OSHA: Proposed Standard For Indoor Air Quality: ETS Hearings, January 5, 1995
OSHA: Proposed Standard For Indoor Air Quality: ETS Hearings, January 5, 1995
UNITED STATES DEPARTMENT OF LABOR
OCCUPATIONAL SAFETY AND HEALTH ADMINISTRATION
PUBLIC HEARING
PROPOSED STANDARD FOR INDOOR AIR QUALITY
Thursday, January 5, 1995
Department of Labor
Washington, D.C.
The above-entitled matter came on for hearing, pursuant to notice, at 8:30 a.m.
BEFORE: HONORABLE JOHN VITTONE
Administrative Law Judge
AGENDA
PAGE
Oak Ridge National Laboratory
Roger A. Jenkins 9689
Questions:
Dr. Samet 9733
Ms. Sherman 9762
Dr. Hammond 9844
Mr. Hopper 9867
Mr. Myers 9904
Mr. Rupp 9935
Ms. Sherman 9957
Dr. James Bridges 9959
Questions:
Ms. Sherman 9983
Mr. Myers 10022
Mr. Weinberg 10039
Mr. O'Neill 10041
EXHIBITS
EXHIBIT NO. IDENTIFIED RECEIVED
195 9761 9761
196 9761 9761
197 9762 9762
198 9762 9762
199 9983 9983
P R O C E E D I N G S
8:42 a.m.
JUDGE VITTONE: We resume our hearings into the proposed rule on indoor air quality.
A couple of minor housekeeping matters before we begin. Mr. Paul Kamenar of the Washington Legal Foundation was scheduled to testify. He was here on December 16th and we asked him to come back today. I had a conversation with him yesterday. He has requested that his testimony that he had previously submitted be received into the record in this proceeding and he said that he would be most willing to respond to any questions that people wanted to submit to him in writing and submit his answers for the record in writing. As7 I understand it yesterday, there were probably very few people, if any, who had any real questions for Mr. Kamenar. I think basically his statement is a policy as opposed to a scientific argument with respect to the rule. So if there is no objection, I will permit Mr. Kamenar, his statement will be received into the record of the proceeding, it already is in the record of the proceeding but if any of the parties have any questions they should share them in writing, submit them to Mr. Kamenar by the close of what I call the first phase of the proceeding where we have the scheduled witnesses as opposed to the people who will be coming at a later date, after January 20th, on an unscheduled basis. So if anybody have any questions for Mr. Kamenar they should prepare those questions by January 20th or the day whenever we close this first phase of the proceeding.
A change also in tomorrow's schedule. Mr. Collett, I received a fax yesterday from Mr. Sterling of Thomas D. Sterling and Associates stating that Mr. Collett, because of an illness in his family, would not be able to testify. And I will just briefly read the letter. It says, "Due to the same family matters that have restricted his travel, my associate, Chris Collett has made the difficult decision to take an extended leave of absence. He will not be available to testify as scheduled January 6th. Because it is essential that our testimony material be integrated, I, too, with withdraw from testifying. I sincerely apologize for any inconvenience our change of plans may cause the hearing proceedings." They were supposed to testify tomorrow so they will not be testifying.
Our first witness this morning is Mr. Jenkins.
First off, let me ask, is there anything else? Any preliminary matters anybody needs to raise?
(No audible response)
JUDGE VITTONE: Mr. Jenkins, would you state your full name, please, and the organization that you're affiliated with, please?
DR. JENKINS: I'm Roger Alan Jenkins. My affiliation is Oak Ridge National Laboratory.
JUDGE VITTONE: I understand you have a presentation with a number of slides. You're going to use that machine there.
DR. JENKINS: That's correct.
JUDGE VITTONE: Okay. Let me get out of your way and you can proceed. You're going to be approximately one hour?
DR. JENKINS: About an hour and five minutes or so, something like that. Unless I get really going fast and then we'll get it done faster.
Is there any water available some place?
Okay. Why don't the slide projector and then I can get going?
Good morning. As I said, my name is Roger Jenkins. I hold a Ph.D. from the University of Wisconsin-Madison in analytical chemistry. I am employed Oak Ridge National Laboratory.
ORNL is a multi-program basic and applied research facility in eastern Tennessee. It's staffed by more than 4500 scientists and supporting staff. The facility itself is owned by the U.S. Department of Energy. My current position is that of Group Leader for Environmental Monitoring in the Chemical and Analytical Sciences Division of ORNL.
In the more than 19 years since I've earned my doctorate, I've been principal investigator for a number of projects for the Departments of Energy, Defense, National Institutes of Health and private and educational concerns.
My professional interests include the chemical characterization of airborne mixtures. That includes process effluents, fugitive emissions, military obscurance and tobacco smoke. I've worked in the general area of characterization of tobacco smoke and methodology for the definition of animal inhalation exposures and the definition of them since 1976.
The studies regarding tobacco smoke with which I've been involved have been supported by the Center for Indoor Air Research, the Council for Tobacco Research, the University of California at Davis, the National Institute on Drug Abuse and the National Cancer Institute.
I am the co-author of a monograph entitled "The Chemistry of Environmental Tobacco Smoke, Composition and Measurement" and I have published both in the general area of symposium proceedings and peer reviewed literature concerning ETS levels in public places.
I am not here today as an advocate for or against this rulemaking but as a scientist that's been conducting a study and gathering some information I think is going to be useful to OSHA.
Along with Dr. Michael R. Guerin I am the co-principal investigator for what is to my knowledge the largest study ever conducted in the United States of personal exposure to environmental tobacco smoke. On the 16th of March last year, I privately briefed OSHA and EPA representatives on the initial status of this study and since that time, the field and laboratory analyses of the samples have been completed. The data interpretation is still ongoing but I think that some of the initial conclusions ought be interesting and sort of bear on some of the findings and some of the thinking that's going into this proposed rulemaking.
Now, it's important to bear in mind as I go through this discussion that much of the data has been assembled over the last six weeks or so because we didn't get all of the laboratory data until relatively recently, about early November, so you may see some minor modifications when you see the final published reports or peer reviewed literature on these things and also because I did these slides myself, there's a couple of errors through them and I'll try to point those out as we go along.
This is slide number two for those of you that are following along. There were copies anyway of the slides outside and also my testimony.
This study has been sponsored by the Center for Indoor Air Research and is a joint effort of the Chemical and Analytical Sciences Division and the Computing Application Division of Oakridge National Laboratory.
Now, what we were trying to do in this study was to determine personal exposure to ETS. We recruited subjects from 16 cities around the United States, about 100 subjects in each of those 16 cities. We asked the participants to wear sampling pumps at their workplace and then at their away from work location and that includes anything that they did away from work, sleeping, eating, dining, shopping, whatever.
We collected both particle and gas phase constituents and we assessed the smoking status through the use of salivary cotinine. We began the field operations in the city of Knoxville, Tennessee in May of 1993 and we finished up 15 cities later in New Orleans, Louisiana in the middle of June of this past year.
The cities were chosen on the basis of a number of factors, the time of year, the likely ambient weather in the location. By the way, in this presentation, there's no difference between the red stars and the green stars, those are just the cities. You can see that they're scattered throughout the four major census regions of the country. We also picked them on the basis of logistics and marketing research support in the particular area where we were doing the study.
We determined a number of constituents. The particle phase ETS markers were collected on the membrane filter at flow rates of approximately 1.7 liters a minute and you can see a listing of the particle phase constituents.
The RSP is a 3.5 micron cutoff. The gas phase constituents were 3-ethenyl pyridine, myosmine and nicotine and we also measured salivary cotinine. The RSP was determined gravimetrically. UVPM and FPM were determined by high performance liquid chromatography. The gas phase constituents were analyzed by gas liquid chromatography and we used radio immune assay to do the analysis of the salivary cotinine.
Now, in any kind of a study design, you need to decide who you're going to include the study and who you're going to exclude in the study. How we did the recruiting and the inclusion of the individuals is that a marketing research firm was contracted in each of the 16 cities to randomly recruit the subjects. The recruiting began a minimum of three weeks before the testing was to take place. Telephone numbers in each metropolitan area were dialed at random.
The factors that we used for the inclusion or exclusion from the study was that the individuals had to be at least 18 years of age, no tobacco use within the last six months and this included prescription use of a nicotine patch or nicotine gum. They had to work at least 35 hours a week outside the home on a regular shift and we requested regular shifts just because of the logistics of conducting this. The pumps had to go out and then come back in on some kind of a schedule that we could orchestrate all this.
We didn't want people, no offense to the attorneys in the group, but we didn't want any attorneys in there because attorneys, my experience is that all are advocates in one form or another. We wanted to eliminate certain people in certain branches of government that might have a stake in the outcome of the study, such as EPA or OSHA people. And we also didn't want individuals that were in smoking-related advocacy groups whether it's the pro-tobacco people or the anti-tobacco people. So there were a number of questions in the screening questionnaire.
The screening questionnaire interview, if it went all the way to completion on the telephone, took about 20 minutes. Each person was either terminated or they were invited to participate based on their responses to the screening questionnaire.
Let me see if there's anything else I want to say about that.
We told the subjects that this was a study related to indoor air quality but no specific reference was made to environmental tobacco smoke. As soon as the survey forms were completed for individuals, they were rechecked to make sure that there in fact weren't any mistakes according to the categorization.
Now, I want to sort of take you through in a couple of slides what happens to the participants so that you get a better feel for what they're doing, what's going on and that sort of thing.
This is slide seven, if you're following along.
Okay. The first thing that happens is the person is contacted in a random telephone survey. They pass the screening questionnaire. Then on the evening of the first day, they arrive at the test coordination site and they are re-screened to make sure that they are giving us the same kind of answers that they did over the telephone. We eliminated several people from the study based on changes in their answers when they got there.
When they get to the coordination site and they all get together, they watch an instructional video that takes about 22 or 23 minutes. They answer sort of what we call a first visit survey concerning their lifestyle and other kind of habits and whatever that they have. They receive their sampling pump systems, they're trained on how to actually turn the sampling pump on and off and they provide a saliva sample.
They go home and on the morning of day two when they arrive at work they turn on their workplace sampling pump that they're wearing around them. They fill out a workplace diary on an hourly basis. At the end of their workday -- they're asked to remain at their desk during the lunch hour if possible.
At the end of the workday, they turn off their workplace sampling pump, they complete this workplace diary. They sort of fill everything out and finish the work part of the survey. Then they put on their away from work sampling pump which is outfitted with a larger battery pack.
They go home via whatever errands they've got to run wearing this sampling pump. And they continue to fill out an away-from-work diary on an hourly basis.
At bedtime, they take off the sampling pump and they put it alongside their bed and the pump continues to sample while they are sleeping. We also asked that they remove the pump while they take a shower or a bath. And also tell us what times that they may have removed the pump for those kinds of activities or other forms of personal activities which they may not want to describe in detail. However, some of them did describe in detail exactly what they were doing when they took the pump off.
Okay. On the morning of day three, after they get to work, they turn off the away-from-work pump, they complete their diary and then after work on day three they go back to the coordination center. They take all the test materials back, they complete another survey, they provide another saliva sample and receive their gratuity.
Clearly, for 1600 subjects this is a huge study and these are the people and the major role players in the study and what their responsibilities are.
This is slide nine, institutional responsibilities. The three organizations that were directly involved in the study were my organization, Oak Ridge National Laboratory. Our responsibility was overall study design and oversight. We provided independent surveillance of the field and laboratory analyses. We also assisted with some of the field operations. It was our responsibility to do all the data integration, interpretation and reporting and quality assurance.
Bellomy Research was the primary marketing research firm. They provided initial questionnaire design to us. They were also responsible through subcontractors with field recruitment of the subjects. They assisted with the field operations and they were responsible for coding all the subject demographic data.
R.J. Reynolds Tobacco Company essentially handled the field sampling operations and logistics, the sample management, and they performed all the laboratory analyses for both the ETS components and salivary cotinine.
Now, it's important to remember that while RJR and Bellomy research maintained the primary responsibility for the initial development of the recruiting and training questionnaires, ORNL had final approval on all of the materials, we reviewed the scripts for the training videos and stuff like that and we were responsible for the revisions in the questionnaires to meet some of our additional needs and concerns.
In order to assure that the data was reported to us objectively, neither Bellomy Research nor R.J. Reynolds had access to each other's data during the course of the study. They all had to feed the data directly to us and it wasn't fed back and forth. So essentially RJR did not know whether the subjects for the samples which they were analyzing in fact were from smoking homes, non-smoking homes or what have you.
Now, because of the sensitive nature of the data and we understood that a lot of critics may question the appropriateness of tobacco company scientists performing laboratory analyses of the collected samples, a number of controls and surveillance practices were instituted to ensure the quality and the integrity of the data. And I'm not going to go through all of these in detail but essentially we provided a number of ETS marker component solutions to them under terms of blind analyses, that they had to essentially get the right answers on those things. We provided them with spiked sampling traps. We also sent them XAD-4 traps which had been exposed to known concentration atmospheres of environmental tobacco smoke that were generated in our chambers back at Oak Ridge and we also sent them smoker/non-smoker saliva samples from ORNL staff members just to make sure that they were in fact getting the right kinds of numbers for those things.
We also provided some surveillance of the field operations. As I said, there was an ORNL staff member at each one of the cities in which the field operations were conducted. As I said, both Bellomy and R.J. Reynolds had to provide ORNL with certain key pieces of what we call raw data before they had had a chance to get out in the field and if there was any kind of monkeyshines going on with the data. Not that we didn't trust them but we had to act as thought we didn't trust them. And this included Xerox copies of the participant diaries and also electronic copies of all the pump flow data and what have you. And this, of course, was all checked after we received it to make sure that that was being handled accurately and appropriately.
We also selected at random some of the individual surveys and manually checked the data that had been coded by Bellomy to make sure that the coding was accurate.
Now, I don't have hard copies of the next slides but I just wanted to show you a few pictures just to kind of give you a feel for things. This is some of the field staff recalibrating the pumps after they came back from the individual operations at the participants' home or workplace or what have. The samples were removed, the pumps are recalibrated and then they're set back out again. These are the blue sports bags that contain both the away-from-work sampling pump and the workplace sampling pump and all the materials and what have you. That's being lined up.
Here's a photograph of -- this is actually what we referred to as the home sampling pump. You can't see it too well but there's an extra battery pack right here and this affords sampling for 18 hours. Here's the actual sampler itself. The filter sample, basically the device to use to cut off the particles at three and a half micros and here's the XAD-4 tube here which collected the gas phase components.
These are individuals filling out one of the many questionnaires. Somebody commented to me that they thought for all this work $100 wasn't enough of a gratuity to do all this because it involved two trips to the sampling site and all of this stuff.
And here's the individual couvets. They contain cotton dental dam that was used to take the saliva samples.
And this slide is the individuals providing us a saliva sample. And I think that was taken in Fresno, California.
This is a participant being shown how to turn the sampling system on and off.
Okay. We have a wealth of demographic information about the participants in this study and the data just includes, just to kind of give you a feel for some of the things that we have, data on age, sex, household income, smoking history, the lifestyles of these individuals, what their dietary habits are, what kinds of pets, heating and cooling systems. Some of the questions in the study were clearly red herring questions to make sure that they didn't catch onto the idea of the fact that this was an environmental tobacco smoke study.
We have a great deal of information on self-reported exposure information, that is, the number of cigarettes, cigars, pipes, whether or not they burn incense in the home, the kind of food cooking smells that they have, whether not there are woodburning fireplaces. And also we have data from the National Weather Service on essentially hour-by-hour climate information while the field operations were being conducted in the cities so we can tell whether or not there were any effects of reduced ventilation in the various private homes during particularly hot or cold weather.
Okay. Now, given this, what I want to go through now is just sort of what are the demographics of our subject data base.
This first slide compares the male/female ratio with that of the U.S. population. This data was taken from the 1993 statistical abstract. That information is from the 1990 census. What all the demographic information that you are about to see has all be age adjusted so those individuals under 18 years of age have been excluded from the demographic comparisons.
One thing you can see right away is that we have more females in our study than the U.S. population as a whole. I'm told by my marketing research colleagues that this is not surprising. First, in general, women are more likely to answer the telephone in homes than are men and also in this kind of a study that requires this much volunteerism, women are more likely to participate in a study that requires this much effort.
Now, it's also important to realize this, that all the comparisons that I'm going to show you now are with the U.S. population as a whole. Remember that in fact what we are looking for, we've asked for non-smokers in our study and we have not yet made the correction for the non-smoking fraction of the United States population over the age of 18. And we know that there are some important demographic differences between smokers and non-smokers and so you want to just keep that in mind when we're making these comparisons.
The next slide, slide 21, is the educational distribution in our study or the age distribution, rather. Because of the requirement that individuals work outside the home, many individuals that are older than 60 years were excluded from the study and so that's why you see some of these differences here, at least that's what I speculated as to why that's the case.
This is the educational level of individuals in our study versus the U.S. population as a whole. Again, this hasn't been corrected for just the non-smoker population in the United States.
Differences in household income, this doesn't show up too well in this brightened room here but essentially we have a slightly higher household income than what the U.S. average is. The U.S. median is about $30,000 a year and in our study the median income was about $40,000 a year and this is partially reflected in this higher educational distribution.
This is the racial distribution of the subjects in our study relative to the U.S. population as a whole. Slightly fewer blacks, slightly more whites, otherwise, relatively close.
This is the breakdown of occupational categories. This is slide 25. And you can see that we have a preponderance of sort of what we might refer to as white collar occupations, technical, sales, administrative, management, professionals and those sorts of things.
We don't have as many service occupations represented in this study and those kinds of individuals that work in factories and that could very well be because the latter kinds of individuals decided not to participate because of safety reasons. That is, they couldn't wear this kind of a sampling pump around their job that wasn't provided to them by their own industrial hygiene operation. We ran into that. Some people had volunteered to participate in the study and then they came to the study site telling us that basically they're sorry but their boss wouldn't let them participate because they would have to wear the sampling pump during the day.
And also the requirement that the subjects hold at least a 35-hour-a-week job outside the home and work on a regular shift also tends to produce a greater fraction of the white collar subjects and that probably also contributes to this general boost in overall income of the participants in our study.
Now, as I indicated previously, we did an awful lot of work associated with field and laboratory surveillance to ensure that the data was being provided to us in an accurate and appropriate manner.
I could show you charts of this but you're going to see enough charts already and I don't want to bore you with long numbers of 92 percent versus 93 percent and things like that. But essentially what the data shows from the quality assurance and surveillance activities, in general, the laboratories were doing a pretty darn accurate job of conducting the analyses.
But I wanted to share with you some of the results of one of the tests and tell you this because I've talked to other people and they seem to be interested in the idea and this is what we refer to as the surreptitious exchange or surreptitious switch.
What we did was as follows: In nine of the 16 cities in the study, an ORNL staff member on site would essentially abscond with one of the blue sports bags during the operations that evening. What they would do -- now, you understand that the field team that were comprised of R.J. Reynolds staff knew that this was going on but they didn't know which one of the bags were going to be essentially stolen that night and there was about 25 or 30 out at any one evening.
Now, when the ORNL staff member got back to his or her hotel room, what they would do is they would turn on the sampling pumps and let them run for an appropriate length of time and then after the sampling pumps had run, say, the workplace sampling pump running eight or nine hours, something like that, and the away-from-work running 16 or 17 hours, then what we would do is turn the pumps off and switch the traps that were in the units with traps that we had loaded either with environmental tobacco smoke atmospheres or known quantities of tobacco smoke markers in the traps.
R.J. Reynolds had provided us with copies essentially of the same kinds of labels that they were using in their study for the individual traps so we relabeled our own traps with their -- and they're the identical traps, they can't tell the difference between one trap and another, and then we put them back into the sample stream a couple of nights later as these individual pumps were coming back in.
So R.J. Reynolds thought in fact that these were real people in the study, in fact, they were just dummy samples that we had put into the sample stream. And here is the data from just those surreptitious switches.
These are the ETS loaded samples. You can see the mean recovery was about 95 percent or so and here for about 89 percent on the spiked samples. To us, this indicated that in general these kinds of analyses were going pretty well.
On the actual ETS loaded samples, when we analyzed duplicate samples of the atmospheres, we hit about the same 93 to 95 percent recovery, so we were getting essentially the same answers as they did.
Now, any study of this kind of magnitude which requires this kind of a voluntary participation, we know is not going to be exactly representative of the United States as a whole and so one potential criticism of any study is that it's going to exclude those individuals in lower socio-economic status groups.
Obviously, there's individuals in this study, we recruited mostly by telephone survey so we exclude all those individuals that don't have telephones in this country, which tends to be lower income groups. And so it would be easy to criticize and say, well, you know, but this is too many high socio-economic status individuals that are participating in this study and that's really going to affect the results by a huge amount.
What we tried to do here is just segregate some of our data using four of the smoke markers as a function looking at the concentration of those smoke markers. These are time averaged concentrations, sort of the average exposure that an individual received over the course of the study as a function of household income, whether or not they were present in smoking homes or non-smoking homes.
And what you see is that there is this general trend towards decreasing exposure with increasing household income. But it's important to see here that while this inverse correlation appears to exist that the range of variation of these things is certainly not a couple of orders of magnitude, that they're in fact about a factor of two to six changes in the average exposures.
Now, we haven't all this particular data for statistically significant differences between these two but the important point here is that if there are differences, these are the median values, if there are differences, the differences are not, say, more than order of magnitude, they're probably a factor of two, three, four, five, something like that those segments of the population which may be reflected in the higher income relative to lower income segments of the population.
Okay. Now I want to talk about some of the results of this study sort of as a whole that relate to the kinds of issues that OSHA is considering.
In this slide, I have presented the overall distribution of nicotine and 3-ethenyl pyridine 24-hour time averaged concentrations for all the subjects in our study which were not excluded because of potential misclassification. I'm going to talk later about misclassification of smokers. We, of course, wanted only non-smokers in the study. We found that a lot of people in the study, it would not have been appropriate to include them in the study. I think the euphemistic term is misclassification.
What we see here, essentially this is -- I want to explain to you because you're going to be seeing time averaged concentration a lot in some of the data that I'm showing. These values that you are seeing are equal to the sum of the concentration time products for the workplace and the away-from-work sampling systems divided by the total time of measurement of the two sampling systems. That's about 24 hours. Essentially this time averaged value is the average level of exposure an individual is exposed to both at work and away from work.
Okay. Now, the other thing I want you to note here in this is that the size -- this is a histogram but the size of the cells are different in the lower end because most of the data occurs in the lower end although there's a few, and you can't see it too well in this slide right here but there are a few at the higher exposure levels.
These are exposures to nicotine and 3-ethenyl pyridine, gas phase markers, and what you see is that most of the time weighted average exposures are considerably less than a microgram per cubic meter over the course of that 24-hour period.
So what you can draw from this slide is that while there are a few subjects that enc