicotine decays or disappears from the air more rapidly than other ETS constituents. This provides further evidence of nicotine adsorption.

Because of this, nicotine and cotinine concentration could be found in biological fluids of individuals without their being exposed to ETS. For people with no or low ETS exposure, the detection of nicotine or cotinine in the body could be improperly ascribed to ETS exposure.

Now, let's look at the second NRC criterion, that the markers should be easy to detect even at low smoking rates. Does nicotine meet this criteria? Certainly. In fact, one of the main reasons nicotine has been measured so often is its ease of detection.

The third criterion, that cigarettes emit similar amounts of nicotine, is demonstrably untrue. There are currently over 1,300 different brand styles of cigarettes on the U.S. market. Different styles, the same brand of cigarettes -- for example, Winston Box, Winston 100s or Winston Lights -- don't necessarily contain the same blend of tobacco or have the same combustion characteristics. In other words, they may generate different amounts of ETS.

As you can see here on Slide 7, nicotine yields of the top 50 U.S. brand styles show wide variation. An investigator testing a specific brand style of cigarette cannot simply assume that the cigarette tested represents the entire brand family. Likewise, an investigator who tested a small segment of the market, say 10 different cigarettes, cannot claim that such a sample is representative of the entire market.

The fourth criterion listed by the NRC, that there should be a consistent ratio between the marker and the contaminant of interest, is probably the most important criteria because it establishes the link between the marker and the agent of concern. This criterion has three elements.

First, there must be a contaminant or contaminants of interest.

Second, the ratio between the marker and the contaminant of interest must be consistent across a range of tobacco products.

And third, the ratio between the marker and the contaminant of interest must be consistent across a range of commonly encountered environmental conditions.

For many reasons, the use of nicotine as a marker for ETS fails to meet this criterion.

The first problem is this: If you don't know what you're looking for, how can you possibly choose a marker for it?

OSHA tends to deal with ETS as if all of its components behaved in exactly the same way. But they don't. Approximately 100 compounds have been identified in environmental tobacco smoke. Experimental data indicate that each of these constituents may potentially behave differently when introduced into indoor spaces. And nicotine's unique decay characteristics make it unsuitable as a marker for most identified constituents of ETS.

It's generally thought that if there are any health problems associated with ETS they might be related to compounds in the particulate phase. Like ETS in general, ETS RSP, or respirable suspended particles from ETS, is a complex mixture. Very little is understood about its composition. Even less is known about the relationships among ETS RSP constituents.

In the proposed rule OSHA has suggested that nicotine is a quantitative marker for ETS exposure. Therefore it must also be quantitatively related to ETS RSP exposure. But nicotine, a vapor phase component of ETS, is not a good marker for ETS RSP.

Why? First, nicotine and RSP yields of cigarettes are poorly correlated. Recently, Reynolds Tobacco conducted a study of the top 50 cigarette brand styles -- and this was the largest number of brands surveyed to date, accounting for 64 percent of the U.S. market.

As you can see here on Slide 9, emission rates for nicotine vary considerably. The sales-weighted average nicotine yield was about 1.6 milligrams per cigarette. But even near that average, as you can see here on this chart, RSP yields vary from below 10 to up close to 20 milligrams per cigarette. The nicotine yields of individual brands vary between 1 and 2.3 milligrams per cigarette. A regression of these data indicates that nicotine and RSP yields are only poorly correlated, an "r" value of .47. And that's "r" as opposed to "r"-squared. And the relationship between ETS RSP yield and nicotine yield cannot be expressed as a simple ratio.

A second problem with using nicotine as a marker for ETS is that nicotine has not been demonstrated to be found in a consistent ratio to ETS RSP across a range of environmental conditions typically encountered. Studies at our laboratory have shown that the relationship between nicotine and ETS RSP can be consistent in a controlled environment test chamber operated at a single ventilation rate. However, as the data I have provided to OSHA previously show, that relationship breaks down at different ventilation rates or when measurements are taken in the field.

In a single environment, such as an office or a restaurant, one might expect that the relationship between nicotine and ETS RSP would remain constant.

Does it? Sometimes, yes. But often not.

During her testimony at this hearing, Dr. Katherine Hammond discussed the study she and Dr. Brian Leaderer conducted that purports to demonstrate that nicotine is a good predictor for RSP concentrations. These study has two main failings as applied to OSHA's proposed rule.

First, it was performed in homes, not workplaces. Ventilation rates in homes are generally low. Less than one air change per hour. Ventilation rates in workplaces are typically higher. This is important because the ventilation rate is a key factor associated with the ratio between concentrations of nicotine and ETS RSP.

Second, nicotine concentrations measured in homes do not appear to be highly predictive of RSP concentrations in this same environment. This is especially true at low nicotine concentration, which happen to be those most often encountered.

In recent years our company has conducted studies in several buildings owned and operated by Reynolds Tobacco. The methods and results of those studies have been submitted to the docket. I would like to discuss those results that pertain to the use of nicotine to predict exposure to ETS RSP.

In one study, RSP and nicotine were measured in several locations on a single floor of a large office building. The relationship between RSP and nicotine for the different sampling locations is presented on Slide 10. RSP and nicotine were not significantly correlated at any of the five sampling locations. That is, they were independent of each other at every location.

Slide 11 incorporates the data for all sampling locations. Looking at the overall data, there was no statistically significant correlation between nicotine and RSP.

We also looked at the relationship between nicotine and ultraviolet particulate matter, or UVPM. UVPM is more specific that total RSP as a marker for ETS RSP. In discussing field measurements, I will use UVPM as a surrogate for ETS RSP.

Slide 12 shows the relationship between nicotine and UVPM for each of the sampling locations. UVPM was statistically significantly correlated with nicotine in only two of the five sampling locations. Where correlations were statistically significant, they were also poor -- "r" values were .63 and .64 for the two locations.

Slide 13 incorporates all the data points from all locations. The overall correlation between nicotine and UVPM was statistically significant. But the precision of the prediction was poor. So significant errors in accuracy could occur if nicotine were used to predict ETS RSP concentrations in the office building.

Why is that? Well, let me give you an example.

If I were in this building -- and it's the one illustrated on the slide -- and I were exposed to 2 micrograms per cubic meter of nicotine, the corresponding amount of ETS RSP I was exposed to could have ranged from as low as 5 to as high as 25 micrograms per cubic meter, a five-fold difference.

Similar results were obtained in a study of an office area in a different building. The existence of data sets that show a correlations between nicotine and RSP do not provide sufficient justification for assuming that nicotine is found in constant proportion to other ETS constituents within or across environments.

To put it very simply, if you see a couple of red cars on the road, you can't assume that all cars are red. Many data sets, including references cited in the proposed rule, demonstrate the nicotine is not found in constant proportion to RSP within single environments. It would be an error for OSHA to assume that the ratios between nicotine and other ETS constituents are consistent, or that nicotine is a good quantitative marker for ETS RSP.

Slide 14 shows the results of 480 concurrent area measurements of nicotine and RSP from workplaces in 15 U.S. cities. As you can see, the correlation between nicotine and RSP is not significant, and also very poor.

In an effort to compare Dr. Hammond's home data and the data derived in workplaces, I plotted the workplace data on the same axes that Dr. Hammond used in her submission to the docket. As you can see... And these are the axes: 0 to 20 micrograms per cubic meter nicotine and 0 to 200 micrograms per cubic meter RSP.

As you can see on this inset graph, the data are gathered in what appears to be a random pattern. In addition, many individual data points correspond to a wide variety of RSP and nicotine ratios. Because of the diversity in U.S. buildings in terms of the interior surface and ventilation characteristics, any correlation between RSP and nicotine derived from small sample sets or from samples collected in similar micro-environments, such as homes, cannot be applied across the range of environments encountered in the U.S. workplace.

During her testimony, Dr. Hammond suggested that variable background concentrations of RSP from sources other than ETS explain the variety of RSP-nicotine ratios observed in the field, and especially the variation and low nicotine concentration. It is true that non-ETS RSP can lead to unusually large ratios in some circumstances. However, it is unlikely that background RSP can explain all the variations in RSP to nicotine ratio.

The two data sets I'm about to describe illustrate this point. Back here on Slide 14, and I'm referring now to the inset on Slide 14, at nicotine concentrations of 4 micrograms per cubic meter there's a series of data points where the RSP concentrations are between 150 and 200 micrograms per cubic meter. The RSP to nicotine ratio is large, approximately 35 to 1.

Using Dr. Hammond's model, which is based on a ratio of approximately 11 to 1, one would expect ETS RSP to approximately 43 micrograms per cubic meter. That would require an ambient background in excess of 100 micrograms per cubic meter. A concentration much greater than Mr. Hammond testified that she would typically expect.

Obviously, many data points on this graph demonstrate the possibility of RSP-nicotine ratios greater than 11 that are unlikely to be explained solely on the basis of high background levels of RSP. Likewise, many data points correspond to a ratio much less than 11.

In the group of points between 7 and 8 micrograms per cubic meter, there are several data points where RSP concentration is less than 50 micrograms per cubic meter. For these points, the RSP-nicotine ratio would average about 3 to 1.

There is no way to explain this low ratio on the basis of background RSP. Dr. Hammond has correctly observed that background RSP can play havoc with RSP-nicotine ratios. And it can have a large influence. But it is not the sole reason that RSP-nicotine ratios are variable. And it may contribute only a small amount to the variation of ratios seen across a wide range of environments.

What other factors are important? Ventilation patterns, and by that I mean air distribution, as well as ventilation rates, or the amount of air you put into a space, and interior surface characteristics can also play a large ratio in the inconsistent relationship between ETS RSP and nicotine.

Now let's look at a second data set that illustrates that background RSP is not solely responsible for the wide variability in RSP-nicotine ratios.

As I noted earlier, ultraviolet particulate matter, or UVPM, is a marker that is more specific to ETS RSP because it does not measure most non-ETS particles. Consequently, the ratio of UVPM to nicotine should be relatively constant if, as Dr. Hammond suggests, non-ETS particles are the sole reason for the variation in RSP-nicotine ratio.

As the data on Slide 15 show, the ratio between UVPM and nicotine is not constant. We have a number of cities and locations, offices and restaurants -- I'm just explaining the chart. These are the average ratios, and that's the ratios of the measurements determining those individual, from individual measurements within these locations. Some standard deviation, minimum and maximum measurements observed. Anyway.

For example, in Winston-Salem offices, the average UVPM to nicotine ratio was 21.3, but the ratios range from as little as 1.3 to as high as 211. And the average ratio determined in the Winston-Salem offices is different from that determined in Dallas offices. These variations in ratio between ETS RSP and nicotine could not occur unless factors other than just background RSP are responsible for the variation in RSP to nicotine ratios.

So what do these results mean?

First, nicotine cannot be used to quantitatively predict concentrations of either RSP or ETS RSP in the indoor environments. These types of predictions are not accurate because the ratio of nicotine to RSP and ETS RSP is highly variable and dependent upon many environmental factors.

Second, no data demonstrate that the ratio of nicotine to any other ETS constituent is constant across the wide range of the indoor environments encountered in the U.S. workplace.

Now that we've explored some of the problems with using nicotine as a quantitative ETS marker, let's look at the problems in using nicotine and cotinine as quantitative ETS biomarkers.

As we've seen, the relationship between exposure to nicotine and other ETS constituents is poor and highly variable. As a result, the prediction of exposure to individual ETS constituents on the basis of measurements of nicotine and cotinine in biological fluids must also be poor.

Some of the most significant problems was relating nicotine and cotinine concentrations in biological fluids to nicotine exposure include -- and they're here on Slide 16 -- a lack of standardized measurement techniques, inter-individual variation in the metabolism and clearance of nicotine and its metabolites confounding in some individuals by exposure to nicotine from sources other than ETS and the time dependent function describing nicotine metabolism in the body.

To the first point. Since standardized analytical methods don't exist -- and that's for measuring nicotine and cotinine in body fluids -- the results of nicotine or cotinine determinations obtained in one laboratory cannot necessarily be compared to those obtained at another laboratory.

A study performed by Dr. Jeffery Idle showed that there were considerable differences in reported cotinine concentrations from identical urine samples tested at different laboratories. These results are described in more detail in a previous submission to OSHA.

Point two. There are significant inter-individual differences in metabolism and clearance rates and to the relative proportion of metabolites generated. Exposure to a single concentration of nicotine can result in a wide range of nicotine or cotinine concentrations in different individuals.

As an example. If Drs. Coggins, Sears and I had a 10 microgram nicotine uptake right now, our salivary cotinine levels would almost certainly increase by different amounts. The differences can confound attempts to use spot nicotine or cotinine measurements to accurately and precisely back-calculated nicotine exposure.

The third problem, confounding by exposure to environmental nicotine in the absence of ETS or ingestion of nicotine from dietary sources has been discussed in detail in my previous submissions to OSHA.

Now I'd like to spend a couple of minutes exploring the fourth problem with using nicotine and cotinine as quantitative markers for ETS: The fact that concentrations of these compounds in biological fluid are highly time dependent.

At Reynolds Tobacco we have developed a physiologically-based pharmacokinetic, or PBPK model, that can predict general trends in concentrations of nicotine or its metabolites in the body fluids of smokers. We have applied this model to time dependent ETS exposure data.

Before I proceed, I should point out that PBPK models cannot predict nicotine or cotinine concentrations in any individual unless specific, relevant pharmacokinetic data are available and exposure information is know. Or multiple biological fluid samples have been obtained from that individual.

In this case, our model was used to predict plasma nicotine concentrations for an ideal individual. And by that I mean one for whom all the aforementioned information is available. We assumed an eight hour workplace exposure at 10 micrograms per cubic meter nicotine. We also assumed that the individual had no non-workplace nicotine exposure.

What does that profile look like?

Slide 17 shows the predicted plasma concentration of nicotine versus time. The data illustrate three important points that make it difficult to predict nicotine exposure on the basis of spot nicotine measurements.

First. The predicted serum concentrations of nicotine are extremely low.

Second. Nicotine concentrations in serum are extremely transient or variable. They vary throughout the course of the day. In fact a wide range of spot plasma, and therefore urinary and salivary, nicotine concentrations result from a single exposure.

Third. Nicotine is metabolized quickly. Within 16 hours following exposure -- the exposure ending here -- there is almost no nicotine particulate to be found in the plasma. Less than a day after exposure, the nicotine is essentially gone.

So what about using cotinine concentrations instead? There are similar problems.

As you can see here on Slide 18, cotinine concentrations also vary considerably across time. Because of this, at some time points, different nicotine exposures can result in identical spot cotinine concentrations. So that predicting ETS nicotine exposure based on cotinine or nicotine measurements in biological fluids is difficult to impossible, assuming the ideal conditions used in the PBPK model.

Where field measurements are concerned, the use of nicotine and cotinine as biomarkers to quantitatively predict ETS nicotine exposure are further complicated by inter-individual variability in nicotine and cotinine metabolism and inter-individual variability of physiological parameters related to adsorption, distribution and elimination of nicotine and/or its metabolites.

(3) Non-workplace exposure to ETS.

(4) Exposure to non-ETS nicotine.

(5) Much lower concentrations of ETS in the real world than used in the model and corresponding problems with working near the detection limit of analytical methods.

And, finally, poor inter-laboratory precision of nicotine and cotinine determinations.

Because of these factors, at low exposure levels nicotine and cotinine might not even serve as valid qualitative biomarkers for ETS exposure.

Even Dr. Neil Benowitz, one of OSHA's own witnesses at this hearing, has acknowledged many of these difficulties. A study conducted by Reynolds' scientists illustrates that the difficulties associated with predicting nicotine exposure on the basis of cotinine measurements.

Slide 20 shows the salivary cotinine measure in 20 subjects before and after ETS exposure. Throughout the week of the experiment the subjects were instructed to avoid ETS exposure in the evenings and at night. During days 1, 2, 4 and 5, the subjects stayed in an ETS-free environmental chamber throughout the day. On the third day of the experiment, the subjects were exposed to an average of 59 micrograms per cubic meter ETS nicotine for seven and half hours.

And I should point out that this is a typographical error on my written testimony and on the slide. This is seven and a half hour duration.

The ETS RSP concentration in the exposure chamber was 200 micrograms per cubic meter. Salivary cotinine was collected at the beginning and end of each day.

What were the results?

Well, looking at Slide 20, you can see that on average ETS nicotine exposure increased salivary cotinine. However, the important point here is that salivary cotinine concentrations cannot even be used to distinguish non-exposed individuals from those who have heavy ETS nicotine exposure.

As you can see, the salivary cotinine concentration of some individuals who had claimed they were not exposed to ETS were higher than the concentrations of some individuals who were exposed, or following the seven and a half exposure to ETS nicotine.

In addition, immediately following exposure small increases in salivary concentrations were detected in some individuals and substantial increases were noted for others.

Fifty-nine micrograms per cubic meter nicotine is much, much higher than anyone would normally expect to find in the vast majority of workplaces. As numerous scientists have recently testified, data obtained from field studies suggests that in the past average nicotine concentrations in smoking workplaces was approximately 10 micrograms per cubic meter and is considerably less in more modern workplaces.

Small increases in salivary cotinine following the experimental exposure would be considerably smaller upon exposure to ten, two or one microgram per cubic meter nicotine. It is likely that inter-individual variations in nicotine and cotinine metabolism and excretion would far outweigh the small, incremental increase in cotinine concentration following exposure to typical levels of ETS nicotine.

In other words, the variation between people is larger than the variation caused by normal exposures.

In any event, it is improper to suggest that nicotine, and hence ETS exposure, can be quantitatively determined on the basis of single, and perhaps even multiple, nicotine or cotinine measurements. If anyone who is interested in determining ETS exposures then they must first define specifically what exposure they are referring to and then measure that exposure directly. Or use a properly validated marker.

Now, if you say what do I mean by a properly validated marker, one that satisfies the NRC criteria. Especially the criteria that the marker be a consistent ratio to the component of interest, and also one that yields similar results regardless of the laboratory performing the analysis.

In conclusion, the information I have just presented only highlights some of the problems associated with the use of nicotine and cotinine as quantitative markers or biomarkers for ETS exposure. And that brings me back around to my first slide and the real question here.

Are nicotine and cotinine quantitative ETS markers or biomarkers?

I think when all the information is considered the evidence strongly suggests that no quantitative relationship exists between ETS and nicotine and other ETS constituents. Spot measurements of nicotine and cotinine in body fluids cannot be used to predict ETS nicotine exposure. And nicotine and cotinine cannot be used to accurately quantify or differentiate among typical levels of exposure to other ETS constituents.

As a result, nicotine and cotinine cannot, and should not, be used to evaluate or predict risk associated with ETS exposure.

Thank you.

JUDGE VITTONE: Thank you, Dr. Nelson.

Dr. Nelson's printed remarks will be identified in the record as Exhibit 230, and the copy of his slides which he used in his presentation will be identified as Exhibit 231.

(The documents referred to were marked for identification as Exhibit Nos. 230 & 231 and were received in evidence.)

JUDGE VITTONE: Gentlemen, it's 12:00, and I propose that we break for lunch now.

Let me ask Dr. Coggins. Dr. Sears is next?

DR. COGGINS: Dr. Sears is next.

JUDGE VITTONE: Followed by whom?

DR. COGGINS: Dr. Ogden will come back, and then myself, then Mr. Bohanon. Four more speakers.

JUDGE VITTONE: Okay. One of those is a relatively short presentation, right?

DR. COGGINS: Mine's about 45 minutes. Ogden's is about 30 and Mr. Bohanon's is about an hour.

JUDGE VITTONE: About an hour. Okay.

MS. SHERMAN: Would it make sense to get one more before lunch?

JUDGE VITTONE: I was going to say, do you want to go one more time? Are you all ready?

Okay. Let's take Dr. Sears, and then we'll break for lunch.


JUDGE VITTONE: Dr. Sears, you may begin.

DR. SEARS: I'm Stephen Sears. I am a scientist in the research and development department of R.J. Reynolds Tobacco Company. By training I'm a theoretical chemical physicist, and I have a Ph.D. in theoretical chemistry from the University of North Carolina at Chapel Hill.

For the past four years my work at Reynolds Tobacco has focused on environmental tobacco smoke and issues surrounding ETS risk assessment. Today I will talk about just some of the many problems with the epidemiologic approach the proposed rule uses to quantify risk. Specifically, I will explain why the epidemiological evidence that OSHA relied on does not support the contention that exposure to environmental tobacco smoke increases the risk of lung cancer or heart disease.

This discussion is detailed in the written critique that was prepared by Mr. Thomas Steichen and myself for submission to the OSHA docket. Mr. Steichen is a statistician and analyst in the R&D department of Reynolds Tobacco. He holds a Master's degree in statistics from the University of Kentucky and he is co-author of today's presentation. Mr. Steichen has been involved in ETS risk assessment issues for more than three years.

To reach its conclusions concerning ETS the proposed rule relies on the epidemiological evidence that has three key shortcomings, shown here in Slide 1.

First, the evidence is inappropriate for the workplace.

The second shortcoming. The evidence is methodologically flawed.

The third shortcoming. The evidence is insufficient to conclude increased risk.

There are also errors in the way this evidence was interpreted and applied. Therefore the conclusions drawn from this evidence are invalid.

During the past several years Mr. Steichen and I have extensively investigated the epidemiologic evidence concerning ETS exposure, lung cancer and heart disease. We have also analyzed how that evidence is used in the proposed rule. Based on that research there are a couple of points I'd like to make before I detail some of the problems in the proposed rule.

I show these points on Slide 2.

First, the epidemiologic evidence in general does not support the contention that ETS exposure increases the risk of lung cancer or heart disease. Among other flaws, the proposed rule relies on world-wide spousal exposure data to characterize U.S. workers. In fact, as you will soon hear, there are a number of good reasons for not using it. That is, there's a very strong argument for confining the evidence to data about U.S. citizens, and ultimately to data about U.S. workers.

Second. Even if the proposed rule had limited itself to U.S. spousal data, it still could not have reached the conclusion that ETS exposure increases the risk of lung cancer or heart disease.

For example, an EPA-style meta-analysis that includes the Brownson, Stockwell and 1994 Fontham data does not show an increased risk of lung cancer at the 95% level. Concerning ETS and heart disease, as everyone on the panel knows, the body of evidence is so inconclusive that the EPA abandoned its effort to deal with heart disease in its risk assessment.

And that evidence did not even include the results of three large studies that were recently introduced at these hearings. Those studies -- the American Cancer Society's two cancer prevention surveys and the National Mortality Followback Survey -- have greatly increased the amount of data and the strength of the epidemiologic evidence concerning ETS and heart disease. They showed no statistically significant increased risk.

As an aside, the data from these three studies was available in 1972, 1988 and 1986, respectively. The fact that the results were never reported by the researchers highlights the substantial publication bias that exists with ETS.

Now. Let's look at why the epidemiologic evidence in the proposed rule in inappropriate for quantifying risk in the U.S. workplace.

The proposed rule uses world-wide spousal studies instead of U.S. workplace studies to evaluate risk to workers in the workplace. To justify this questionable approach, the proposed rule relied on three false assumptions. These were shown on Slide 3.

First, that spousal residential data is relevant to workplace risk.

Second, that workplace and residential ETS exposures are comparable.

And third, that the expected risk for workplace and residential exposures are equivalent.

The errors in these assumptions lead OSHA to the unjustified belief that world-wide spousal data are an adequate substitute for U.S. workplace date. This unjustified belief undermines the validity of the approach used in the proposed rule.

Let's look at each of the three assumptions in order. Slide 4.

First, are spousal data relevant to workplace risk? The studies cited in the proposed rule used smoking status of the spouse as a surrogate for ETS exposure. This conveniently let researchers label cases and controls as exposed and unexposed.

But these categories, in fact -- Slide 5 -- represent the cumulative and interactive effects of all the differences between the groups. That is, the full array of risk factors associated with living with a smoker. These factors include diet, exercise, socio-economic status, alcohol consumption and a host of others. So rather than being a surrogate for ETS exposure, living with a smoker is in fact a surrogate for the full range of independent risk factors that living with a smoker entails. As a result, risk ratios can only be computed for this aggregate of risk factors rather than for the ETS exposure factor alone.

The attendant statistical tests of significance do not and cannot indicate the specific risk factors that caused the two groups to appear significantly different. That is, the statistical test merely provides evidence that the two groups are or are not different. It does not provide evidence about why the groups are different.

In a November, 1994 publication, The American Journal of Epidemiology, Professor Sander Greenland noted that confidence intervals represent the minimum level of uncertainty associated with a conclusion. That is, the actual uncertainty is in fact higher than the interval predicts.

He goes on to note that the use of confidence intervals as estimates of causal effect must assume four things, as shown on Slide 6.

(1) That all other risk factors have be randomized;

(2) That there are no uncontrolled measurement errors;

(3) That there is no uncontrolled selection bias; and

(4) That the correct statistical model has been use.

Greenland indicates that each of these assumptions elicit concerns, and he specifically notes that the first assumption, that all other risk factors have been randomized is "by definition" never fulfilled in observational studies. Instead, he says, and I quote, "It gets replaced by an assumption of no uncontrolled confounding which, when operationalized, amounts to nothing more than a vain hope that nature randomized exposure within the strata of controlled variables."

We agree that this is a vain hope concerning ETS. Why? Because there is now abundant and still-growing evidence in the literature that spousal smoking status is strongly confounded with a host of lifestyle, socio-economic, dietary and environmental variables.

Because there is now abundant, and still growing, evidence in the literature that spousal smoking status is strongly confounded with a host of lifestyle, socioeconomic, dietary and environmental variables. In 1988, Koo et al. observed that their results, quote, "indicate some correlates of passive smoking that can act as important confounders when evaluating health risk among families with smoking husbands."

In 1993, Dr. Marcia Angell, Executive Editor of the New England Journal of Medicine stated, quote, "In study after study socioeconomic status emerges as one of the most important influences on mortality and morbidity."

In 1994, Cress et al. reported that, quote, "Women exposed to passive smoke differed from those not exposed on several factors that should be considered in future studies that seek to investigate smoking-related disease risk."

And then a submission to the OSHA document that analyzed NHANES I data, Matanoski noted, quote, "Those who were exposed to spousal smoking were more likely to be older, have lower education, live in the city and have other health behaviors which could increase their risk of lung cancer, compared to those with non-smoking husbands."

She went on to list a number of dietary confounders and she concluded that, "Future studies of non-smokers should examine the influence of both passive smoking and diet on the risk of disease, not just a single factor."

Many of the factors that are associated with having a spouse who smokes have been shown to be independent risk factors for lung cancer or heart disease. And, in general, they appear to be more specific to spousal relationships than work relationships.

For example, subject and spouse are more likely to share a common diet than subject and co-worker. As a result, spousal based risk calculations typically overestimate relative risk for the risk factor of concern, ETS exposure.

Confounding of this type is likely to be a less serious problem in the workplace than in the home. And studies that report both types of data typically indicate that workplace risk estimates are in fact smaller than spousal risk estimates. This could be verified both by reports of the individual studies and by pooled estimates of summary reviews.

Given the disproportionate role that confounders play in the home versus the workplace, it is clear that it is inappropriate to use data from home environments to estimate risk that might be present in workplace environments.

The second question we need to address is this, slide 7. Are ETS exposures at the workplace comparable to those in the home?

The proposed rule claims that residential ETS exposures are expected to, quote, "accurately reflect" workplace exposures but this claim is dubious for at least two reasons.

First, as detailed in Dr. Nelson's presentation, airborne nicotine which is the foundation for most of the exposure data in the proposed rule, is, at best, a problematic surrogate for quantifying ETS exposure.

Second, the proposed rule cites a number of studies to show that workplace exposure is as great or greater than household spousal exposure. But as Dr. Ogden's presentation has clearly shown, rather than providing evidence of comparable exposure, some instead provide strong evidence that workplace exposure is substantially lower than residential exposure.

As Dr. Ogden noted, the docket submission by Jenkins and his colleagues at Oak Ridge National Laboratory provides data that are extremely relevant to this issue. Their results show that actual ETS exposure for subjects, quote, "exposed at work only" is less than one-fourth of the ETS exposure for subjects "exposed away from work only." Slide 8.

The Oak Ridge study measured ETS specific markers using personal air sampling pumps. This study alone provides definitive exposure data that directly contradict the contention that workplace and home exposures are comparable.

In light of all the evidence, there is no justification for assuming that workplace ETS exposure is comparable to exposure in the home. We are measuring grapefruit and proposing rules for grapes. The body of evidence clearly shows that people are exposed to much less ETS at work than at home.

Now, let's explore the third assumption by asking, slide 9, are the expected risks for workplace and spousal exposures equivalent?

In a word, no. Slide 10. Since spousal smoking epidemiology is strongly confounded by many more known variables than workplace epidemiology and since actual ETS exposures are not comparable in the two environments, there is simply no logical reason to expect that the risks would be the same. And, indeed, they are not.

An EPA-styled meta-analysis of the workplace lung cancer data demonstrates no statistically significant association for the data in its entirety or for any subgrouping of U.S. workers. The pooled risk estimate of 1.0 is significantly lower and substantially different from the 1.2 to 1.5 range of relative risks the proposed rule has reported for data derived from spousal studies. Even the upper limit of the confidence interval for the pooled workplace risk estimate is only 1.11.

It is difficult to draw any conclusions concerning heart disease risk in the U.S. workplace because virtually no information on the U.S. workplace exists. As detailed in our written submission, there are only two relevant studies that have been conducted. One is flawed to the extent that it cannot be used to derive reliable risk estimates. The other is not significant.

To sum up, the answer to this third question, the hypothesis that residential and workplace risks are equivalent is just that, a hypothesis that is not supported by any U.S. workplace data. Available data, in fact, contradict the assumption of equivalent risks.

Furthermore, it is illogical to assume that risks would be equivalent when the exposures are not comparable and the confounders are so different.

Given these facts, it's clear that the proposed rule's assumptions are wrong. Slide 11 shows the corrections.

Spousal residential data are not relevant to workplace risk; secondly, workplace and residential ETS exposures are not comparable; and, third, actual risk estimates for workplace and residential exposures are most definitely not equivalent.

Therefore, there is no justification for the belief that worldwide spousal data are an adequate substitute for U.S. workplace data. The truth is they are not. So it's clear that if we want to learn about ETS in the U.S. workplace, we need to look at U.S. workplace studies.

What can we learn from the U.S. workplace studies?

To date, eight lung cancer studies and two heart disease studies have used data from the U.S. workplace.

Slide 12. Six of the lung cancer studies do not even meet the most basic statistical measure of credibility for claiming increased risk, significance at the 95 percent confidence level. That means these six studies do not provide sufficient information to conclude that their estimates are not the result of random chance.

Two report a decrease in risk. Two report an increase. And two provide no numerical estimates but their authors conclude that the results show neither an increase nor a decrease.

The remaining two studies provide limited evidence of increased lung cancer risk due to ETS exposure in the workplace. But one of them is only marginally significant for its small subset of males and it is negative for the larger subset of females.

I will discuss the remaining study, Fontham et al. in a few moments.

In the case of heart disease, slide 13, there are only two U.S. workplace studies. One study of the ASHMOG cohort reported in a dissertation by Butler is of marginal significance but contains serious design and execution flaws. The other study, of the MRFIT cohort reported by Svendsen, is not significant for even these high CVD risk subjects.

Despite abundant evidence to the contrary, the proposed rule assumes an increased risk of both lung cancer and heart disease associated with ETS exposure in the workplace. And it goes on to quantify a lifetime occupational risk based on two studies that do not reliably demonstrate any increased risk.

The proposed rule uses data from Fontham et al. to quantify lung cancer risk and it uses data from Helsing et al. to quantify heart disease risk. As I will detail, both studies are unsuitable as a basis for the quantification of workplace risk.

Because of these problems, the lifetime occupational risk estimates for lung cancer and heart disease that are reported in the proposed rule are invalid.

Since the proposed rule has relied exclusively on the Fontham and Helsing studies to provide quantitative data for its risk calculations, I now turn my attention to those studies.

I would like to note up front that this presentation highlights just a few of the more glaring problems with each study. Specifically, I intend to highlight problems that have not yet received a great deal of attention in this hearing. A much more detailed and complete analysis is contained in our written submission.

Let me begin with two important points on slide 14. First, my comments on Fontham deal primarily with Fontham et al. 1994, which has superseded Fontham et al. 1991. Second, Helsing is not a workplace study. I will discuss Helsing only because the proposed rule quite inappropriately relies on it to quantify lifetime occupational risk of heart disease.

Now let's look at Fontham et al.

On the basis of statistical significance alone, Fontham et al. 1994 which reported a 1.39 relative risk diverges from the body of the evidence concerning lung cancer risk in the workplace. The other workplace studies except for the small male data subset of Kabat and Wynder find no effect in point estimate or trend.

In addition, a recent analysis of the largest of these other studies, an analysis of the workplace data from the Brownson study that was submitted to the OSHA docket by Dr. William Butler yields a non-significant relative risk estimate of .09.

This study had a similar number of subjects as Fontham and reported results consistent with every study but Fontham. Our submission presents details on nine serious shortcomings in Fontham et al., including the critical misclassification issue that will be discussed in detail in a subsequent presentation by Dr. Ogden.

For now, I will focus on just two important flaws shown on slide 15.

First, the study population is not representative of the U.S. workplace. The Fontham subjects are exclusively female and are not even representative of the non-smoking U.S. female population. Almost all of the subjects are urban with four-fifths from the major metropolitan areas of Los Angeles and San Francisco. Rural subjects are under-represented by at least a factor of one-third and the percentage of minority cases in the study is twice as high as that in the U.S. population. Asians are over-represented by 600 percent.

Because these demographics diverge so greatly from those of the overall U.S. population, Fontham's risk estimate simply does not apply to the U.S. population, much less to the U.S. workforce population.

A second flaw in Fontham et al. is that its use of frequency-matched controls rather than controls that were individually matched inflated the risk estimates. In general, it is essential to match cases with controls as carefully as possible in case-control studies, especially when any risk elevation is likely to be small. Simply put, a case-control study can be no better than its matching.

Fontham used age and race frequency-matched population controls as opposed to the more precise individually-matched controls. That is, the age and race of controls were generally rather than specifically matched.

Age matching was done using four ten-year categories with no assurance that the age distribution within each age category was similar between cases and controls. In fact, because of excess accrual of younger controls and a deficit of older controls, Fontham was forced to change study protocol during its third year. Specifically, Fontham stopped adding people to the youngest age group and specifically sought an additional source for older controls.

Because cancer incidence is tied so directly to age differences, this frequency-matching bias most probably resulted in artificially inflated risk estimates.

Furthermore, the use of a new source for controls also introduced a new source for bias, confounding and variability. And it eliminated the benefits of accruing controls through random digit dialing.

Using frequency-matched controls also introduced biases resulting from racial differences. By broadly categorizing individuals, Fontham et al. failed to account for important lifestyle differences within categories, especially within the large Asian subset of this study.

This problem raises substantial concerns because it introduced uncontrolled confounders into the study.

Specifically, after the start of the study, Fontham changed the protocol to allow non-English speaking Chinese and Spanish subjects into the case population without requiring that they be matched to non-English speaking controls.

The protocol change increased the probability of acquiring non-smoker lung cancer cases but the fact that these new cases were only matched by race and not by primary language and culture greatly increased the possibility of bias.

For example, there are well documented risk factor differences between native born and U.S. born Chinese women. Specifically, the risk for adenocarcinoma of the lung among non-smoking native born Chinese women is significantly elevated.

This elevated risk has been attributed to practices associated with a traditional Chinese lifestyle, including cooking with smoky coal. It is likely that these practices have been far more prevalent among non-English speaking Chinese than among those Chinese-Americans whose assimilation into U.S. culture includes adoption of the English language.

Matching non-English speaking Chinese cases to English speaking Chinese controls cannot properly account for the biases introduced by these critical lifestyle differences. Again, this bias inflates the risk estimates.

Fontham's protocol changes resulted in the following:

First, the accrual of a disproportionate number of non-English speaking Chinese subjects.

Second, the introduction of controls from an additional source.

Third, a substantial expansion of the San Francisco study area.

And, finally, the termination of accrual from more demographically balanced study centers.

These are shown in slide 16.

Since these changes occurred after the study started but during the accrual of cases, they may have inflated initial relative risk estimates. And they may explain the increases in relative risks observed between Fontham's interim and final reports.

It is clear that Fontham's protocol changes progressively and serious compromise the integrity of case control matching during the five-year period of the study and that it inflated all of the study's risk estimates.

And, as I noted earlier, there are seven additional critical flaws in the Fontham study that we have discussed in detail in our written submission. The combination of such serious design and execution flaws along with the conspicuous deviation of Fontham's result from the body of workplace epidemiology make Fontham et al. unreliable for estimating workplace risk.

It is therefore inappropriate for the proposed rule to use the Fontham et al. risk estimate to quantify workplace risk.

Now, I'll turn to the Helsing study. As I said earlier, Helsing et al. is used in the proposed rule as the basis for quantifying heart disease in the workplace but it is not a workplace study. In addition to its irrelevance to workplace risk estimation, Helsing et al. also suffers from other serious shortcomings as shown on slide 17.

Specifically, the study inadequately adjusts for confounding variables.

Second, it produces perplexing changes during the adjustment process.

Third, it is inconsistent with a later publication on the same data.

Fourth, it has substantial design flaws.

And, finally, it lacks meaningful dose response trends.

As with Fontham, I will highlight only a few of these flaws today. The remainder are detailed in our docket submission.

I'll refer now to slide 18.

First, Helsing adjusts for only four potential confounding variables: age, marital status, housing quality and years of schooling. It fails to adjust for some of the most important risk factors for heart disease, including hypertension, total cholesterol and the ratio between HDL and LDL levels.

Second, Helsing's adjustment process is unclear and it results in unexpected consequences. In fact, among women, adjustment changed the direction of the risk estimate which inexplicably went from a statistically significantly negative relative risk of .68 to a statistically significant positive value of 1.24. Among men, the adjustment changed the relative risk from a statistically non-significant 1.17 to a statistically significant 1.31.

It is surprising to observe quantitative differences of that magnitude and it is even more surprising to observe qualitative differences. That is, changes in the direction of statistical significance for adjustments involving so few and such relatively unimportant confounding variables.

These changes lend themselves to only two possible explanations. Either the observed population was entirely inadequate to represent the target population or the adjustment process used by Helsing and his colleagues are seriously flawed.

Third, in 1989, Sandler et al. reported a relative risk of 1.19 for the same population of females compared to the 1.24 reported by Helsing. It seems clear that OSHA mst resolve this discrepancy before any quantitative use of these risk estimates can be considered.

A fourth problem with Helsing et al. is that smoking status of subjects was obtained at the start of the study and it was never corrected for any changes. No attempt was made to adjust for this misclassification and we have no way to know the consequences.

Finally, no true follow-up was performed on any of the subjects so Helsing et al. could not say how many subjects were actually in their study. Instead, they generated statistical estimates of the number of subjects based on a small mid study sample.

In addition, only deaths in Washington County were recorded. Subjects who left the county and subsequently died of heart disease elsewhere were not counted.

Despite these non-standard epidemiological procedures, Helsing inappropriately applied standard statistical analysis methods. Consequently, the resulting risk estimates and corresponding confidence intervals are meaningless.

Butler discusses this critical flaw in detail in his submission to the docket.

Clearly these and other serious flaws in Helsing et al. render it useless as a source of risk estimates for the proposed rule.

Because of the problems I have discussed today and the additional problems that are detailed in our written submission, none of the lifetime occupational risk estimates in the proposed rule are valid.

The available workplace exposure data do not substantiate any increased risk of either lung cancer or heart disease due to ETS exposure in the workplace. Clearly, without any increased relative risk the formulas in the proposed rule yield zero for the annual and lifetime occupational risks.

In summary, there are problems with many of the steps the proposed rule uses to reach its conclusions. These are shown on slides 19 and 20.

First, evaluation of available data does not reveal an increased risk for heart disease or lung cancer due to ETS exposure.

Second, suitable data is necessary to perform a quantitative risk assessment but data of sufficient quality are simply not available for either lung cancer or heart disease.

Third, it is wrong to assume that spousal exposure data and reported relative risks are relevant to the workplace and the workforce.

Fourth, on slide 20, key values derived from Fontham et al. and Helsing et al. to compute annual and lifetime occupational risks are inappropriate.

And, finally, the resulting lifetime occupational risk estimates contained in the propose rule are invalid.

This topline examination of problems with the proposed rule makes it clear that the epidemiological evidence OSHA relied on does not support the contention that exposure to environmental tobacco smoke increases the risk of lung cancer or heart disease among American workers.

Thank you.

JUDGE VITTONE: Thank you, Dr. Sears.

Dr. Sears' printed statement will be identified in the record as Exhibit 232.

(The document referred to was marked for identification as Exhibit 232 and was received in evidence.)

JUDGE VITTONE: The slides that he used during his presentation will be identified as Exhibit 233.

(The document referred to was marked for identification as Exhibit 233 and was received in evidence.)

JUDGE VITTONE: We will return and resume this afternoon with the remaining RJR witnesses at 1:45.

Thank you very much.

1:50 p.m.

JUDGE VITTONE: Before we resume this afternoon, Ms. Janes, when we were breaking, as I understand there's been a change in the schedule, that you want to add the day of January 24th to the hearing and bring in Dr. Greenfield from ICF Kaiser and Mr. Tiffany from AIHA. Is that right?

MS. JANES: That's correct.

JUDGE VITTONE: So that the witnesses scheduled for Monday, January 23rd will be Dr. Ford and Dr. Bayard, is that right? Does everybody understand that?

Mr. Furr?

MR. FURR: I was not present last Friday, but it was my understanding that a motion was made by Reynolds to reschedule Dr. Bayard on a day in which Dr. Ford was not also appearing.

JUDGE VITTONE: That's correct.

MR. FURR: I understood that you took it under advisement at the time.

JUDGE VITTONE: I took it under advisement. I guess with respect to the motion made by Ms. Ward last week, do you want to respond to that?

I just stated for the record so that people would know, that January 24th would be an additional day of hearings with Mr. Tiffany and Dr. Greenberg, and that the two witnesses, as I understand that you have scheduled for the 23rd are Dr. Ford and Dr. Bayard, is that correct?

MS. SHERMAN: That is correct, Your Honor. Dr. Greenfield was the last on on the 23rd. He had not been scheduled to be on on the 23rd. His schedule allowed him to be rescheduled. Dr. Bayard's did not. While we appreciate Reynolds' concern, we're willing to start the hearing early that day. However, I think that it's within OSHA's discretion as to who to move on a certain day.

JUDGE VITTONE: I would be willing to start early on that day, the 23rd, and also stay late that day, as necessary.

MR. FURR: Unless we're talking about staying until the middle of the night, I don't think that even comes close to addressing the concerns that we have about attempting to examine both of those witnesses the same day.

I actually have a number of points I'd like to make as to why we believe that's true, and if this is the time I'll be happy to go now, Your Honor.

JUDGE VITTONE: Why don't we hold off on that. Let's bring it up, I think it's more important that we get through today with these witnesses. I was just trying to give everybody a little advance information, but let's hold off on that, okay? We'll take it up this week. But I think it might be better if we got through these witnesses for this afternoon and made sure we made substantial progress today.

MR. FURR: Thank you.


DR. COGGINS: Thank you, Judge. The next speaker is Dr. Ogden, who will discuss the problems associated with misclassification.


DR. OGDEN: Thank you, Dr. Coggins, Your Honor.

I am Dr. Michael Ogden. My academic qualifications and affiliation were outlined in my previous testimony on environmental tobacco smoke exposure.

My testimony that follows will deal with one major bias in the spousal smoking epidemiology of ETS exposure. This bias that I will address is that due to the misclassification of current smoking status among self-reported, never smokers.

In Slide 1 I pose five questions. They are:

One, what is smoking status misclassification? Two, how do we detect smoking status misclassification?

Three, why is smoking status misclassification important to OSHA?

Four, how many smokers actually deny that they smoke?

Five, what does proper misclassification correction do to lung cancer risk estimates observed in the spousal smoking epidemiology?

In Slide 2 I turn to the first question I posed. What is smoking status misclassification? Quite simply, smoking status misclassification occurs when a person misrepresents his or her smoking history. As you might imagine, there are several types of smoking status misclassification.

As Slide 3 shows, for example, current smokers could report themselves as former smokers, implying that they have quit smoking. Also, current smokers could report that they are never smokers. That is, that they have never smoked. In addition, former smokers who have quite smoking could deny that they ever smoked, and report themselves to be never smokers.

As you can see, there are other types of misclassification, but the ones I just described are the most important as biases in risk assessment. Of these, the largest misclassification bias is due to current regular smokers reporting themselves to be never smokers. This is the so-called regular smoker misclassification rate and is the only type of smoking status misclassification that I'll discuss here.

Misrepresentation of smoking status is potentially a major bias in the epidemiology based risk assessment of ETS exposure. Proper adjustment to individual epidemiology studies or to groups of meta-analyzed studies requires that misclassification rates be known precisely for the target population under study.

Slide four states that almost exclusively, epidemiologic investigations that attempt to study ETS risk, have used spousal smoking as a surrogate for ETS exposure of self-reported, never smoking women. As a result, a major target population for ETS related epidemiology is married females, and the most important type of misclassification is the degree to which regular smoking married females report themselves to be never smokers.

The resulting misclassification is a recognized bias is ETS risk assessment based on spousal smoking epidemiology. What is not generally recognized, however, is just how prevalent this misclassification really is. I'll come back to this point later.

Beginning with Slide 5, I'd like to answer the second question I posed. That is, how do we go about measuring smoking status misclassification?

True current smokers who misrepresent themselves as non-smokers, can be detected by measurement of cotinine, a major nicotine metabolite in biological fluids. However, the ability of cotinine to detect current smokers disappears in two to three days after smoking, assuming a cotinine half-life of approximately 17 hours.

As discussed earlier in the testimony of Dr. Neil Benowitz, there are large variations in cotinine half-life among different individuals. Accordingly, some smokers' cotinine levels would decrease to non-smokers' levels in one day after smoking. Other smokers' cotinine levels might not decrease to non-smokers' levels until five or six days after smoking.

This brings me to Slide 6, and a very important point. Even though there is considerable variation among different people, using body fluid levels of cotinine as a guide, virtually all smokers will appear to be non-smokers within a few days after last smoking.

The reason this is an important point is because of some recent criticism against certain types of studies that have determined misclassification rates. Some have argued that misclassification rates determined from population surveys over-estimate the rates that would be found among cases in epidemiologic studies. However, based on the previous discussion, it would appear that just the opposite is true.

In epidemiology studies, significant numbers of the subjects may be infirm, institutionalized, or otherwise limited in access or motivation to smoke. Accordingly, the ability to detect regular smoker misclassification by cotinine determination is diminished among cases in epidemiologic studies.

Before proceeding with a more detailed review of smoker misclassification, let me pause for a moment and try to answer a fairly simple question, and that is, in proposing a workplace ETS rule, why should OSHA be concerned with smoking status misclassification among married couples?

At first thought, one might be tempted to answer that OSHA shouldn't be interested. However, that's not the complete answer. The best answer is because the NPR places so much emphasis on ETS exposure in the home.

As stated in the NPR and reproduced in Slide 7, "Risk estimates based on residential exposures are expected to accurately reflect occupational risks in most workplaces and possibly underestimate the risk in some workplaces."

These risk estimates based on residential exposure are, of course, derived from the spousal smoking epidemiology. Being derived from exposure to ETS among spouse pairs in the home, this is the same epidemiology that is biased by smoking status misclassification among spouse pairs. This is also the same epidemiology that is complicated by a number of confounders that group differentially between smoking and non-smoking households.

Dr. Sears has addressed this latter topic in more detail in his testimony.

The fourth question, how many smokers actually deny that they smoke, is actually the crux of the whole misclassification issue. It is also a question that simultaneously generates a little controversy and a lot of confusion.

Before I begin discussing misclassification mathematically in terms of rates and percents, it is absolutely imperative to understand the two things shown on Slide 8.

Number one, that most misclassification adjustment models require that smokers be categorized into so-called regular and so-called occasional smokers. And two, that different misclassification adjustment models require different ways of expressing misclassification rates.

First let me explain the difference between an occasional and a regular smoker. The differentiation between the two is typically based solely on cotinine concentration. In general, people with cotinine concentrations above some high level are classified as regular smokers. Those with cotinine at some intermediate level are classified as occasional smokers. People with cotinine below this intermediate level are classified as non-smokers.

As shown in Slide 9, please realize that within certain broad ranges the choice of cotinine cut points is entirely arbitrary. Thus, one important consideration in comparing misclassification rates derived from cotinine measurements is to know what cotinine cut points were used to differentiate regular smokers, occasional smokers, and non-smokers.

A second important consideration in comparing misclassification rates is to ensure the rates are expressed numerically in the same way. I know it's cliche`, but you must make sure you're comparing apples with apples.

Slide 10 illustrates the three ways that are commonly used to express the most important misclassification rate. For simplicity, this most important rate is the only rate I'll discuss here, because it accounts for the largest bias. In essence, this involves true, regular smokers who have been misclassified as never smokers.

One way of expressing this misclassification is as a percentage of reported never smokers who are determined to be regular smokers. A second way of expressing this misclassification is as a percentage of true regular smokers who report themselves to be never smokers. Both of these calculations are intuitive. Both results are probabilities, since the numerator is a subset of the denominator.

The third way of expressing this misclassification rate is not so intuitive. However, it is the method I will use here because it is the format required for input into the recent Environmental Protection Agency risk assessment on ETS and respiratory health, including lung cancer.

This calculation requires that the number of reported never smokers who are regular smokers, be divided by the number of reported current smokers that are regular smokers.

In 1992 we conducted a nationally representative study of smoker misclassification among the U.S. married female population. The reason we studied only married females is because this is the subject group in almost all of the ETS epidemiology.

This is the largest, most representative study of smoker misclassification ever performed in the United States. Also, this study was conducted in such a way to enable calculation of nationally representative misclassification rates by any one of the three methods I just described.

Most importantly for the present discussion, this study allows us to calculate nationally representative misclassification rates exactly in accordance with the protocol used by EPA.

These new data on misclassification, along with a critique of EPA's misclassification adjustment, were presented to OSHA in May 1993 and submitted to the docket. A slightly more detailed report was included in my written submission. A final technical report will be entered into the docket during the post-hearing comment period.

Since this study has been described in some detail previously, both in our oral and written submissions to OSHA, I'll discuss only the major findings here.

Regular smoker misclassification rates from this study are tabulated here in Slide 11. As I mentioned earlier, and as you can see in Slide 11, these rates can be expressed in many different ways. I've shown only two of the three common calculation methods: the percentage of true regular smokers who are misclassified and the method used by EPA.

I've also shown misclassification rates calculated at two different cotinine concentrations. The higher cut point used to identify regular smokers, that is 106 nanograms per milliliter is 30 percent of the mean cotinine determined in all self-reported current smokers. This is the exact definition used in the EPA misclassification model.

The lower cut point shown, 35 nanograms per milliliter, is 10 percent of the mean cotinine level determined in all self-reported current smokers. This value, as a cut point, is in better agreement with most of the published studies that have tried to detect smokers among self-reported non-smokers.

However, there is yet another variable represented in this tabulation. That is, reporting of observed misclassification rates or rates that have been weighted in accordance with the target population.

The observed misclassification rate using a cotinine concentration of greater than 106 nanograms per milliliter to indicate regular smoking, is 6.02 percent. This is the rate which I reported previously to OSHA. Weighting this value, the 6.02 percent, so that it becomes nationally representative, yields a result of 2.81 percent.

In a moment, I will illustrate the effect of applying some of these misclassification rates in the EPA misclassification adjustment model. However, before I do, let me reiterate one very important point. This study is the largest, the most representative, and the most comprehensive study done to date on the target group studied in the majority of the ETS epidemiology. All parameters necessary to calculate the misclassification rates were measured directly. Also, all parameters necessary to weight the observed rates to the national norm were also measured directly.

The rate of misclassification among regular smokers, between 2.8 percent and 4.1 percent, is substantially larger than the EPA-assumed rate of 1.09 percent.

In previous presentations and submissions to OSHA, I outlined some of the problems with the misclassification rate determined by EPA. Some of these are outlined here on Slide 12.

Specifically, it was explained to OSHA how one, EPA failed to demonstrate that the EPA rate of 1.09 percent is representative of the U.S. population.

Two, EPA failed to demonstrate that this rate was even representative of the studies cited.

Three, new analysis indicates misclassification bias alone is likely to explain any risk elevation observed in the U.S. epidemiology of ETS and lung cancer.

A misclassification rate of 3.05 percent is all that is necessary to nullify EPA's conclusions regarding significant elevation in lung cancer risk from their meta-analysis of U.S. epidemiology studies. Only three percent. That's a very small number.

The NPR states, as quoted here in Slide 14, that "Many potential sources of bias such as publication bias, misclassification bias, and recall bias, cannot account for the elevation in risks seen in these various studies." This is not true.

In a presentation to OSHA in May 1993, I demonstrated how just one of these biases, the misclassification of current smokers as never smokers, could account for all observed risk in the U.S. epidemiology of spousal ETS exposure and lung cancer.

Recall that only 3.05 percent misclassification is required to reduce the current EPA meta-analysis to non-significance. This is well within the range of likely misclassification rates for the target population.

Shown on Slide 15 is the reduction in EPA's meta-analyzed risk estimate for lung cancer with increasing misclassification rate. The central line is the risk estimate, and the upper and lower lines are the upper and lower 90 percent confidence limits respectively, as used by EPA.

As you can see in Slide 15, the lower bound of the confidence interval crosses the horizontal line at about three percent. This is the point at which the risk estimate loses statistical significance, applying the 90 percent confidence level used by EPA.

Also note that the risk estimate itself is reduced to one at about 7.5 percent misclassification.

Also on Slide 15, I've illustrated the two misclassification rates from our study which are applicable to the EPA misclassification model.

Even higher misclassification rates than those reported here have recently been observed. For example, there is a recently published study of Apseloff et al in three clinical settings in the U.S. This study found nearly 16 percent of self-reported non-smokers had urinary cotinine levels consistent with being regular smokers. Thus, and this is very important, this result, recently obtained from clinical settings, is over four times higher than the result we found in our study. In fact, it's even off my graph here.

In his testimony at this hearing, Dr. Judson Wells made several criticisms against our study. He claimed that our numbers, which were submitted to EPA in 1992, were confusing. However, the numbers were presented in exact accordance with the requirements of the EPA misclassification model.

Further, Dr. Wells cited an example in which he claims we reported 80 occasional smokers were misclassified in the sample that probably only contained about 20 occasional smokers to begin with.

The record of our submission to EPA will, of course, speak for itself. However, our written submission to EPA concerning our misclassification study does not claim anything whatsoever about 80 occasional smokers.

Remember earlier in my statement that I outlined several different ways of expressing smoker misclassification. I also pointed out that there is a potential for confusion when interpreting and comparing these rates. Interestingly, Dr. Wells, the author of the EPA misclassification model, apparently misunderstood our submission to EPA.

We were explicit in stating that our reported misclassification rate was calculated in exact accordance with the EPA model. However, somehow, he missed that point. He then inappropriately adjusted our observed misclassification rate upward by over 100 percent. Then he criticized our misclassification rate for being too high.

In short, Dr. Wells, wasn't comparing applies with apples.

Misclassification bias is widely recognized as a potential problem in spousal smoking epidemiology because of the concordance of smoking habits between spouses.

Based on concordance alone, smoking status misclassification would not be expected to bias workplace ETS epidemiology. However, there would be smoking status misclassification bias in workplace epidemiology if there was a differential rate of misclassification between exposed and unexposed groups. We have observed this type of differential misclassification.

For example, in the study of ETS exposure we conducted in New Jersey in 1992, we found five percent of the self-reported non-smoking subjects actually had salivary cotinine levels that indicated they were smokers. I describe this study in more detail in my written submission, and also in my prior testimony on ETS exposure.

All of these misclassified subjects worked and reported exposure to ETS in their workplace. Thus, these same subjects would be in the category of "ETS exposed" in workplace ETS epidemiology. It is not known to what extent this type of bias has affected the reported workplace ETS epidemiology.

In conclusion, which I show here on Slide 16, new data indicate that misclassification rates are much greater than the 1.09 percent assumed by EPA. Proper consideration of misclassification bias is likely to nullify the observed lung cancer risk in spousal smoking epidemiology. Further, this same bias may be significant in the workplace ETS epidemiology as well.

Thank you.

JUDGE VITTONE: Thank you again, Dr. Ogden. Dr. Ogden's printed statement on smoking status misclassification will be identified as Exhibit No. 234. And his exhibits used on the presentation on misclassification will be Exhibit No. 235.

(The documents referred to were marked for identification as Exhibits No. 234 and 235, and were received in evidence.)

JUDGE VITTONE: Dr. Coggins, who's next?

DR. COGGINS: I am, Your Honor.


DR. COGGINS: I am Chris Coggins. I'm a Board Certified Toxicologist, and my specialty is inhalation toxicology.

In my first slide, I review the studies with experimental animals that were included in the Notice of Proposed Rulemaking. Those selected studies were flawed for at least two major reasons. The wrong test material was used; and the concentrations used were completely inappropriate. Rather than using ETS or a surrogate, most of the studies used fresh mainstream smoke, which is known to be physically and chemically very different from ETS.

Moreover, the smoke concentrations used in some of the studies were "phenomenally high", according to Dr. Neil Benowitz in his September 26th comments at this hearing.

I point out in my second slide that studies without major drawbacks were only considered in a cursory way, if at all, by the agency.

My presentation today will review those latter studies that OSHA did not examine in detail. These studies were performed on rats by RJR, and in rats and hamsters by the German consortium of cigarette manufacturers.

The studies show that the toxicological activity of aged and diluted sidestream smoke used as a surrogate for environmental tobacco smoke, ETS, was extremely minimal, even at massive exaggerations of real-world ETS concentrations.

The only change that we saw in our smoke-exposed rats was a very slight thickening of the lining of the inside of the tip of the nose. We saw no changes whatsoever in the lungs or in any other organ except for the tip of the nose.

Similar findings have been reported by the German researchers who found no responses at all in smoke-exposed hamsters.

At the end of my talk I will review some new data from our laboratories, data from a very sophisticated technique named cell proliferation, which again show minimal toxicological activity at very exaggerated concentrations.

My third slide asks, "Why were studies performed?" We performed studies to obtain data on the toxicological responses of experimental animals exposed to different concentrations of ETS for extended exposure durations. The NPR contains a brief review of an earlier experiment which we performed using only 14 days of exposure Our published papers, referring to the longer exposures, 90 days, were evaluated only very superficially in the NPR, although we did present the data verbally to OSHA in September 1993. The German 90 day study was not included at all. Today I will review some of the data from our longer term experiment.

In my fourth slide, I then ask, "What exactly is ETS?" ETS is largely aged and diluted sidestream smoke, that is the smoke that comes from the lit end of the cigarette, mixed with smaller amounts of aged and diluted exhaled mainstream smoke that comes from the smoker. Because of this, aged and diluted sidestream smoke, or ADSS, acts as a good surrogate for ETS. Aging and dilution are critical factors in the formation of ETS.

In my fifth slide I state that ETS is not the same as either mainstream or sidestream smoke. That there are fundamental physical and quantitative chemical differences between ETS, mainstream, and sidestream smoke, largely due to the aging and dilution mentioned in the previous slide. This is one reason why the animal studies discussed in the NPR are inappropriate. The wrong test agent was used.

Slide 6 examines the amounts of ETS that people are exposed to. A recent review by Holcomb of respirable suspended particulate -- RSP -- concentrations in area measurements of 951 smoking and 905 non-smoking homes showed mean concentrations of 49 micrograms per cubic meter in the former; 22 micrograms in the latter; giving an overall difference of 27 micrograms per cubic meter.

You will recall that Dr. Ogden has personal sampling data showing lower values still for the amounts of ETS that people are exposed to. Exposure, of course, consisting of the product of concentration and time. Area measurements with concentrations expressed in millionths of a gram per cubic meter are extremely low.

Holcomb found similarly small differences in area measurements in Slide 7, of smoking and non-smoking offices, 22; restaurants, 42; and transportation, 30 micrograms per cubic meter. This is another reason why the animal studies mentioned in the NPR are inappropriate. The doses used were phenomenal exaggerations of the very low concentrations seen in the field.

In Slide 8, I review our decision to expose animals at an assumed, but very high, real world RSP concentration of 100 micrograms per cubic meter, and then at 10 and 100-fold exaggerations. In other words, .1, 1, and 10 milligrams per cubic meter.

The selected concentrations encompass the single concentration of four milligrams per cubic meter used in the German 90 day inhalation study, a study which used 10 hour daily exposures. That study showed minimal changes in the nasal cavities of smoke-exposed rats. No changes of any kind were noted in smoke-exposed hamsters.

Slide 9 reiterates that 0.1 milligrams per cubic meter is an over-estimate of a real world RSP concentration as we heard earlier from Dr. Ogden. One milligram a cubic meter is a worst case situation that would not normally be encountered even with the poorest ventilation. Finally, ten milligrams per cubit meter is a massive exaggeration of any ETS situation which we perform to establish dose relationships. It was difficult to see the animals in the inhalation chamber at this massive concentration.

For the remainder of the talk I will use the color coded abbreviations shown on Slide 10. Thus, HRW stands for high real world, or .1 milligrams per cubic meter, which will be shown in yellow. Ten X stands for ten times high real world, or one milligram per cubic meter, and this is shown in green, which doesn't appear to be too visible. One Hundred X is 100 times high real world, and this is shown in cyan, and that was the ten milligrams per cubic meter.

As you can see in Slide 11, the high RSP of ten milligrams per cubic meter is approximately 370 times the mean RSP difference shown in Holcomb's smoking and non-smoking homes and workplaces. A huge exaggeration. It is more than 300 times higher than the absolute RSP figure for workplaces reported by the Oak Ridge team.

Slide 12 shows a schematic of the animal exposure room. In the schematic, the symbol H represents a filter called a HEPA-filter. HEPA stands for high efficiency for particulates in air. So the entire room was HEPA-filtered.

Within the HEPA-filtered room we had a sidestream smoke generator, and we had five exposure chambers. The exposure chambers and a common plenum or duct in yellow, were all connected to vacuum represented in this layout as V.

Fresh sidestream smoke was drawn into this common plenum and was diluted and was allowed to age. Different amounts of this aged and diluted sidestream smoke were then drawn through the chambers -- high real world, ten-fold high real world, and 100-fold high real world -- by connecting these chambers to vacuum. A separate chamber was used for control animals treated identically as the experimental animals, but only exposed to HEPA-filtered air. These are what's called "sham" exposures. A fifth chamber contained animals that we used as sentinels for infectious disease. None occurred. This novel layout was recently published.

The animals lived permanently in the chambers. They were exposed to ADSS for six hours a day, five days a week, for 13 weeks.

Slide 13 shows a picture of one of the stainless steel whole body inhalation chambers. Each chamber held 96 rates. Chambers and the associated equipment are all commercially available. The chambers had a number of sampling ports to allow analysis of the chemical and physical composition of the material being presented to the animals. Such analyses are required to demonstrate the surrogacy for ETS. In general, only minimal chemical analyses were performed in the studies described in the NPR.

Slide 14 is a picture of a nose-only restraint tube into which the rats were placed during the six hour exposures so as to minimize dermal and oral (from preening) routes of administration. So the rat would normally occupy this position here. Here would be the nose, and the tail would hang out here.

This concept of minimizing non-inhalation routes of administration was not considered in the studies reviewed in the NPR. Nor was it considered in other work, including the German study and the work from Dr. Glantz's and Dr. Penn's laboratories. The deficiency is important, because large amounts of material can be ingested when whole body exposures are used, with the potential to render studies meaningless.

Slide 15, as mentioned earlier, we made a number of measurements of the smoke composition, and these are listed on Slide 15. For particulates, the parameters were total particulate matter, particle size distribution, and solanesol. Vapor phase analytes were nicotine, carbon monoxide, and the nicotine combustion product,
3-ethenylpyridine. The smoke composition data have been published elsewhere, and only a small part will be presented today.

We measured the variability in the aerosol total particulate matter over the 65 exposure days. The targets, remember, were .1, 1, and 10 milligrams per cubic meter. Shown on the vertical scale is a log of total particulate matte measured in milligrams per cubic meter. This is Slide 16. Note that this is a log scale. It shows that the targets were met, and that there was very minimal day-to-day variation in the quantities of aerosol presented to the animals. Clearly, we had a very good control of the experimental conditions.

Yet another weakness in most of the published studies is that they did not measure the size of the aerosols to assess whether they would be respirable by the animals. Slide 17 reviews the critical analysis of particle size distribution which we estimated using cascade impactors with cutoff diameters in the range of .5 to 4 micrometers. The results were that median diameters were less than one micron in each of the groups. The distributions were fairly homogeneous with sigma g's of around 1.25 as defined by the American Conference of Governmental Industrial Hygienists. These particles were thus completely respirable by the animals. Published studies in general did not attempt to assess respirability of the experimental aerosols.

Slide 18 plots the mean concentrations of carbon monoxide on the vertical scale, expressed in parts per million, in the different chambers over the 90 days of the study. The two white bars indicate CO concentrations in the control chamber and just in the exposure room. Remember the color coding, the yellow bar indicates the high real world; the green bar, which is almost invisible, is the ten-fold real world; and the cyan bar, the 100-fold high real world.

Also shown on this exposure is the ACGIH threshold limit value of 25 parts per million. So not only were animals exposed to extremely high concentrations, some of those concentrations in one group were more than twice the TLV.

Following the chemical and physical characterization of the smoke presented to the animals, we made measurements of the amounts actually inhaled by them. These are shown in Slide 19. Again, such analyses were in general not performed in the studies discussed in the NPR, there being effectively no information there on dosimetry.

Specific biomarkers for ETS are only available for vapor phase components -- that's carbon monoxide, which results in blood concentrations of carboxyhemoglobin, and nicotine, which results in plasma concentration of nicotine and of cotinine.

Slide 20 shows the blood carboxyhemoglobin concentrations in percent on the vertical scale for the four different groups. Again, the white bar indicates the control; the yellow bar indicates the high real world; both of these groups show negligible concentrations. The ten-fold high real world in green, with approximately half a percent of COHb; the cyan bar indicates the hundred times real world with a mean of approximately five percent. This is extremely high, indicating that large amounts of smoke had, in fact, been inhaled.

Also noted on the chart is the ACGIH "Biological Exposure Index", BEI, above which ACGIH says that there is "increased health risk," and that's at 3.5 percent.

Slide 21 depicts the plasma nicotine concentrations in nanograms per mil, again in the different exposure groups, right at the end of the six hours of exposure. Both the control and the high real world had negligible values. Concentrations in the ten-fold real world group were around 20 nanograms per mil, and in the 100-fold high real world group, concentrations were around 60 nanograms a mil. Very high values indicating the inhalation of very large amounts of smoke.

Slide 22 shows the plasma concentrations of cotinine, the main metabolite of nicotine, immediately after six hours of exposure, again in nanograms per mil. Concentrations were, again negligible in the control and high real world groups. The ten-fold high real world, approximately 40 nanograms per mil; and in the 100-fold high real world, approximately 150 nanograms per mil. Again, the data are indicative of the inhalation of large quantities of smoke.

The data from the last three slides show that the most common biomarkers for ETS exposure -- that's carboxyhemoglobin, plasma nicotine and plasma cotinine -- are not sufficiently sensitive to detect six continuous hours of exposure to high real world ETS concentrations. As concluded by Dr. Nelson earlier in the day, the proposed biomarkers do not biomark.

Slide 23 shows the body weight change in grams in the vertical scale, from across the different stages of the experiment, from delivery, through a quarantine period, through the smoke exposure period, and through a reversibility period. I've used the same color coding. The white indicates the sham controls; the green, the high real world; the yellow the ten-fold high real world; and the blue, the 100-fold high real world. At no time were there any differences between any of the groups in terms of their mean body weight. This is despite the absolutely massive amounts of smoke that the animals were exposed to. No effects whatsoever on body weight. That was Slide 23.

In Slide 24, we made examinations to see if the smoke exposures produced any chromosomal damage. In the assay used, free cells were obtained from the lungs of the animals. These are so-called pulmonary alveoli macrophages. These macrophages engulf and digest foreign material that may enter the deep lung. The cells were examined under the microscope to see if there was any increase in the normal background rate of damage to the chromosomes.

As stated on this slide, there were no differences observed between any of the groups at any time point in any of the experiments.

Slide 25 shows the rate, on the vertical scale the aberrant cells in percent, the different groups across the bottom. Adjacent triplets with the same color represent 14, 28, and 90 days of exposure. That is to say, we used interim kills.

Here is the baseline of background chromosomal aberrations. There is no difference between any of the groups at any time. We used animals treated with the positive control cyclophosphamide to show that our test was working. Exposure to massive amounts of ETS, the cyan bars, did not produce any changes in the sensitive genetic toxicology end point.

In Slide 26 I review the use of DNA adducts as a non-specific marker for exposure. An adduct is simply a molecule which has somehow become bound to the DNA molecule. DNA adducts are currently the subject of intense research in toxicology circles -- research that has shown that DNA is constantly being damaged or adducted, and that modified DNA is just as constantly being repaired.

The fact that DNA adduct are routinely found in tissues where cancer is virtually unheard of, in particular heart, indicates that the assay is probably a non-specific marker of exposure, and that it's not a precursor for cancer. We use the 32P post labeling technique to study adducted DNA. this is currently the most sensitive technique.

Slide 27 shows on the vertical scale DNA adducts expressed as adducts per billion nucleotides. We're talking of a very sensitive test. In heart tissue. Again, we have triplets representing 14, 28, and 90 days for control. High real world, ten times high real world, and 100 times high real world.

There were no increases over control responses for yellow or for green. All you have to do is compare this group with this group, this one with this one, and this one with this one.

There were statistically significant increases at 14, 28, and 90 days for the 100-fold-plus groups, that's the asterisks, when compared with controls. Note the increase in this group when exposures were increased from 14 days to 28 days.

Note also that there are increases in the background amounts of adducts in the controls with increasing age.

Slide 28 shows similar information as the last one, but this time the organ is the larynx. Clearly, there were no effects over controls for the high real world, or 100-fold real world, or for the 14 day exposure. There was a statistically significant increase for 28 and 90 days, and the difference between 28 and 90 days also was statistically significant.

The only other organ which showed effects of exposure was lung, and the data are shown in Slide 29. Again, there was no increase over control values in the high real world or ten-fold high real world. The increase over control values in the 100-fold high real world group was statistically significant.

Not shown on my slides are the DNA values from liver and bladder. They did not contain increases in DNA adducts even in the 100-fold high real world group.

Necropsy observations are described in Slide 30. They included terminal body weight, organ weights, both absolute and relative, gross pathology, hematology where we had 14 tests, and clinical chemistry, 16 tests. There were no differences observed between any of the groups at any time point in any of the experiments.

The only histopathology noted, and this is Slide 31, was a mild hyperplasia, an increase in cell numbers, of the epithelium of the nasal turbinates in one section of the nose in the 100-fold, high real world group only. Translated, this means a slight thickening of the lining of the inside of the tip of the nose. No histopathology was noted in any organ other than the nose.

Hyperplasia is an extremely minor change, much less dramatic, for example, than the formation of skin calluses in manual laborers. The pathologists, and we used three, consider this hyperplasia, and this is Slide 32, consider this hyperplasia to be an adaptive and reactive response to repeated irritation, not a sign of toxicity.

Slide 33 is a cross section, schematic of a cross section of the rat nasal cavity. This is right at the tip of the rat nose. Here is the septum running down the middle, here are the teeth. The blue area indicates the airway. This is a nasoturbinate. This is a maxilloturbinate. This is the nasal lacrimal duct and the vulmerol nasal organ which is how boy rats find girl rats.

This is the site of the change, so in this entire cross section, the only change that we saw was in this tiny area here. This is an actual cross-section of the rostral, forward part, of the nasal turbinates.

So here is the septum down the middle, here are the teeth, the nasoturbinates, maxilloturbinates, vulmerol nasal organ. This is the airway. The air flows in and out these two sides of the nose.

The only change that we saw was right on the tip of this part here. That was Slide 34.

This is Slide 35, which is a schematic of that rat nasoturbinate. That's the tip of a nasoturbinate. The underlying tissue is cartilage and bone. The airway is, again, here in blue. There are blood vessels distributed along the tissue, but there is a lining to the tissue. This lining is called an epithelium. The epithelium rests upon a basement membrane. That's the basement membrane. The epithelium lies upon the basement membrane and goes right out to the airway. That was No. 35.

Slide 36 is a higher-power magnification of the tip of a "real" nasoturbinate from a control animal, it looks like Florida, note that the thickness of the outer lining, the epithelium, here's the basement membrane, the thickness of this epithelium from the basement membrane to the airway is about two cell layers thick. Here we're talking a very large magnification. That's 50 millionths of a meter.

Slide 37 is a side-by-side comparison of nasoturbinates from a control animal and an animal in the hundredfold high real-world group after 90 days of exposure. This is from, as I said, a non-exposed control animal, it's the same as the last slide.

Ignore this overall size. It's just that two animals are sectioned in slightly different positions.

The important point to note is that in the hundredfold high real-world animal the thickness of the lining from the basement membrane, which runs like so, through the airway is six layers thick. This is the only histopathological change that was noted in the experiments.

Slide 38 notes that the nasal histopathology was completely reversible in those subgroups of animals kept without further exposure for an additional 90 days. We made peer review of the pathology results. The pathologists concurred in their conclusions and here are the three pathologists that we used.

The horizontal axis of slide 39 shows the grades of the severity of the hyperplasia so there -- the pathologist graded them as being no hyperplasia and minimal change, mild change, moderate change, marked change or severely changed.

On the vertical axis is shown the percentage of animals with the different change. Control animals are in white, high real-world in yellow, tenfold real-world in green and in the blue color here, hundredfold high real-world.

There were no statistically significant differences between the white, yellow and green distributions. Clearly there is a significant effect of the high exposure.

The slide that I showed you side by side was from this group here.

Slide 40 notes that reevaluations were made of the nasal histopathology in the hundredfold high real-world groups only using blind evaluations to assess progression of the change with additional exposure. So this was done independently by two of the pathologists, a mixed set of slides from the hundredfold high real-world animals for 4, 28, 90 and 14 days were put together, mixed. The pathologists had knowledge of the treatment but not of the duration of the exposure.

Slide 41 shows the unmixed results. Again, none, minimal, mild, moderate, marked, severe. Percentage of the animals in the hundredfold high real-world group after different exposure periods, four days in yellow, 14 days in green, 28 days in blue, 90 days in red.

The data clearly show that there is no difference between the distributions for four, 14, 28 and 90. In other words, there is no progression of the change with very significant increases in the exposure periods.

Slide 42 points out that the larynx or voice box has been shown to be a very sensitive organ in rodent inhalation studies. Here is the reference.

The main histological change reported in the larynx in these studies is a metaplasia. Metaplasia I define as the conversion of one differentiated cell type to another. Of the ventral epithelium of the larynx.

Like the hyperplasia discussed earlier, metaplasia is reversible but it is considered toxicologically more important than hyperplasia. Metaplasia is, however, often described as adaptive. A simple example of metaplasia is the callouses mentioned earlier in the manual workers.

We did not find laryngeal metaplasia in this study, even in the hundred high real-world group. And no laryngeal changes were reported in either of the species used in the German inhalation study.

Slide 43 is a schematic of a low-power view of a cross-section of a rat larynx showing the vocal chords and the site opposite them where the metaplasia is seen. The site is directly above subepithelial glands. The blue indicates the airway. The white here is the esophagus and other sets of cartilage around the outside. So the air flows up and down the blue part. Directly opposite the vocal chords is this site.

Slide 44, here is the vocal chords. This is the airway. Here are the glands. This area here is the site where we would expect to see metaplasia.

The pathologists were in agreement that there were no changes at this site, even in the hundred times the high real-world group.

To ensure that there were no subtle changes at this site that would perhaps not be detectable by conventional light microscopy, we made examinations using the sophisticated technique of Scanning Electron Microscopy.

This is slide 45 and it's a Scanning Electron Micrograph of the epithelium at exactly the site where the metaplasia would have been seen. Clearly, the epithelium contains non-ciliated and ciliated cells. These are cilia.

No metaplasia had occurred, even in the hundredfold high real-world group. This is from an animal in the hundredfold high real-world group in the area where metaplasia is seen.

This absence of a more severe histopathological change than simple hyperplasia in a particularly sensitive part of the respiratory tract and with long daily exposures to very substantial concentrations of smoke is a major finding that was again confirmed by two independent pathologists.

Slide 46 refers to an additional end point which was built into our 90-day ETS study, namely, cell proliferation. Cell proliferation was assessed in subgroups of animals exposed for the same periods of time as the histopathology animals just described using incorporation of the thymidine analog 5-bromo-2'-deoxyuridine called BRDU. Incorporation of BRDU into DNA takes place during replicative DNA synthesis. This is the process by which cells multiply and, again, here's the reference.

Slide 47 notes that detection of increased rates of replicative DNA synthesis has been advanced as being a component of rodent subchronic toxicology studies that may provide information about the tumorigenic potential of test compounds. The theory is that higher rates of cell turnover could mean an increased risk of errors in the process of DNA synthesis.

Alternatively, increased rates of cell proliferation may have nothing to do with the process of carcinogenesis as reviewed in slide 48. There's an article by Ward, "Cell Proliferation Not Associated Carcinogenesis in Rodents and Humans."

In slide 49, I review the method of administering the BRDU, namely, by placing pumps under the animal's skin for 72 hours immediately prior to necropsy. Such a long exposure period ensures maximal sensitivity for measuring any potential proliferative response.

Slide 50 points out that detection of the incorporation of BRDU into DNA in cells undergoing replicative cell proliferation is made by a complicated series of reactions.

The result of this series of reactions, slide 51, is that cells that are undergoing DNA replication have nuclei which are stained red. Cells with red nuclei are easily counted using light microscopy, allowing them to be counted and quantified using morphometry.

Slide 52 is a section of a rat duodenum. This is used as a positive control because there are high rates of cell turnover in this organ. In this section, cells with nuclei colored in red are undergoing DNA synthesis.

Slide 53 states that the entire respiratory tract was examined plus heart and, of course, duodenum. We made an estimate of cell proliferative rates by examining labelled cells, the number of labelled cells per unit length of epithelium.

Rates of cell proliferation were increased only in the tips of the nasoturbinates, of the rostral nasal cavity and the hundredfold high real-world group after 90 days. That's the only place that we saw an increased rate of cell proliferation. And, of course, this is the same site as the histopathology that we saw earlier.

Slide 54 is a pair of high power photomicrographs of the tips of the nasoturbinates of two different rats. the left section is from a control animal and the right section is from an animal exposed to a hundredfold high real-world concentration for 90 days. Cells undergoing DNA replication are marked in red. Note that they are lined up at or near the basement membrane, exactly where you would expect them to be.

Note also -- it's difficult to see in this slide because there isn't a counterstain, this is the hyperplasia, this thickening here. Here is the basement membrane, along with the cells start to replicate. And then you have that small amount of hyperplasia. But it's occurring at exactly the same site as the hyperplasia we saw earlier.

Note also obviously that there are more red cells in the hundredfold high-real world group than in the control.

Slide 55 shows DNA incorporation rates expressed as cell count per unit length of the nasal epithelium for the four groups after 90 days of exposure. There was no increase over control values for tenfold high real-world. There was a significant increase over controls in the hundredfold high real-world group. This is after 90 days. It's the only time we saw it.

This is after 180 days, slide 56, after 180 days, the cell count again, the same units.

The increased rate of cell proliferation in the rostral nasal cavity of the hundredfold high real-world group was not seen at the end of the reversibility period. The mild proliferative response in part of the rostral nasal cavity is thus completely reversible.

Slide 57 shows some topline conclusions.

There was no effect at all noted in animals exposed to high real-world concentrations at the one or to tenfold higher concentrations. No effect at all. Minimal histopathology changes were noted in animals exposed to hundredfold high real-world concentrations in one area of one organ only.

These latter changes were, of course, reversible. My conclusions are continued in slide 58. Most of the end points showed no effects of exposure at all, even at the massive smoke concentrations that we used. The sensitive proliferation assay was essentially negative, apart from a small region at the very tip of the nose.

Massive exaggerations of real-world concentrations produced tiny changes, changes which in general have been replicated by others, including work in a second species.

Slide 59 gives the bibliographic information for our 90-day inhalation study. Not only have we published the data in the peer reviewed scientific literature, we have donated the animal tissues for archival by the Armed Forces Institute of Pathology. In this way, the tissues from our ETS exposed animals can be examined by other interested researchers.

Slide 60 was given earlier as the reason for performing the experiment. We feel that we achieved the goal of obtaining data on the toxicological responses of experimental animals exposed to different concentrations of ETS.

In summary, the papers reviewed in the OSHA document on inhalation studies with experimental animals are inappropriate. Among other reasons, this was because they did not use the correct test material and because they used grossly exaggerated concentrations of smoke. The study which I have described today is free of the first of these shortcomings. We did use some extremely high concentrations but we also had more realistic concentrations.

Our study shows effectively no toxicological activity of ETS even at massive smoke concentrations in animals exposed for long periods.

The study did not receive the attention it merits in the OSHA document. I hope that my talk today will encourage you to make a full analysis of this important toxicology study on ETS.

My final slide, number 61, states that OSHA's conclusions that "results are supported by animal studies" is simply not supported by the data from those studies.

Thank you very much.

JUDGE VITTONE: Thank you, Dr. Coggins.

Your prepared statement with the slides will be identified as Exhibit 236 for the record.

(The document referred to was marked for identification as Exhibit 236 and was received in evidence.)

MS. SHERMAN: Excuse me, Dr. Coggins. Do you have a full size, hard copy of the slides you showed as opposed to the reduced size in your written text?

DR. COGGINS: Can provide those, yes. I thought it would be easier for you to have them imbedded in the text but they can be provided.

MS. SHERMAN: I think that in terms of a point of reference it is but in terms of actually studying the slide, it would probably be a little bit more useful to have them full-size.

JUDGE VITTONE: Some of the slides that you showed I don't think are included in here, particularly the ones of the --

DR. COGGINS: The photomicrographs have been published. They're in our published manuscripts.

JUDGE VITTONE: Okay. You can provide those at a later time, though.

Thank you very much. We're going to take a 10-minute recess.

JUDGE VITTONE: We resume with Mr. Bohanon.

Sir, would you state your full name for the record, please?

MR. BOHANON: Yes. My name is Hoy Bohanon. I am a professional engineer employed by R.J. Reynolds Tobacco Company. I have studied indoor air quality for the past three years and was involved in many aspects of building management in the ten years prior to that.

Today I will talk to you about engineering solutions for good indoor air quality. I will point out that the solutions proposed by OSHA for indoor air quality should be evenly applied to all substances including environmental tobacco smoke.

There is no need for singular restrictive provisions applied only to smokers either as workers or as customers. Specifically, I will provide details supporting the following three points:

First, there is no reason to single out environmental tobacco smoke from all other indoor air constituents.

Secondly, properly ventilated spaces have very low levels of indoor air contaminants.

Third, OSHA's proposed rules for environmental tobacco smoke are costly and unnecessary.

There is no reason to single out environmental tobacco smoke from all other indoor air constituents.

Without citing any data to indicate the effects of dilution ventilation on environmental tobacco smoke concentrations, OSHA maintains that dilution ventilation is inadequate to address environmental tobacco smoke control.

Dilution ventilation, however, has long been recognized as an effective way to reduce contaminant concentrations and add to indoor comfort.

The concept is very simple. If you burn your breakfast toast, you open the window and turn on a fan. There are many engineering methods to do the same thing with modern HVAC systems. Engineers generally agree that dilution ventilation and filtration are acceptable methods for minimizing concentrations of substances in the indoor air.

It is therefore puzzling that the proposed rule quickly dismisses engineering controls as a viable alternative for control of environmental tobacco smoke levels.

OSHA did not investigate nor discuss the many types of controls that are available.

It is also puzzling that OSHA dismisses dilution ventilation as an adequate environmental tobacco smoke control but mandates it to control a host of substances such as benzene, formaldehyde and styrene, which OSHA identifies as suspected carcinogens within this same proposed rule.

According to OSHA, these substances are emitted not only from cigarettes but also from building materials, interior furnishings, appliances and office equipment and supplies.

In the proposed rule, OSHA has developed several lists of compounds, specifically Tables II-2, III-1 and III-2. Now, many of the chemicals in OSHA's list of compounds "identified in tobacco smoke" are also typical compounds found in offices.

In fact, 29 of the 43 compounds in tobacco smoke can also be found within the same document as emissions from building materials, interior furnishings, office equipment and supplies. Amines or alkanes can be found in adhesives, floor coverings, paints, caulking materials, ceiling tiles and particle board.

How can dilution ventilation be the accepted engineering control for chemicals found in building materials, in furnishings, in office equipment, in supplies, and not be acceptable for the identical chemicals found in environmental tobacco smoke?

In fact, OSHA concluded that dilution ventilation can effectively control occupational exposure to formaldehyde. OSHA's final rule for occupational exposure to formaldehyde established eight-hour time weighted average exposure levels of 0.75 ppm or below.

The rule describes the use of dilution ventilation as the primary means of control in the apparel industry to lower the exposure levels from 1 ppm to 0.75 ppm.

In the Federal Register, the discussion on the apparel industry states, "The workplace is treated like an office or store and air is recirculated rather than exhausted and replaced, allowing formaldehyde concentrations to build.... A relatively simple solution to this problem of air stagnation is to install roof exhaust fans."

The exhaust fan specified by OSHA is a means of using dilution ventilation to lower the formaldehyde concentrations.

Certainly, the primary means of contaminant control in the office environment is dilution ventilation.

Many indoor air quality problems can be solved by designing, installing and operating HVAC equipment with proper rates of ventilation.

The ANSI/ASHRAE Standard 62-1989 entitled "Ventilation for Acceptable Indoor Air Quality" addresses proper ventilation in controlling indoor air problems. The standard's purpose is "to specify minimum ventilation rates in indoor air quality that will be acceptable to human occupants and are intended to minimize the potential for adverse health effects."

The ventilation rate procedure does consider smoking activity within this procedure.

Mr. John Janssen, an ASHRAE fellow and the chairman of the committee charged with writing the ASHRAE 62-1989 standard presented a written statement to Congress in 1991. Quoting from that statement:

"Research on tobacco smoke odor at Yale's John B. Pierce Laboratory has also shown that with today's reduced smoking rate, 15 cfm of outdoor air will dilute environmental tobacco smoke to a level acceptable to 70 percent of people entering an occupied space.... Other calculations on the tobacco smoke perception of non-smoking occupants in a room for 15 minutes with smokers show that non-smoking occupants will register 90 percent acceptance under the assumed conditions. Thus, Standard 62-1989 appears to be able to control tobacco smoke odor under minimum smoking conditions."

Now, for conditions that are suspected to be above minimum smoking conditions, higher ventilation rates were specified within the standard. Bars are required to have twice the ventilation rate, that is, 30 cfm per person, and the rate for smoking lounges is four times the minimum, that is, 60 cfm per person.

OSHA has specified dilution ventilation to be a solution to IAQ problems in the workplace. We agree. OSHA maintains that dilution ventilation will reduce the concentration of chemicals found in the indoor air. We agree. For some unknown reason, OSHA concludes that if the source of the chemical is tobacco smoking, ventilation will not be effective. We strongly disagree.

It appears that OSHA is not making decisions based upon scientific information but is making a moral judgment.

As an analogy, if we are on an elevator that has a capacity of 1200 pounds and the current load weighs 1150 pounds, it matters not if the next person to board the elevator is a felon or a priest, the elevator is overloaded. If an indoor environment is at its capacity, you raise that capacity by ventilating. That is, changing the air changes.

With the many assaults on the air indoors, the only way that smoking can overload the air is if the ventilation is inadequate. Evidence shows that for most cases levels of ventilation established to handle other constituents will also handle smoking activity.

There is no technical reason for making different rules for different sources of indoor chemicals. There is no reason to single out environmental tobacco smoke from all other indoor air constituents.

All evidence shows that properly ventilated spaces have very low levels of indoor air contaminants.

For a number of years, R.J. Reynolds has had an interest in testing the effectiveness of ventilation in real world environments. To assist OSHA in understanding the availability and effectiveness of engineering controls, we have presented OSHA with a number of engineering papers, examining concentrations of chemicals in the indoor environment.

The data show that real world concentrations of chemicals regardless of source can be reduced by dilution ventilation and filtration. These controls have been shown to be effective in various types of indoor environments and such controls can be used to effectively minimize non-smoker exposure to environmental tobacco smoke in virtually every workplace under OSHA's jurisdiction.

These papers and data are unique in three aspects:

(1) Ventilation rates were measured in addition to measuring indoor concentrations of various chemicals.

(2) Smoking activity was assessed by counting people and cigarettes.

(3) Chemicals measured were truly representative of the presence of environmental tobacco smoke in a quantitative sense.

Before reviewing the results of the testing, let me elaborate on these three points.

If you are to study concentrations of environmental tobacco smoke for purposes of characterizing these concentrations within buildings, factors that affect those concentrations must be measured. So the questions are:

Were ventilation rates measured in addition to measuring indoor concentrations of various chemicals?

In order to evaluate data from an indoor air test or perspective or for meaning or certainly before using it to represent a typical national condition, one must know the ventilation conditions.

Now, why do I say that? Why can't one just walk into a room, take a reading on a meter and then generalize on the basis of that data?

Because, as shown in slide 3, ventilation has a major effect on concentrations of chemicals measured in an indoor space. A non-ventilated area can have concentrations from identical sources that are several times higher, as shown on the left side of the curve, than those observed in a ventilated space, shown toward the right side of the curve.

Now, many occupational settings are already well ventilated and almost all have the potential to be. A properly ventilated office will have on average at least twice the air change that a residence will have. A restaurant should be as much as 10 times that of an office. Ventilation prevents buildup of contaminants and does not act in a linear fashion but is very dependent upon air exchange quantities.

The effects of ventilation cannot be discounted or ignored in measuring concentrations in the indoor air.

The second important question is was smoking activity assessed by counting people and cigarettes.

Now, if environmental tobacco smoke is the subject of a study, smoking activity must be measured. A situation where one 100 percent of the people observed are smoking six cigarettes an hour is extraordinary and cannot be generalized as a typical workplace exposure. Likewise, trying to measure environmental tobacco smoke where no smoking occurs is similarly misleading.

The third question, do the chemicals represent the presence of environmental tobacco smoke in a quantitative sense? Measures should include a particulate marker such as UVPM, FM or solanesol, and a vapor phase marker, 3-ethenylpyridine. Nicotine measurements can give a qualitative indication that smoking activity has occurred but that activity may have been days ago. It does not necessarily indicate the presence of smoke.

RSP and carbon monoxide have many sources in addition to tobacco smoke and are most useful quantitatively in a controlled laboratory environment.

The data that OSHA has received from RJR includes four specific tests that I will review. All of the tests meet the criteria that I have just mentioned: ventilation rates are measured, smoking activity is quantified, and an array of environmental tobacco smoke indicators are used.

The first test was conducted in 1991 and was submitted to OSHA in response to its request for information. The test was conducted in order to examine the effect of smoking activities on indoor air quality.

Four buildings were selected. Two of the buildings had constant volume HVAC systems that were approximately 35 years old. The office areas were primarily private offices. The other two were modern buildings. These newer buildings employed VAV systems and the office areas were set up into an open cubicle type arrangement.

Two types of smoking policies were in place within the building. Smoking was unrestricted in two of the buildings and restricted to smoking lounges in the other two. Thirty-eight substances were sampled in the four buildings. Ventilation rates in the two newer systems were monitored and controlled to essentially the ASHRAE recommended minimum rate of 20 cfm per person.

Comparing the results of the two newer buildings, ventilated at the ASHRAE minimum rate, slide 7 shows the carbon dioxide results. For the smoking restricted building, the level is 575 ppm. For the smoking unrestricted building, the level is 604 ppm. Essentially the same level.

Let's look at some more of the chemicals.

Methylene chloride, slide number 8. Smoking unrestricted building, less than 1.1 micrograms per cubic meter. Smoking restricted building, 8 micrograms per cubic meter. In reference to the TLV of 174,000, this is about 20,000 times lower.

Toluene. The smoking unrestricted building, the level was 19 micrograms per cubic meter. Smoking restricted building, the level is 21 micrograms per cubic meter. The threshold limit value of 377,000 indicates the levels were about 10,000 times lower.

Looking at the chemical styrene measured in the two environments. In the smoking unrestricted building, we measured 1.9 micrograms per cubic meter, smoking restricted building, the results were .8 micrograms per cubic meter. Versus a threshold limit value of 213,000, approximately 100,000 times lower.

Formaldehyde. Smoking unrestricted building measured .034 parts per million. Smoking restricted building measured .033 parts per million. The referenced TLV of .75 parts per million, about 20 times lower.

Carbon monoxide. The smoking unrestricted building, the level was approximately 1 ppm. Smoking restricted building, 2.4 ppm. The TLV is 25 or approximately 10 times lower.

Benzene. For the smoking unrestricted building, we measured 4.8 micrograms per cubic meter. Smoking restricted building, 1.7 micrograms per cubic meter. A TLV of 32,000, approximately 10,000 times lower.

Nicotine. In the smoking unrestricted building we found a level of 2.3 micrograms per cubic meter. The smoking restricted building less than 0.1 micrograms per cubic meter. A TLV of 500, the levels observed were about 200 times lower.

RSP. For the smoking unrestricted building, 30 micrograms per cubic meter. For the smoking restricted building, 5 micrograms per cubic meter. A reference TLV of 5000 in slide 15.

The second test submitted to the current docket was designed to evaluate the effects of ventilation rates and distance on environmental tobacco smoke in a large office building.

This test measured the concentrations of many environmental tobacco smoke indicators in several areas on a floor in a large high rise office building where smoking is unrestricted

As a part of the test, outside air ventilation rates were varied from no mechanically delivered outside air to the outside air dampers wide open.

The conclusions of this test were:

That for open office areas on one floor of a large modern office building with VAV HVAC technology, where 31 percent of the occupants were smokers who each consumed on average 9.97 cigarettes per seven and a half hour day, changes in the ventilation rate, the smoking rate or distance from smoker were accompanied by measurable changes in environmental tobacco smoke concentration.

Statistical models indicated that ventilation, smoking activity and distance from smokers are all significant factors in explaining levels of environmental tobacco smoke. The results of this study are consistent with predictions of models that are based upon fundamental physical principles.

These results indicate that increased ventilation or smoker separation can reduce environmental tobacco smoke exposure.

Concentrations of most environmental tobacco smoke indicators varied inversely with ventilation rate and directly with smoking activity. These indicators were nicotine, 3-ethenylpyridine, UVPM and FPM. Again, such behavior is consistent with predictions from physical models relating ventilation and contaminant.

However, RSP, one of the more commonly used indicators of exposure, did not exhibit this relationship. The absence of both ventilation effect and smoking activity effect is theorized to be due to RSP originating from other sources. This result underscores the limitations associated with this indicator.

The third paper presented and examined the effects of ventilation and separation of smokers and non-smokers on the exposure of non-smokers to environmental tobacco smoke in an office using both personal and area monitoring techniques.

Slide 17 shows a diagram of that 5500 square foot office space.

Due to the apparent lack of information regarding the relationships of dilution ventilation and environmental tobacco smoke compound levels, R.J. Reynolds conducted this test. The results of this test clearly demonstrate the effectiveness of ventilation and separation as engineering control techniques.

Now, in this test, the term separation means establishing an area for smokers and non-smokers. It does not mean separate enclosed rooms.

A 5500 square foot office space was tested. The space was served by one dedicated HVAC system. The HVAC system in this test was a constant volume reheat system with the capability to provide outside air by opening an economizer damper.

The test protocol was to measure indoor concentrations of chemicals in the space for two days with no smoking allowed. Then for eight following days 160 cigarettes were smoked each day. For the first four days, no outside air was provided by the HVAC system by keeping the damper closed. The following four days, the ventilation system damper was opened to provide outside air. For two days under each condition smoking was restricted to a designated area located near the return air vent. This designated area is shown in yellow on slide 17. Note once again that the separated area was not enclosed by walls or doors.

Personal monitors were used to measure nicotine and 3-ethenylpyridine. The results for specific markers show that all levels are low.

Slide 18 shows that carbon monoxide levels were not affected by smoker and non-smoker separation. Carbon monoxide levels were affected by cigarette smoking activity, outside air levels and outside air ventilation rates.

Indoor carbon monoxide levels exceeded outdoor levels by approximately 2 ppm during the test days of cigarette smoking and no outside air ventilation.

Now, carbon monoxide levels outside were not statistically different from outside levels during days of cigarette smoking with the outside air damper open.

Slide 19. Nicotine levels were most affected by smoker/non-smoker separation. Nicotine levels also were affected by outside air ventilation rates.

Personal nicotine sampling measurements are not significantly different in this case than the work area sampling measurements.

RSP levels on slide 20 were not statistically different in smoking or non-smoking locations nor in either mode of ventilation. The average RSP levels for the test duration were 32 micrograms per cubic meter and 29 micrograms per cubic meter for smoking and non-smoking respectively. In fact, the RSP levels inside were not different from the outside concentrations of 30 micrograms per cubic meter.

UVPM shown here on slide 21 and FPM levels were found to be good tracers of environmental tobacco smoke particles. These particles were found to be at essentially the same levels in the office area as the outside air during the days when the outside air dampers were opened.

Slide 22 shows the same result for FPM.

Slide 23 shows that environmental tobacco smoke RSP based upon solanesol measurements were affected by ventilation rates and indeed found to be below the limit of detection on all days when the dampers were opened.

The conclusions:

During the days where outside air dampers were opened and smoking was restricted to the designated area, carbon monoxide, carbon dioxide, RSP and UVPM levels were found to be essentially the same as outdoor air levels.

Overall levels are low when compared to OSHA's estimates of environmental tobacco smoke concentrations or exposures in the workplace.

Under all conditions the levels are many times lower than PEL limits and, indeed, even below the limits set for outdoor air.

By understanding these relationships and applying separation and ventilation procedures, exposure to environmental tobacco smoke can be virtually eliminated in non-smoking areas.

If ventilation can be effective in this case, then the technology is a valid engineering control and should not be discounted by OSHA.

This research was designed to test the effect of engineering controls in effectively minimizing environmental tobacco smoke concentrations. Additional research is needed to determine optimal ventilation rates in separation distances between smokers and non-smokers.

The fourth paper deals with the effects of ventilation and filtration on reduction of exposure in restaurants.

OSHA requested specific information regarding bars and restaurants. RJR has submitted data and analysis that demonstrate the effectiveness of ventilation and filtration in reducing environmental tobacco smoke components in bars and restaurants.

These reductions were possible with existing standard technologies. For all areas of the restaurants, the concentrations are very low. For the non-smoking sections, concentrations are significantly lower than the smoking sections.

In order to test the effectiveness of HVAC technology in reducing environmental tobacco smoke concentrations in the real world, we sought out restaurants that served a lot of smoking patrons and had poorly performing HVAC systems.

A test and balance contractor, that is, a contractor certified in measuring HVAC systems, evaluated and documented the performance of the existing ventilation systems. Air flow rates on kitchen exhaust hoods, makeup air units, HVAC supply airflow rates and outside air flow rates were measured.

We then made cost effective changes projected to lead to better indoor air quality. These changes included increasing outside air ventilation rates and, in some cases, improving filtration.

Air samples were taken during operating hours in both the smoking and non-smoking sections of the restaurant. Before we changed anything in the HVAC systems, we sampled for seven consecutive days. After we modified the HVAC system, we again sampled for seven consecutive days.

Slide 26 shows the layout of the restaurant with the highest levels measured. The smoking section concentrations were measured in room C. The non-smoking concentrations were measured in room A.

The results were dramatic reductions in RSP in both the smoking and non-smoking sections. We used the solanesol method developed by Dr. Ogden to analyze the contribution of environmental tobacco smoke to respirable particulate matter.

The environmental tobacco smoke RSP levels were reduced from a median of 110 micrograms per cubic meter to a media of approximately 20 micrograms per cubic meter. Reductions in median values of particles were found to be a direct result of increased ventilation with outside air and the electronic air cleaner installed on the HVAC unit serving the smoking area.

In cases where high levels of carbon dioxide, nicotine, 3-EP and particulate matter were found, reductions were achieved. In other cases, better control of thermal comfort resulted.

Positive comments were received from both customers, restaurant owners and workers after the changes were made.

In summary, when one looks at all of the data that was presented, the conclusions are clear.

All markers and indicators fit the theoretical ventilation concentration model at high concentrations.

Secondly, for very low levels of any substance the relationships are hidden by noise from uncontrolled factors and variance of measurement techniques. The substances most affected by these factors are RSP, carbon monoxide and nicotine. Carbon monoxide and RSP are confounded by multiple sources, and nicotine is affected by physical instability, that is adsorption and re-emission from surfaces.

The markers of environmental tobacco smoke exhibit the same characteristics as markers for other substances, such as CO2 or TVOC.

Therefore, one must conclude that the same techniques, namely, ventilation and filtration, can be used for tobacco smoke.

A review of the submitted papers and a critical examination of the research available to OSHA clearly demonstrates that adequate ventilation, coupled with separation of smokers and non-smokers effectively minimizes environmental tobacco smoke concentrations in large offices, small offices and restaurants.

Effective engineered solutions available to OSHA may be much less costly than the OSHA prescribed smoking rooms and would certainly be less disruptive to American workers and the American workplace.

We have observed that there is no reason to single out environmental tobacco smoke from all other indoor air constituents. Scientific evidence indicates that properly ventilated spaces have very low levels of indoor air contaminants.

My third and final summary point in my comments today is that OSHA's proposed rules for environmental tobacco smoke are costly and unnecessary.

Many economic, enforcement and logical deficiencies of the proposed rule would be eliminated by treating environmental tobacco smoke as any other indoor air component.

Now, apart from the lack of substantial scientific evidence to justify the proposed smoking regulation, which is extensively detailed in other comments submitted by R.J. Reynolds and others, and apart from the fact that the proposed rule would unnecessarily duplicate efforts that are currently well handled by the free market, the proposed rule suffers from major deficiencies in three key areas: cost, enforcement and a justification.

First, cost. OSHA has overestimated the benefits of environmental tobacco smoke regulation while dramatically underestimating the implementation costs both in terms of real dollars and disruptions to the workforce and workplace.

In addition, there are hidden costs that have not even been considered in the proposed rule. OSHA has dramatically underestimated the cost of constructing smoking rooms, apparently because the agency misinterpreted R.J. Reynolds estimates of smoking lounge costs.

In 1992, R.J. Reynolds published a brochure entitled "Developing a Smoking Lounge." This publication, which apparently is the source of some of OSHA's cost estimates, was intended for businesses in an office setting and envisioned a typical low rise office building. It was assumed that an office or other room would be existing and that some furnishings would be available. This would be the case in the typical office, where it was then estimated that the cost of ventilating a smoking lounge would range from $1500 to $4000.

It is important to note that this estimate does not include any room construction or furnishing costs and that the estimate would in no way apply to high rise buildings, restaurants or other types of facilities.

Based on these figures, OSHA chose one number, $4000, and applied it universally to determine the national cost of constructing smoking rooms. This is clearly an inaccurate approach.

To provide more accurate and detailed information for estimating costs, R.J. Reynolds commissioned Raytheon Engineers & Constructors to evaluate potential costs of providing a 150 square foot room for smokers in a variety of workplace settings. As detailed in our submitted comments, the costs range from a low of $2481 to a high of $21,253.

The costs are provided by building category and regional adjustment factors are also provided.

In order to make an accurate estimate of the total national cost of constructing smoking rooms, the agency should calculate according to building type and geographic region. When such calculations are made, it is clear that OSHA dramatically underestimates the economic impact of the proposal.

OSHA has overestimated the cost savings that could result from smoking bans because the agency uses unreasonable estimates for the costs of accommodating smokers.

For example, OSHA has estimated the cleaning cost savings from smoking bans to be $500 per smoker per year. Using numbers from BOMA to calculate cleaning costs, the average annual cleaning cost per person in an office building is about $250 per year. How can a business save $500 per year when it is only spending $250?

The second area, enforcement.

The proposed rule presents enforcement problems that fall within two major areas. If the proposed rule takes effect, it will divert and dilute the resources that OSHA and businesses now devote to enforcement of important existing regulations that protect the health and safety of American workers.

Even if additional resources were available to focus solely on environmental tobacco smoke regulations because of the broad scope of the proposed rule, those resources would ultimately be ineffective because the proposed rule as written is essentially unenforceable.

The OSHA smoking police would have to be in every bar, every restaurant, every mechanic's garage, inspect the American Legion hall and follow the truck drivers to completely enforce this rule. The potential for arbitrary enforcement is enormous.

The third area, justification.

OSHA has not demonstrated any justification for significant provisions and premises within the proposed rule. For example, the provision that no work of any kind can take place in an area where smoking occurs.

If smoking is restricted under rule to enclosed exhausted rooms and we assume that these rooms properly protect non-smokers, why can no one work in the room and smoke at the same time?

How does such a restriction improve productivity?

What valid scientific reason can there possibly be for preventing work from taking place?

The provision that rooms where smoking takes place must be negatively pressurized and totally exhausted. What about filtration? Why is it discounted when this technology is used for many other substances?

There is no evidence cited or examined to support OSHA's unreasonable proposal.

The underlying premise that smokers and non-smokers alike need to be protected from environmental tobacco smoke exposure, even if they would voluntarily allow themselves to be exposed.

What significant workplace risk is eliminated by preventing a waiter who is a smoker from working in a smoking-permitted section of a restaurant?

Justification for this rule is lacking simply due to the fact that the free market system is working.

As OSHA must know from examining the docket, the more recent the measurements of workplace concentrations or exposures to environmental tobacco smoke the smaller the quantity.

Is this because instrumentation is becoming less and less sensitive?

Certainly not. It is because in a free society personal preference and opinion mater. If employees desire that levels be reduced, then employers respond.

Businesses are establishing smoking policies in increasing numbers and most businesses now have formal smoking policies. Through these policies, businesses determine the best solution for their particular employees, their building situation and their geographic location.

There is no evidence that the process of setting policies has failed in any manner nor is there any reason to believe that setting a single standard, such as in the proposed rule, will create any advantages not inherent in the free market approach.

If anything, setting a single standard will diminish the ability of businesses to best accommodate their employees and their customers.

In fact, OSHA has ignored the workings of the free market as well as incorrectly estimated the potentially exposed population by assuming a uniform 73.01 percent non-smokers in every workplace.

People employed in different occupations have different rates of smoking. Smoking policies that evolve from market forces attempt to accommodate those differences.

For example, physicians are about 94 percent non-smokers. Their workplaces, hospitals and clinics, are virtually all non-smoking areas. Therefore, this population of non-smokers have virtually no exposure to environmental tobacco smoke.

In the other hand, many construction workers are smokers. Fifty-eight percent of roofers, 46 percent of carpenters, 55 percent of drywall installers, 52 percent of masons and 47 percent of painters are smokers. Many of these workers work outdoors. Some of them work indoors. And smoking policies in the construction industry are not the same as for hospitals. Nor should they be. The employee base and the workplaces are different.

It makes sense for the service station attendants, 44 percent of whom are smokers, to smoke indoors away from the gas pumps to protect their own safety and the safety of others.

In addition, there are differences in enclosed spaces in different geographic regions. Construction and ventilation are very different in a typical building in Minnesota compared to a typical building in Puerto Rico.

A free market system can accommodate and reasonably arrive at unique solutions, taking into account building construction, ventilation systems, geography, demographics, business types and individual attitudes to arrive at an optimum answer.

A dictate from Washington that represents an all or nothing, one size fits all approach can never accommodate all the factors that a free system can.

In conclusion, OSHA should seek out and use the best available data to determine what rules should be established regarding indoor air.

Careful evaluation of the data that is representative of real workplaces would lead OSHA to conclude that there is no reason to single out tobacco smoke from all other, and in some cases identical, chemical compounds found in indoor workplaces.

OSHA should change their proposed rule to be consistent in logical application of ventilation recommendations by treating environmental tobacco smoke like all other substances because there is no reason to single out environmental tobacco smoke from other indoor air constituents.

There are indeed a wide variety of engineering controls that can effectively minimize environmental tobacco smoke concentrations to which non-smokers are exposed.

These are the same techniques prescribed by OSHA for controlling identical chemicals emitted from other sources into the indoor air. These techniques work.

Properly ventilated spaces have very low levels of indoor air contaminants. American businesses should have the option of employing any effective controls in lieu of the restrictive environmental tobacco smoke provisions contained in the proposed rule.

Because of cost considerations, the proposed rules would effectively prohibit smoking in most workplaces, a prohibition that cannot be justified when judgment is made on a scientific basis.

OSHA's proposed rules for environmental tobacco smoke are costly and unnecessary because properly ventilated spaces have very low levels of indoor air contaminants.

R.J. Reynolds Tobacco Company agrees that indoor air quality is an important issue and that additional safeguards are necessary. We also believe that the ventilation based approach that the proposed rule takes concerning indoor air contaminants in general represents a sound and rational basis for improving indoor air quality.

Buildings should be ventilated to the rates established by the American National Standard/ASHRAE 62-1989. For those buildings unable to comply during all weather conditions, they should be operated to the design capacity for their original code providing that the use has not changed.

The best way to assure that buildings are properly operated is to assure that the operators are qualified. Qualified operators will recognize the vital role that maintenance plays in assuring good indoor air quality. Properly ventilated spaces have very low levels of indoor air contaminants.

For the overwhelming majority of employers in the United States, there is little doubt that the proposed rule would result in total smoking bans. Costs associated with the proposed smoking restrictions, including the costs of lost productivity, would be too great for most employers to bear.

Employers are already free to institute smoking policies if there are legitimate reasons for doing so. Federal regulation provides no clear benefit that is not already being realized through the current market efforts and, in fact, would unnecessarily burden American business with overly restrictive and inappropriate regulations.

Thank you.

JUDGE VITTONE: Thank you, Mr. Bohanon.

Mr. Bohanon's printed statement will be identified as Exhibit 237 for the record.

(The document referred to was marked for identification as Exhibit 237 and was received in evidence.)

JUDGE VITTONE: And the slides that he used during his presentation will be identified as Exhibit 238.

(The document referred to was marked for identification as Exhibit 238 and was received in evidence.)

JUDGE VITTONE: Let's take five minutes.

JUDGE VITTONE: We resume our hearing. Let me turn first to Ms. Sherman.

DR. COGGINS: Your Honor, I do have a summary.


DR. COGGINS: A one page summary, yes.


DR. COGGINS: To sum up Reynolds Tobacco's presentation, we believe that if one strips away the emotional issues involved with smoking and the political and social climate that currently surround cigarettes and smokers, it becomes clear that there is no justification -- scientific or otherwise -- to regulate ETS separate and apart from other indoor air components. A perspective that views ETS within the context of its contribution to total IAQ is clearly the most appropriate and effective approach.

Now our panel will be happy to answer questions concerning our presentations. As I noted earlier, I will direct the questions to the appropriate panel members.

JUDGE VITTONE: Thank you, Dr. Coggins.

Now let me turn to Ms. Sherman.

MS. SHERMAN: How many facilities do you have in the United States?

DR. COGGINS: Does R.J. Reynolds have?


DR. COGGINS: I have no idea. Does anybody on the panel know anything about this?

MR. BOHANON: Give me a minute to look in our submittal.



JUDGE VITTONE: Ms. Sherman, you've stumped them n the first question.


MS. SHERMAN: It was not my intent.


MR. BOHANON: Sorry. It's in this table. Most of our facilities are in North Carolina.

Let me just say, since I can't put my finger on the number right now, it is several hundred buildings in approximately...

MS. SHERMAN: I believe you said in one of your submissions that you're the owner of 372 buildings. You being R.J. Reynolds, not Mr. Bohanon.



MS. SHERMAN: However, that really wasn't my question. My question was, how many facilities do you have? by that I mean, after all, you may have some workers in work sites that are not owned by the company in various organizational units, et cetera.

MR. GROSSMAN: Ms. Sherman, our submission on page one says that Reynolds is a major U.S. manufacturer of tobacco products, and employees approximately 9,000 employees in over 370 buildings in North Carolina.

MS. SHERMAN: I understand that. Have you no facilities in any other state?

MR. BOHANON: The answer to that, if I don't have to find an exact number, yes, we do. We do have some regional sales offices and do have some other operations that are minor in nature outside of the State of North Carolina.

MS. SHERMAN: Can you provide for the record the number of facilities that you have within the United States, and can you break it down into industrial facilities and non-industrial facilities?

MR. BOHANON: Not at this time, but I think we can provide... You're talking about buildings where our employees work, whether we own them or lease them or...

MS. SHERMAN: That is correct.

MR. BOHANON: I think we can come up with a good estimate on that number.

MS. SHERMAN: To the extent that you have data on the type of leased facilities therein, whether there are other employers in the same facility, we would find that useful, also. And to the extent that you can give us an idea of the percent of your employees who are located in non-industrial facilities, we would find that data useful also.

MR. GROSSMAN: If I may clarify this. By when would you like this data?

MS. SHERMAN: I believe during the post-hearing comment period would be good enough.

MR. BOHANON: Let me just make sure. You want owned and leased information. You want industrial and non-industrial.


MR. BOHANON: Numbers of facilities and numbers of employees.


MR. BOHANON: Also you want to know...

MS. SHERMAN: Industrial and non-industrial.

MR. BOHANON: Right. And you also had a question about the number of facilities where there are employees of other companies that are in the same facilities with our employees?

MS. SHERMAN: In the same building, not necessarily in the same facility. You might have separate office space, and 16 other people may have office space in the same building.

MR. BOHANON: Okay. When we speak of facilities, I think of buildings. I count buildings. So in the same building...

MS. SHERMAN: It's perfectly possible, especially when you're dealing with your out-of-state sales establishments, that you would not occupy an entire building, is it not?

MR. BOHANON: It certainly is. In fact it's possible and exists right now in North Carolina in some of our own buildings.

MS. SHERMAN: So I'm trying to get some sort of an idea of a workplace facility profile of your many workers.

MR. BOHANON: We will do our best to provide that.

DR. COGGINS: Workplace facility profile, okay.

MS. SHERMAN: That's a term of my own creation right now, but I thought it was descriptive of what I was asking.

DR. COGGINS: If Mr. Bohanon understands it, we're in good shape.


Do you have an indoor air quality program at R.J. Reynolds?

DR. COGGINS: I'm going to let Mr. Bohanon answer that question.

MR. BOHANON: Yes, we do have an indoor air quality program, and it really hinges on two factors. One is the operational factor which I view as being very important, probably the most important, which is the preventive aspect, and that is that we operate our buildings to the national ventilation standards. We make sure that our buildings are properly ventilated. Then the second factor involved with that is that the maintenance operations are under the direction of professional engineers, so that those maintenance practices are good maintenance practices, and we feel that that provides a high quality of air within the buildings. In fact one of the buildings that I talked about where we have done extensive measurements is, indeed, one of our own facilities.

The second piece of that has to do with how we might respond to complaints regarding indoor air quality. We have within our various operating units health and safety technicians that look after many aspects of the health and safety of our operations and compliance with all the many health and safety rules that are in place. Those health and safety technicians really are the first line of investigation for any sort of issue that may be an indoor air related issue. If indeed they find that it's something that's not readily dealt with, they then communicate with the industrial hygienist who then engages our occupational physician to take over and investigate the issue.

MS. SHERMAN: For how long have you had this indoor air quality program?

MR. BOHANON: Well, Reynolds has had a very long history of having a safety program that goes back into the 1920s. The industrial hygiene portion, of course, was added later as that became a field. The current organization with the health and safety technologists I believe was put into place in bout the last five years with that structure that I have described to you.

As far as the operating policy in the maintenance side, which is the other part which I view as being the preventive side and the part that's going to prevent having problems, to my knowledge, and certainly since I have been with the company...

MS. SHERMAN: You said that was three years?

MR. BOHANON: No, I have been with the company 15 years.

MS. SHERMAN: Oh, okay.

MR. BOHANON: Three years investigating indoor air quality, ten years prior in building operations and management. And the people who are accountable for operating the air conditioning systems, the ultimate accountability, always rested and still does rest with people who are professional engineers who evaluate good practices and standards. So we look at many different factors.

We have an extensive cooling tower water treatment program, we have a lot of aspects in our specific maintenance policies that have been in place as I said, as long as I have been with the company. In fact that was one of my accountabilities at the time was I had the accountability for maintenance of about 1.5 million square feet of office space, and it was always an awareness to make sure that the systems were in good, clean operating order and that they were properly operated.

MS. SHERMAN: I believe you said earlier in your answer that you maintain your facilities to, I think you termed it national ventilation standards. Are you talking about ASHRAE standards or some building codes? To what were you referring?

MR. BOHANON: I'm referring to the ASHRAE standard, the current American national standard.

MS. SHERMAN: So all of your non-industrial facilities you believe would meet ASHRAE 62-89?

MR. BOHANON: Yes, I believe all of our non-industrial facilities are designed to meet ASHRAE 62-1989. On the specific ones where we have made measurements and measured accurately the ventilation air, we know that the performance does, indeed, comply with that standard.

MS. SHERMAN: To the extent you've made measurements, you've found that you're providing 20 cfm per occupant, I believe, or is it 15?

MR. BOHANON: For an office building, the recommended number is 20, and yes, as a minimum we would provide that. We also, in general, for variable air volume in other types of systems have economizer types of systems which would then increase the amount of outdoor air during many weather conditions that in effect save energy and provide high rates of ventilation concurrently.

MS. SHERMAN: When you were talking about the maintenance of your ventilation systems, et cetera, these are under the supervision of professional engineers?

MR. BOHANON: The individual accountable in the organization has always been, to my knowledge, a professional engineer. There are many technicians that actually carry out the duties, and in fact some of the work like cooling tower water treatment is subcontracted to a specialty chemical contractor that can properly maintain the chemicals in the cooling towers.

MS. SHERMAN: That was going to be my next question. In terms of the maintenance of your HVAC systems, do you tend to contract it out or do it in-house?

MR. GROSSMAN: You're asking, Ms. Sherman, with regard to every facility in the United States?

MS. SHERMAN: No, actually at this point I believe I'm asking about facilities that Reynolds owns.

MR. GROSSMAN: And only those facilities?

MS. SHERMAN: Yes. However, I may get into what they don't own and if they do have a policy.


MR. BOHANON: Thanks for that clarification.

When I am speaking of my direct experience and knowledge, it has to do with those buildings that I'm familiar with that we own in North Carolina. I can't speak to locations that I don't know about.

For the buildings in North Carolina, which are the bulk of the buildings in square footage that we occupy, and I'm sure when we go back and add up the numbers that will be correct.

For those buildings, we have had over the years, I guess a number of building owners and operators have had various times where activities are out-sourced or contracted out, and then for other reasons you decide to bring the work in-house, and then you decide again to contract it out. So my experience has been with both. There are times and specific things that have been done by in-house personnel, and there are things that have been contracted out. I can specifically give an example like filter changing. There are times when that has been done solely by in-house personnel, and there are other times when it has been contracted out.

When it's contracted out, then there are specific requirements as far as timing of changing filters. Again, it's going to depend on a system-by-system basis, because we have many different kinds of filtration technologies within our buildings, and the proper and most economical way to make sure that filtration is adequate without any kind of sloughing off and having the filters not working, depends upon the type of filter.

MS. SHERMAN: Do you have any procedure for qualifying a contractor to work for you?

MR. BOHANON: The answer to that is yes, there are many qualification procedures that I certainly can't cite that our purchasing department administers relative to a contractor's qualifications -- financial and otherwise. For specific contractors to do a specific job, I can recall and give you the instance where we needed someone to perform preventive maintenance on a large, 600 ton centrifugal chiller. In that case, since I was accountable, I was very interested in that contractor's qualifications. I think I interviewed two or three. So I guess that I was interested in that, I interviewed, I talked to other people that the contractor was doing work for. I talked to their engineers, examined their protocols and proposals before I made a decision on which contractor was going to give us the best job, which includes not only the cost but the quality of the work that they perform.

I cannot say that everyone in every instance goes through that, but I would generalize that the accountable engineers involved are concerned with their accountability, and concerned that contractors perform properly.

DR. COGGINS: If I could just add a comment there. As an end user, I've been with the company for ten years and been in a number of different facilities, and the rest of the panel has, too. I've always found that we've gotten very, very good service from these people. We get E-Mail messages saying the water's going to be cut off this day, the air's going to be slow this day. As an end user, we work in very, very comfortable facilities that are extremely well maintained.

MS. SHERMAN: Is there any company policy as to any particular tasks involving your heating and ventilation services which you would not contract out?

MR. BOHANON: That's an interesting question. Let me think about it.


MR. GROSSMAN: Are you responsible for deciding whether to contract out the services?

MR. BOHANON: I am not currently, but in my previous job experience I was, and I do feel comfortable in characterizing how those decisions are made. I would say that other than the overall management and strategy of how to perform the maintenance that there are probably qualified contractors that can be obtained to do all of the things that we do. We choose to do some things in-house because of just access, work schedule, knowledge of power plants and how those facilities inter-relate to... The utilities operations interrelate to the industrial operations. So there are a lot of reasons why we make the choices that we make.

DR. COGGINS: Ms. Sherman, you are going to find out we have a very unruly panel.

One of my colleagues now wants to make a comment, and we're going to give you answers as and when they come. This is Dr. Sears.

DR. SEARS: Thank you.

Actually, I'm not sure that this is a comment so much as it is a question of my colleague, Mr. Bohanon.

Am I correct that our OSHA responsibilities, that is the OSHA requirements for maintaining safe laboratory spaces, is something that we don't contract out? That we ourselves are responsible for that type of...

JUDGE VITTONE: Can you answer that question?

MR. BOHANON: Yes, I can. That's into another area. The OSHA compliance type of activities we do not contract out. We assure that we are in compliance with all of the OSHA regulations. Our safety and health technologists are all company employees as are the people who have the primary operations accountabilities, and indeed the supervision of all of the maintenance accountabilities regarding our air handling systems.

MS. SHERMAN: I believe you said that you're primarily familiar with your buildings that you own within the State of North Carolina. Would you happen to know what the oldest of these buildings is?

MR. BOHANON: We have undergone a lot of changes in buildings in the 15 years I've been there, which is why, if I give you a current number, the number is constantly changing because we're selling buildings and in transition.

I would speculate at this point in time that the oldest building that we currently own is an office building. In fact the Reynolds building which was built in 1927.

MS. SHERMAN: This building is able to provide 20 cfm per occupant?

MR. BOHANON: Yes. Keep in mind that air conditioning didn't come along really until the '50s and '60s. The type of system that was installed in that building which was installed in 1957 is a constant air volume induction system. A system that puts either hot or cold water on a coil under an induction unit, under each window in the office building, and then provides either heated or cooled outside air to induce air flow across the coils in the building. So by design, that system does deliver a high amount of outside air, and I can estimate, I think we have in our written submissions talked about that building. It's maybe between 60 and 100 cfm per person that it delivers all the time the system is operating.

MS. SHERMAN: So you have updated the systems in many of your older buildings to meet ASHRAE standards?

MR. BOHANON: I don't think I can answer exactly that way, that we've updated systems to meet ASHRAE standards.

Let me answer it this way. As we install most, again for most of our major facilities, the big office buildings and the factory buildings, and even the research type buildings and the other buildings, we have found in North Carolina that it's very economical when you have a building that's bigger than a house, that it's very economical to provide a thing called an economizer cycle. When you provide that economizer cycle, it gives you a lot of capacity in being able to bring in outdoor air. The only issue is one of the design end points, the coldest day in the winter or the hottest few days in the summer, when there might be any issue. The issue of bringing air in is not a constraint of the type of air handler that we install, but is more a constraint of perhaps capacity to provide comfort conditions. I think that may be the case in a lot of locations.

MS. SHERMAN: The reason I'm asking you this question is that I understood you to say at the beginning that the buildings that you owned met what you termed to be national ventilation standards, and I asked you what that meant, and we decided it meant ASHRAE, 62-89.


MS. SHERMAN: Now I'm wondering if your statement also included your older building systems.

MR. BOHANON: Yes, my statement does include our older building systems. Indeed, we operate the systems. The systems have the capacity. They move air, they're located so that they can bring air into the building, and we find that with proper ventilation we have very low levels of any sort of indoor air contaminant.

DR. COGGINS: Your Honor, I see that we've reached the time that was set aside for the oral argument.

JUDGE VITTONE: Wait a minute.

Ms. Sherman, where are you right now?

MS. SHERMAN: I have a few more questions to close out this particular area, if I might.

JUDGE VITTONE: Sure. Let's do that.

MS. SHERMAN: As part of your indoor air quality program, do you catalog any employee complaints that you receive about the air quality?

MR. BOHANON: Let me rely back on my experience of 10 to 12 years in building management. Building managers get a lot of complaints. Most of the complaints have to do, you might categorize them as air quality. They really have to do with the temperature. It's too hot. It's too cold. We really don't systematically keep long records of all of those hot and cold calls. In fact, our employees expect a pretty quick response and we try to do that within a few hours time frame to respond and make adjustments.

So I would say broadly that there are a lot of those kinds of issues that aren't recorded beyond the fact that somebody went and adjusted the thermostat or adjusted the system and closed it out and there's a maintenance record that's on-file there for awhile in the building operator's office.

If you're talking about indoor air quality complaints in a more narrow view of not just hot and cold calls, but of symptoms, then yes, those do get recorded, but they're not recorded by solicitation. We don't encourage our employees to tell us every time there's a headache, or every time their eyes are dry. But when people perceive that there's something in the air quality within the building that's causing these issues, then they let the health and safety technologists know or the building manager know, and then it gets into the system of consulting with the industrial hygienists in investigating specifically what may be going on in that area.

MS. SHERMAN: Do you investigate every complaint that is not viewed as a comfort-based complaint? Or you only investigate trends in complaints?

MR. BOHANON: Every complaint that is not a comfort-based complaint is investigated first by a health and safety technologist and then if it's something that they can't resolve it's elevated pretty quickly to the industrial hygiene group and the occupational physician.

MS. SHERMAN: And in your past experience, have you found that this is a useful early warning system of perhaps a deeper problem in a building?

MR. BOHANON: No, I really don't think I could say that because we haven't found any major problems in the building operation when we've investigated those. That's really one of the first things we do, is go look and see if somebody's put some pesticide somewhere, you know, we go check all those kind of records and we really do that in the background before we go traipsing around on a floor with instrumentation. We really want to know how is the system operated. We do a visual inspection to see if there's any visible growth or anything within the system, whether the dampers are operating properly, what sort of temperature conditions because temperature and humidity can also have an effect. If the humidifier is not working in the winter and it gets very dry then you can have dry eyes and humidity type indoor air complaints. So we really try and check all those mechanical things out first. And I don't recall ever finding anything significantly out of order with the HVAC operations within those buildings that we have there in Winston-Salem.

MS. SHERMAN: Does Reynolds have a policy about pesticide application during work hours in its facilities?

DR. COGGINS: Yes. I think, Hoy, you'll have to answer that.

MR. BOHANON: Our practice is to apply pesticides after hours and not during work hours. I think it would be extremely unusual for us to apply a pesticide during work hours except I can recall, again, within my experience a couple of instances where someone wanted a pesticide person to come to their office right then because there was a bug and in that case we thought that the lesser of two evils was to apply some pesticide and that would satisfy the person's complaint. But in general, the more uniform applications or regular applications are done in off hours.

MS. SHERMAN: Is this procedure contracted out or do you do it with your own employees?

MR. BOHANON: No, we contract that out. We are not licensed pesticide applicators. I don't think we have any within our company.

MS. SHERMAN: Your Honor, I think that that ends that line of questions. I could happily go on.


MS. SHERMAN: Happily. Gleefully.

JUDGE VITTONE: Well, I have about seven minutes after. If that's a good point for you to break, then why don't we break right there, okay? And then we will resume with you tomorrow morning on the OSHA panel at 9:30.

And I would really -- and I know I'm probably as much of a violator as anybody but I really want to begin at 9:30 tomorrow. I will within reason without running any red lights or anything be here at 9:30.

MS. SHERMAN: Your Honor, we would be available to start earlier, if you wish.

JUDGE VITTONE: Actually, I have a conference call scheduled at 9:00 tomorrow morning and it won't take me very long but it's one that I had to have scheduled, so I would like to begin at 9:30, okay?


JUDGE VITTONE: I'm sorry, did somebody else over here have something to say?

MR. GROSSMAN: We have a conference at 9:00, too.

JUDGE VITTONE: Okay. We'll begin at 9:30.

Before we break, though, Mr. Furr said he would like to make a point on the record.

MR. FURR: Your Honor, this may take some time.




MR. FURR: Your Honor, with your leave, what I would like to do is address the problem as we see it of Drs. Ford and Bayard testifying on the same day.

I'm going to try to be as brief as possible but we feel very strongly about the problems that are created by them testifying on the same day and would like for you to have the advantage of our full position as you consider this motion.

Both Drs. Ford and Bayard are extremely important witnesses and will require some length of time in order to do a thorough examination of their views and how they impact OSHA's rule. And what I want to do is explain to you why we believe that's so.

With respect to Dr. Ford, he is OSHA's designated witness on the epidemiologic studies of environmental tobacco smoke and lung cancer and environmental tobacco smoke and heart disease. Those issues are critical to the validity of OSHA's rule and Dr. Ford's evidence is critical to those issues.

With respect to the importance of the issues, the quantitative risk assessments included in the rule for environmental tobacco smoke on both heart disease and lung cancer are based upon the epidemiologic studies that Dr. Ford has been asked to testify on. And the reason that his evidence is so important on those issues is that several times during the examination of the OSHA panel they were asked specific questions about the epidemiologic studies of environmental tobacco smoke and heart disease and lung cancer and a frequent response was that they were not prepared to discuss the specifics of those studies but that they had designated a witness from the outside that would be brought in to answer the public's detailed questions about those studies.

And I would like to give you a few examples that illustrate the positions taken by OSHA during their examination.

When Mr. Sirridge asked the OSHA panel the following question, "Was the Helsing study intended to be an investigation of the relationship between exposure to ETS and coronary disease?" Which I would submit is a fairly basic question.

On behalf of OSHA, Mr. Martonik's response was, "I'm not sure that we're prepared to go into a whole lot of detail today regarding all the aspects of that study. We do have a witness that we've hired who is going to testify and who will answer and is extremely familiar with all those factors."

Mr. Martonik went on to add that, "We believe that his viewpoints will provide information regarding the confidence that we have in this study."

Mr. Sirridge later asked OSHA, "What data were gathered in the Helsing study on factors of exposures that might be related to coronary heart disease?"

And again on behalf of OSHA, Mr. Martonik responded in part, "The question that you're asking I already gave a general answer for it. You're getting into details of the study. I've told you already that this was a rough estimate and we're looking forward to interaction with the public on the issue. And we have explained that there is somebody who is going to explain it in more detail, the reasons we think the study was a good study."

And when Mr. Sirridge continued to press OSHA on the details of the Helsing study, Dr. Silverstein explained that, "We have, I believe, 17 additional experts who are going to testify in much greater detail on some points than we here are capable of doing today."

And that type of position was not limited to the Helsing study and heart disease but was also taken with respect to questions on the lung cancer studies.

For instance, when Mr. Grossman was examining OSHA with respect to the lung cancer studies, he asked the following, again a relatively simple question, I would submit, "What weight did you," meaning OSHA, "give to the study by, let's say, Brownson et al. 1987 listed as equivocal?"

And Mr. Martonik's response on behalf of OSHA was, "We're really not prepared to go into detail on the studies."

Mr. Grossman then asked the question, "Could you identify for me the standard epidemiological and statistical criteria that you used to place that Brownson study in the positive category?"

Mr. Martonik's response was, "I just told you that we're not in a position to go into the detail explaining the content of each of those studies."

Again, when Mr. Grossman continued to press OSHA on this issue, Mr. Martonik responded, "We also pointed out that we have some witnesses who are prepared to go into the detail of the studies."

Dr. Silverstein then added, "We are not prepared to give lengthy and comprehensive discussion of all the details of all these studies. Our testimony will be supplemented by the testimony of the experts to follow."

There are other examples in the transcript of OSHA taking this position. I hope the point has already been made.

On this critical issue of whether or not the epidemiologic studies relied upon by OSHA are valid for purposes of quantitative risk assessment, the interested public has been told that if they have detailed questions that they will have an opportunity to ask them later of an outside witness and that witness is Dr. Ford.

I think it's clear that to perform a detailed examination on Dr. Ford regarding these issues a significant amount of time will be required. In this hearing recently full days have been given to the direct examination of Dr. Levois, Dr. Roth and Dr. Jenkins. I don't believe that Dr. Ford will be any less -- in fact, I would submit that Dr. Ford is a more important witness than any of those three, his being OSHA's designated witness.

I also would submit that we're really not asking too much to have an opportunity to examine Dr. Ford thoroughly, given as you pointed out this morning, Your Honor, we're now in the 51st day of these hearings and as I have suggested, Dr. Ford is perhaps the most critical witness that will testify with respect to OSHA's position on why these studies justify the quantitative risk on which the proposed rule is based.

I would also submit that it's not too much to ask for the public to have an opportunity to thoroughly examine OSHA's witness on that issue, especially if the public is to be given an opportunity to provide post-hearing comment in which they can address the positions articulated by Dr. Ford.

I don't see how Dr. Ford can be examined in the manner that I'm suggesting in anything less than a full day, as was devoted to Drs. Levois, Roth and Jenkins.

The problem being that once that full day of examination was completed only then would we turn to Dr. Bayard and I would like to briefly discuss why I think Dr. Bayard is an important witness and why I think that his examination will also encompass at least a full day to a day and a half.

I'm sure I don't need to say much, Your Honor, to convince you that the Environmental Protection Agency's risk assessment on lung cancer has been an important topic in these hearings. If you've heard it mentioned once, you've probably heard it mentioned a thousand times. A number of the outside experts that have appeared and testified here have allocated a large portion of their testimony to examining the conclusions and analysis contained in the Environmental Protection Agency's report.

In addition, and I think perhaps even more importantly, OSHA despite its sometimes protestations to the contrary has in fact relied heavily upon the EPA report and EPA's analysis. And, again, I would like to provide you some examples from the record that demonstrate this point.

As I believe Ms. Ward indicated on Friday, the Federal Register notice itself of the proposed rule states that "The agency believes that available data support proposing regulation of IAQ including exposure to ETS. Further stimulus for this determination was provided by conclusions reached in a report published in December of 1992 by the Environmental Protection Agency addressing hazards associated with exposure to ETS."

And, again, when the OSHA panel was testifying in the early days of this hearing, it was asked a question by Mr. Weinberg with regard to whether the Fontham study, which you've also heard a great deal about, Your Honor, demonstrated the necessary criteria for the establishment of a causal association.

On behalf of OSHA, Mr. Martonik responded in part, "We point to perhaps discussions that were written by EPA in some of their documents as descriptions of the strengths and weaknesses of the various studies."

During Mr. Grossman's examination of the panel, Mr. Grossman asked the panel why it could not explain that the epidemiologic studies had been categorized by OSHA into various slots on their chart.

And Mr. Martonik responded, "I think we told you that the studies are in the record, that EPA in their papers have discussions of these that will give some idea of their content and interpretation which we have generally the same opinion."

When Mr. Grossman attempted to examine OSHA with respect to the Fontham study, Mr. Martonik responded by stating that while OSHA could not answer the specific questions, that there were other descriptions of the study in the record, including the study by the Environmental Protection Agency.

When Mr. Grossman attempted to press OSHA with respect to the Fontham study, Mr. Martonik responded, "I think I answered your question. I said we looked at the EPA study."

I would also note that in the Federal Register it is stated that "OSHA has submitted this proposed standard to the U.S. Environmental Protection Agency which is reviewing it in detail for purposes of submitting detailed comments to the docket."

When you combine OSHA's reliance upon the Environmental Protection Agency's report and analysis, as well as their invitation to the Environmental Protection Agency to perform a detailed analysis of the study and submit that evidence to the docket, I would submit that what you really have is an 18th invited witness, the Environmental Protection Agency. I know that OSHA does not count him as such but I think that it's fair to put Dr. Bayard in that category.

I also want to review quickly the evidence that Dr. Bayard has submitted to the record, which I think further demonstrates the need for a significant amount of time to cross-examine him.

Dr. Bayard has submitted a written statement and eight additional Environmental Protection Agency documents. One of those written statements is the Environmental Protection Agency's risk assessment of environmental tobacco smoke and lung cancer, which alone including its appendices comprises about 450 pages. When you add the other seven documents to that total, his total submission is well in excess of 550 pages.

But it's not just the volume of Dr. Bayard's testimony, it's also the scope. He has submitted evidence on both the EPA report and OSHA's proposed rule. His written testimony alone addresses the hazard identification for environmental tobacco smoke and lung cancer and for environmental tobacco smoke and heart disease. He has also provided testimony on the quantitative risk assessment for both of those end points.

In addition, he addresses exposure to environmental tobacco smoke and the chemistry of both mainstream and environmental tobacco smoke and the implications of that chemical comparison in his opinion for health analysis.

No other witness in this hearing, not even Dr. Glantz, has testified on as broad a range of topics as Dr. Bayard intends to.

And I would also submit to you, Your Honor, that the tobacco industry is not going to be the only party that's interested in examining Dr. Bayard. In fact, his testimony is critical of OSHA's analysis in the proposed rule on a number of counts, including OSHA's treatment of the chemistry, OSHA's review of the toxicology and OSHA's review of the epidemiology.

I would submit that OSHA and other proponents of the rule may have a great number of questions with Dr. Bayard with respect to these criticisms.

R.J. Reynolds, which has submitted a comparable amount of information to the docket as Dr. Bayard, as we all know, is here for three days. And, Your Honor, I would like to suggest that it's really only fair that the interested public be given at least one day to one and a half days to examine Dr. Bayard on a similar volume of information, especially again if the public is to be afforded a meaningful opportunity to provide post-hearing comment on the many issues that OSHA relies upon EPA's analysis for.

Also, I would like to suggest, Your Honor, that this really is not a request that should be burdensome or difficult to accomplish. Dr. Bayard has been in and out of these hearings frequently. In fact, he was here this morning for a while, I'm told. He's shown an ongoing interest in the proceedings. The date for his testimony has been changed at least four times already. I don't know the reasons for that but he seems to have shown some flexibility in the past. He has shown up to examine other witnesses, not only on behalf of the EPA but, as you know, Your Honor, ostensibly on behalf of other participants in the hearing as well.

Again, to place the importance of their testimony, of the testimony of both Drs. Ford and Bayard in context, we're in day 51 of the rulemaking hearing. My understanding is that OSHA expects at least another 10 to 15 days of testimony after the 23rd. When you place the importance of their testimony into context against the amount of time that has been allotted to the examination of other witnesses, largely by OSHA who have far less important information to bring to the hearing, I think the public deserves no less than to have a real opportunity to probe the positions of both Dr. Bayard and Dr. Ford.

Thank you.

JUDGE VITTONE: Thank you, Mr. Furr.

MS. SHERMAN: May I respond?

JUDGE VITTONE: Yes, you may.

MS. SHERMAN: Briefly.

MR. ELY: Your Honor?

JUDGE VITTONE: Hold on a second.

MR. ELY: I have just a minute or two.

JUDGE VITTONE: Okay. I think he's probably going to reaffirm what Mr. Furr had to say.

MR. ELY: Clausen Ely for the Tobacco Institute.

JUDGE VITTONE: Sir, I'm sorry, I've forgotten your name.

MR. ELY: Clausen Ely for the Tobacco Institute.

JUDGE VITTONE: Mr. Ely. I'm sorry.

MR. ELY: I certainly won't take as much time as Jeff did.

I just want to express the Tobacco Institute's strong support for this motion. I would like to note particularly with respect to the EPA report that Dr. Bayard is the only witness that's going to be made available for questioning with respect to that report. I think a good case could be made that as a practical matter the ETS portion of this proposal never would have been proposed if it weren't for the EPA report.

Dr. Bayard has found time to go around the country the last couple of years for many days and discuss the report and answer questions about the report, so I think he ought to find enough time to come to this proceeding and answer questions about the report.

We have a number of questions on behalf of the Tobacco Institute we would like to ask Dr. Bayard. We would be willing to ask those questions any day and a half that he could make available. And since he's here in Washington, it would seem reasonable for him to make that day and a half available.

JUDGE VITTONE: Thank you, Mr. Ely.

Ms. Sherman?

MS. SHERMAN: I think Mr. Andrade wants to say his day in the sun.


MR. ANDRADE: I'll keep it brief.

Tony Andrade for Philip Morris, docket number 51.

We share the concerns that have been expressed by Mr. Furr and by Mr. Ely and agree that these are very important witnesses and that the time that should be allotted for examination of Dr. Bayard and Dr. Ford needs to be substantial, so we strongly support the view and want to join in their request.

JUDGE VITTONE: All right. Thank you, Mr. Andrade.

MR. FURR: Your Honor, just so there is no misunderstanding about what we're asking for, we're not suggesting any specific day for either one of those witnesses, whenever they can be scheduled in a manner that will give the public an ample amount of time to examine them, we will return to do so.

JUDGE VITTONE: Okay. Thank you, sir.

All right. Now, Ms. Sherman.

MS. SHERMAN: First of all, I would like to correct some perhaps mistaken impressions.

Dr. Ford is certainly not the only witness for OSHA who was capable of and did testify about epidemiological studies. I believe that as Mr. Furr pointed out, Dr. Silverstein said that we had 17 additional witnesses. He did not say Dr. Ford this, Dr. Ford that, Dr. Ford the other thing.

The OSHA panel has presented already Dr. Samet, who discussed epidemiological studies; Dr. Glantz, who discussed epidemiological studies; and Dr. Wells, who discussed epidemiological studies.

Dr. Ford is merely one more witness.

However, Dr. Samet spoke quite extensively about epidemiological studies involving lung cancer; Dr. Glantz spoke about epidemiological studies involving heart disease, et cetera.

Now, as to whether or not Dr. Bayard is an OSHA witness, he is not. He represents the Environmental Protection Agency and he is appearing at their request, not at OSHA's request.

Much was made about the fact that there were some witnesses where we only examined them for the whole day. I would note for the record that in the case of Dr. Jenkins, Dr. Bridges also appeared that day and I believe that on the same day as Dr. Levois we had Dr. Ray Witorsch. I would have to check the schedule as to the other two, however, I would note that in some cases if people appeared alone it was because people have dropped off of the schedule.

I very much resent the implication that OSHA is manipulating the schedule and to the extent that Mr. Bayard has been scheduled more than once it because this hearing has gone longer than we expected and many people have been bumped down.

If the rationale about how much material somebody submits were to run true, then I think that R.J. Reynolds would be up here for a week considering the scope and the breadth of their submissions. As everybody notes, they are not going to be up here for a week. They are going to be up here for three days.

R.J. Reynolds submitted three cartons of information and that's just at this stage of the proceeding. They also were very industrious and submitted many other useful pieces of information at earlier stages of the proceeding, as did Mr. Gray Robertson of Healthy Buildings International.

In these proceedings, there is no absolute right to cross-examine everybody who proceeds on everything they've ever said or thought. Our rules talk about asking questions, relevant questions, on crucial issues to help in the development of the record.

At this point in time, based on the amount of presentation that Dr. Ford said he was going to give and the amount of presentation that Dr. Bayard said he was going to give, we thought that this was an appropriate amount of time. We are also on the record as offering to start early that day and I think that's all I have to say about it at this moment.

MR. FURR: Briefly, Your Honor.

I only want to point out that what you didn't hear was why OSHA is resisting rescheduling these two witnesses.

Ms. Sherman attempted to rebut some of the points I made but she did not explain why OSHA won't at least attempt to reschedule these witnesses to give the public a fair opportunity to examine them.

MS. SHERMAN: Mr. Furr, I believe that I have responded to you. The way this motion first came is we are concerned about the number of people that you have scheduled on the 23rd so I took steps to do something about that. Now apparently the next thing I heard you didn't like who I chose to reschedule but the very first instance I heard about this was from Mr. Rupp who pointed out that we had too many people on the 23rd and I took steps to make it a lighter day.

MR. FURR: Your Honor, I don't want this to devolve but I have read the transcript from last Friday and it is very clear --

MS. SHERMAN: Last Friday was not the first time it was raised.

MR. FURR: It was very clear, the concern raised in the motion by Ms. Ward was scheduling Drs. Bayard and Ford for the same day.

JUDGE VITTONE: Okay. But we had discussed it earlier in an off-the-record conversation between Mr. Rupp, Ms. Sherman, myself and a number of the other lawyers. Maybe it was even the day before or earlier that day, I don't recall.

MS. SHERMAN: Or earlier that week.

JUDGE VITTONE: Or earlier that week. I don't recall. But it was some time last week.

OSHA did move Dr. Sidney Greenfield, who was scheduled to testify on January 23rd, next Monday, and added an additional day and put him on on January 24th.

MS. SHERMAN: And in response to concerns about the 20th, we moved the AIHA because it was thought that the 20th was too full a day.

JUDGE VITTONE: Thank you, Mr. Furr.

Let me point out that nobody has been able to do all of the cross-examination that they would have liked to have done in this proceeding.

From the very beginning, I have restricted the examination of various witnesses all through this proceeding. But in addition I think we have been flexible in the sense that -- when I say "we" I mean everybody in this room and that we have started early at times, we have gone late at times, we have restricted the lunch hour at times when it's necessary.

I appreciate your comments, Mr. Furr, with respect to Dr. Ford and Dr. Bayard. However, I note that Dr. Bayard really is an EPA employee who is representing that agency. In response to questions from me, he indicated very clearly he is representing that agency in this proceeding.

In this very real sense, I don't think that Ms. Sherman has any real control over him and I have tried to be somewhat responsive, as much as I could, in this proceeding to various witnesses and their schedules and things of that sort.

JUDGE VITTONE: I am going to go forward at this time with the schedule as it is with Dr. Ford and Dr. Bayard. If on that date, for some reason we don't make any progress with one or the other, I might be willing to reconsider the point. Perhaps Dr. Bayard, like Dr. Smoke, would be willing to come back at some point if we do not have enough time to cross examine him, or have him examined on the 23rd.

I would note that some people have not been able to come back. As I recall, the OSHA staff had some questions for additional witnesses who indicated they were not able to come back for additional questioning. That may or may not occur with Dr. Bayard, I don't know. But I would like to leave the schedule as is. Since Dr. Bayard is not here at the present time, I don't even know what his desires are. He might be perfectly willing to come up here for three days, he might not be able to do that. But I think I'm going to have to hear from him directly before I can just preemptily order that he come back on another day right now.

Quite truthfully, I think your request for a day and a half, in the context of past practices with other witnesses, is on the very high side.

MR. FURR: Your Honor, could I address that very quickly?

JUDGE VITTONE: Just a second.

I don't know of anybody, other than your witnesses and believe me I appreciate that your witnesses are going to be up here for three days. Who has undergone that kind of an examination. I think that one, I do have to hear from Dr. Bayard, and he's not here at the moment. Find out what his druthers are.

As always I will try to be as flexible as possible with everyone in this room, including the witnesses like Dr. Bayard and Dr. Ford. If it's necessary, as I said before, on Monday, the 23rd, I'll be glad to start at 9:00 o'clock, and I will be glad to stay here until the evening hours to get in a much examination as possible.

Now I'll give you a chance to respond, but...

MR. FURR: I just want to make two very brief points. One is, we're not asking fur unlimited cross examination for anything other than, as Your Honor has suggested, the same restrictions on examination that ah ave been applied to all other witnesses. But if one takes one step back from it and realizes that for those other important substantive witnesses like the witnesses that I mentioned earlier, those restrictions resulted in each of them being on the stand from 9:30 until 5:00. That leaves us with two witnesses for that day, each which should in form, be on the stand from 9:30 until 5:00.

Second, what I'm really having the most trouble understanding is why OSHA is resisting even attempting to adjust the schedule. As I understand it, from Your Honor's comments and Mrs. Sherman has not said anything different, OSHA has not even approached Dr. Bayard to see whether he would be able to attend the hearing at some time in what's likely to be the next six weeks that this hearing will run.

JUDGE VITTONE: I think Ms. Sherman did state earlier that Dr. Bayard had indicated to her that he had a scheduling problem. Is that true, Ms. Sherman?

MS. SHERMAN: That is true.

JUDGE VITTONE: She said that earlier in the day.

MR. FURR: For the next six weeks?

JUDGE VITTONE: I don't know, and I don't think we can resolve it here.

As I said, I'm willing to revisit the issue, depending on how Monday goes. And if Dr. Bayard is willing to do this, I'm certainly willing to do it, but I've got to hear from him. As an independent representative of the Environmental Protection Agency, I think we should give him an opportunity to respond.

MR. FURR: How will we go about getting his response, Your Honor?

JUDGE VITTONE: I'll call him.

MR. FURR: Thank you.

JUDGE VITTONE: I will talk to him myself.

If there's nothing else, I would like to recess until tomorrow morning at 9:30.

Thank you very much.

This document's URL is: http://www.tobacco.org/Documents/osha/950117osha.html

Return To: OSHA Hearings page
Go To: Tobacco BBS HomePage / Resources Page / Health Page / Documents Page / Culture Page / Activism Page