UBC Faculty Research and Publications

Generic guidelines for assessing worker exposure to antisapstain chemicals in the lumber industry Teschke, Kay; Fenske, Richard; van Netten, Chris; Jin, Andrew; Marion, Stephen A. Aug 31, 1992

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata


52383-antisapstain_assessment.pdf [ 956.52kB ]
JSON: 52383-1.0048207.json
JSON-LD: 52383-1.0048207-ld.json
RDF/XML (Pretty): 52383-1.0048207-rdf.xml
RDF/JSON: 52383-1.0048207-rdf.json
Turtle: 52383-1.0048207-turtle.txt
N-Triples: 52383-1.0048207-rdf-ntriples.txt
Original Record: 52383-1.0048207-source.json
Full Text

Full Text

    Generic Guidelines  for Assessing Worker Exposure  to Antisapstain Chemicals  in the Lumber Industry     Kay Teschke*, Richard Fenske†, Chris van Netten*, Andrew Jin*, Stephen A. Marion*  * Department of Health Care and Epidemiology University of British Columbia, Vancouver, B.C., Canada  † Department of Environmental Health University of Washington, Seattle, WA, U.S.A.   Report to the Health Monitoring Sub-Committee of the B.C. Stakeholder Forum on Sapstain Control  August 1992    Guidelines for Antisapstain Exposure Assessment  ii Table of Contents    1. Summary .................................................................................................................................................1  2. Exposure Assessment Guidelines ...........................................................................................................5  2.1 Chemicals to be Monitored...........................................................................................................5  2.2 Measurement Techniques:  Dermal Exposure ..............................................................................5  2.3 Measurement Techniques:  Respiratory Exposure.......................................................................7  2.4 Measurement Techniques:  Laboratory Quality Assurance.........................................................8  2.5 Measurement Techniques:  Field Quality Assurance ...................................................................9  2.6 Representative Sampling:  Where, When, and Who ..................................................................10  2.7 Representative Sampling:  How Many........................................................................................11  2.8 Representative Sampling:  Selecting the Random Sample ........................................................11  2.9 Representative Sampling:  Pre-registration Option ...................................................................12  2.10 Supplementary Data to be Collected at the Time of Measurement .........................................13  3. Background ...........................................................................................................................................17  4. Objectives .............................................................................................................................................19  5. Methods.................................................................................................................................................20  5.1 Site Visits....................................................................................................................................20  5.2 Available Data on Exposure........................................................................................................21  5.3 Literature Review on Measurement Techniques and Sampling Strategies...............................21  5.4 Physical and Biological Properties of Antisapstain Chemicals...................................................21  5.5 Design of the Exposure Assessment Guidelines........................................................................22  6. Results...................................................................................................................................................23   6.1 Site Visits....................................................................................................................................23   6.1.1 Representativeness of the Sample...............................................................................23   6.1.2 Types of Treatment Systems.......................................................................................27 Forklift Diptank ............................................................................................27 Forklift and Elevator Diptank .......................................................................28 Automated Elevator Diptank .......................................................................28 Sorting Chain (Trough) Diptank ..................................................................29 Linear Spraybox ...........................................................................................30 Crosschain (Transverse) Spraybox..............................................................31 Carwash Spraybox........................................................................................31   6.1.3 Location of Treatment Systems in the Production Process .......................................32   6.2 Available Data on Exposure........................................................................................................36   6.2.1 Range of Exposures ......................................................................................................36   6.2.2 Influence of Time of Sampling ......................................................................................36   6.2.3 Influence of Type of Contact .......................................................................................37   6.2.4 Influence of Job or Task ...............................................................................................38   6.2.5 Rationale for stratification by season and by work group ..........................................39  Guidelines for Antisapstain Exposure Assessment  iii Table of Contents     6.3 Review of Measurement Techniques..........................................................................................41   6.3.1 Biological Monitoring.....................................................................................................41   6.3.2 Dermal Exposure Measurement ....................................................................................42 Patch Technique...........................................................................................42 Whole Body Technique ................................................................................42 Fluorescent Tracer Technique .....................................................................43 Rationale for Dermal Exposure Measurement Technique ...........................44   6.3.3 Respiratory Exposure Measurement.............................................................................45   6.3.4 Quality Assurance .........................................................................................................46   6.4 Review of Sampling Strategies...................................................................................................48   6.4.1 Shape of the Exposure Distribution .............................................................................48   6.4.2 Duration of Measurement .............................................................................................48   6.4.3 Who to Sample..............................................................................................................50   6.4.4 When to Sample ............................................................................................................51   6.4.5 Where to Sample...........................................................................................................51   6.4.6 How Many Samples to Take..........................................................................................53   6.5 Physical, Chemical, and Biological Properties of Antisapstain Chemicals .................................55   6.5.1 Skin Absorption.............................................................................................................62 Lipid Removal due to Solvent Vehicles .......................................................63 Hydration......................................................................................................63 Chemical Reactivity......................................................................................64 Skin Metabolism ...........................................................................................64 Molecular Weight..........................................................................................65 Octanol:Water Partition Coefficient ............................................................65   6.5.2 Lung Absorption ...........................................................................................................65 Particle Size..................................................................................................66 Solubility.......................................................................................................66 Concentration...............................................................................................66 Association with Particulate Matter ............................................................67 Exposure Frequency.....................................................................................67 Respiration Rate ...........................................................................................67   6.5.3 Additives and Altered Formulation Products...............................................................67   6.5.4 Medical Surveillance ......................................................................................................68  7. Acknowledgments.................................................................................................................................69  8. References ............................................................................................................................................69   Guidelines for Antisapstain Exposure Assessment  iv List of Tables and Figures     Table 1: Overview of analyses required by each study design - Respiratory exposures ................15  Table 2: Overview of analyses required by each study design - Dermal exposures........................16  Table 3: Sampling garments for specific body regions and change intervals ...................................6  Table 4: Characteristics of selected site-visit locations in comparison to .....................................25  characteristics of all mill and wharf locations where antisapstain  treatment is currently done in British Columbia  Table 5: Characteristics of the sites visited .....................................................................................26  Table 6: Antisapstain agents manufactured by the participating companies .................................56  Table 7: Some physical parameters of antisapstain agents, and reported ....................................58  analytical methods  Table 8: Some biological parameters of antisapstain agents...........................................................60  Table 9: Physical and chemical conditions and properties that affect skin ....................................62  and lung absorption   Figure 1: Map showing locations of sawmill using antisapstain agents in BC ...................................20  Figure 2: Observed locations of antisapstain treatment in sawmills.................................................29  Figure 3: Observed locations of antisapstain treatment in planer and ............................................30  remanufacturing mills  Figure 4: Observed locations of antisapstain treatment in shipping terminals ................................31  Figure 5: Example of the distribution of chlorophenate exposures in a B.C. sawmill ......................44    Guidelines for Antisapstain Exposure Assessment  1 1.  Summary   In order to monitor the introduction of new fungicides in the lumber industry, the B.C. Stakeholder Forum on Sapstain Control was formed with interested parties from industry, labour unions, environmental groups, and government agencies.  This committee contracted a Scientific Advisory Panel, made up of university-based researchers, to develop generic exposure assessment guidelines for antisapstain fungicides used or proposed for use in the British Columbia (B.C.) lumber industry.   The project was funded by 17 manufacturers or distributors of antisapstain products.  The objective of the generic exposure assessment guidelines is to provide exposure data which can be used in risk assessments by government agencies or the lumber industry itself, and which will address the unique characteristics of the lumber treatment process not found when pesticides are applied in other settings such as agriculture.   This summary outlines the methods and results of the research used to develop the exposure assessment guidelines.  Several different types of investigation were used to gather pertinent information. To determine the conditions of fungicide use in the British Columbia lumber industry, a list of user mills and shipping facilities in British Columbia was developed, and locations were randomly selected for on-site observations.  To ensure representative and biologically appropriate sampling, the existing knowledge on patterns of exposure to antisapstain treatment chemicals used in the lumber industry was reviewed, as was the scientific and government literature on monitoring techniques and sampling strategies.  Finally, to define the range of physical, chemical, and biological properties of the antisapstain formulations proposed for use in the industry, each participating manufacturer was contacted to provide relevant information.   Fourteen of 67 sites currently using antisapstain chemicals in B.C. were visited to observe the methods of use and opportunities for exposure in the lumber industry.  Three of the sites were shipping terminals and the remainder were sawmills and/or planermills.  They included facilities from 11 different companies, including locations represented by all 4 unions in the B.C. Stakeholder Forum, as well as non- union plants.  The 14 sites visited used 5 different antisapstain treatment chemicals and had 29 separate chemical treatment systems including some of each of the following types:  forklift diptank, forklift and elevator diptank, automated elevator diptank, carwash spraybox, sorting chain diptank, linear spraybox, and crosschain spraybox.            Forklift diptanks and forklift with elevator diptanks were the only types which required the continuous presence of workers to control the chemical application.  In terms of labour required for maintenance and cleaning of the treatment system, crosschain sprayboxes appeared to require the most, followed by linear sprayboxes, and then all other systems.  The amount of free fluid remaining on the  Guidelines for Antisapstain Exposure Assessment  2 surface of lumber after treatment also showed patterns according to the seven system categories; systems treating bundled lumber and sorting chain diptanks produced the most dripping and splashing of chemical in downstream areas.  The tasks required to operate and maintain each type of system showed patterns of similarity between sites, but there was remarkable diversity in the way such tasks were assigned to workers' job categories and in where the treatment system was located in the production flow. Among sites with the same type of production, there was much variability in the routes and frequency with which freshly treated lumber was diverted back into the mill for further processing.   Only one generalization could be made, that systems treating bundled lumber tended to be located closer to the end of the production line than systems treating individual pieces.  This reduced the number of workers who handled wet lumber, and who were needed to maintain contaminated equipment and work areas.   Data from existing antisapstain exposure assessment studies in the lumber industry, mainly measurements of chlorophenates in urine, showed several trends.  The range of exposures observed within mills often spanned several orders of magnitude, with the lowest levels similar to those found in populations not occupationally exposed.  Levels measured in the summer were higher than those measured in the fall and winter.  Graders, lumber pullers, and other individuals who worked continuously in the vicinity and downstream of treatment equipment had the highest exposures, especially if they had opportunities for skin contact with wet lumber.  Maintenance workers who repair antisapstain supply systems had elevated exposures compared to background levels, however, their exposures were not as high as continuously exposed workers.  Certain groups of workers had low probabilities of exposure, including those in buildings or areas where antisapstain treatment was not done, and those working with treated lumber after it had dried, though on occasion even these individuals demonstrated high levels of exposure. Little evidence was available to compare exposures between different types of mills or application methods.   The physical, chemical, and biologic properties of the antisapstain agents used or proposed for use in lumber treatment vary between fungicides.  For example, some of the compounds are highly acidic, some very basic, and others are near neutral pH;  and some compounds are completely water soluble whereas others are insoluble.  Most of the compounds currently proposed for use in the industry appear to have very low vapour pressures, however there are some which are quite volatile.  Many of the compounds have shown mild irritant effects, suggesting that skin and airborne exposure monitoring may be required even if biological monitoring techniques are developed.  For some of the parameters of interest, the data received from the manufacturers was incomplete.  The differences in properties of the antisapstain agents and the extent of unknown data both suggest that generic exposure assessment guidelines should encompass all plausible means of exposure and routes of entry into the body.   Guidelines for Antisapstain Exposure Assessment  3  Biological monitoring, as was done for chlorophenates, represents a highly desirable approach to antisapstain exposure assessment, but may not be feasible for the new antisapstain agents due to insufficient knowledge regarding absorption, distribution, metabolism, and excretion in humans. Development of analytical methods and human pharmacokinetic databases for these compounds is strongly encouraged.  The whole body technique is at present the most practical method for assessing the wide variety of dermal exposure patterns likely to be encountered during antisapstain use.  Fluorescent tracer screening may help to characterize dermal exposure patterns and identify less obvious exposure situations. The patch technique is impractical due to its assumption of uniform exposure over specific body regions. Respiratory exposure measurements should be conducted in all worker exposure assessments, and should include personal monitoring for vapours and inspirable aerosols.  Standard quality assurance techniques are available for dermal and respiratory exposure measurements.   Current government guidelines for pesticide exposure assessment, such as those specified by the Environmental Protection Agency in the United States, incompletely address the issue of how sampling should be done to ensure measurements are representative, but there is a growing body of literature on this issue in the occupational hygiene field.  In order to determine who to sample, most investigators have suggested stratifying the work force into groups with similar exposure potentials, and randomly sampling within these groups.  If exposures are expected to vary over time, as has been shown in the lumber industry, sampling shifts should also be randomly selected from the period of interest.  The number of samples should be large enough to allow estimation of the mean and variance of the exposure distribution with precision;  10 samples has been suggested as a point of diminishing returns for estimating means, but 30 samples or more may be required to adequately assess the variance.  Ideally, the duration of sampling should be related to the biological half-life of the chemical within the body, so that chronic hazards would be measured for long durations, and acute hazards for short durations.  For compounds with both types of outcomes, it is possible to take long-duration (e.g., full-shift) samples and use the mean to estimate the maximum frequency with which a peak exposure could be exceeded.   Based on the above evidence, we have decided on exposure guidelines that include 1. both dermal and airborne exposure measurements to ensure assessment of the major routes of entry, 2. standard quality assurance methods to determine extraction efficiency, potential losses in the field and during storage, and sample contamination during collection and processing, and 3. a plan for randomly sampling locations, workers, and times to achieve representative sampling.   The following report begins with the generic exposure assessment guidelines developed by the Scientific Advisory Panel.  It then details the background to this work, specific aims of the exposure  Guidelines for Antisapstain Exposure Assessment  4 assessment project, and the methods of investigation used to gather information.  The results of these investigations, presented as Section 6 the end of the report, form the rationale for the guidelines.      Guidelines for Antisapstain Exposure Assessment  5 2.  Exposure Assessment Guidelines   The following are the generic exposure assessment guidelines we propose.  They are based on conclusions reached as a result of on-site observations, data about anti-sapstain agents and exposures, and literature on monitoring techniques and sampling strategies, as summarized above.  For additional details explaining the rationale for the guidelines, please refer to the remainder of this report which includes the background, objectives, methods, and results of these investigations.   Tables 1 and 2, included at the end of this section on pages 15 and 16, provide an overview of the analyses required for an exposure assessment study conducted in accordance with these guidelines, showing each of the three study designs described below.  Table 1 outlines the analyses proposed for respiratory exposure assessment and Table 2 outlines the analyses proposed for dermal exposure assessment.  2.1  Chemicals to be Monitored  2.1.1 The active ingredients of the antisapstain product should be monitored according to the procedures outlined below.  If the product contains more than one active ingredient, exposure to all active ingredients should be monitored. 2.1.2 Components of the formulated product present at a concentration of 1% or more, which are not designated as active ingredients, but which are known to cause adverse health effects, may require monitoring, as determined by regulatory agency evaluation.  2.2  Measurement Techniques:  Dermal Exposure  2.2.1 Standard sampling garments should be employed for use in worker exposure studies.  Ideally, a manufacturer should be identified to construct all sampling garments from the same fabric.  Fabric suitable for such sampling garments would be 1) 100% cotton,  2) at least 200 g/m2 in weight, and 3) not treated with finishes which promote water repellence.  Sampling garments should be pre-extracted, dried, and their weights recorded (g/m2 and g/garment) prior to commencement of the study. 2.2.2 Sampling garments should be worn beneath protective clothing and normal work clothing, as close to the skin as is practical, e.g., sampling gloves beneath protective or chemical resistant gloves, sampling shirt beneath workshirt and/or coveralls, sampling pants beneath workpants.  Clothing  Guidelines for Antisapstain Exposure Assessment  6 worn normally by workers should not otherwise be altered.  Sampling garments should fit snugly against the skin (not baggy and not tightly stretched), and should cover the entire body region. 2.2.3 Measurements should be taken over the entire work shift, including clean-up of equipment and materials at the end of the shift.  Sampling garments for specific body regions and intervals for changing sampling garments are listed in Table 3 below.  Sampling garments should be changed more frequently if there is evidence that they are becoming saturated or if the compound under study is volatile.  Sampling times should be recorded.   Table 3:  Sampling garments for specific body regions and change intervals   Type of  Sampling  Number of Changes Approximate  Garment Body Region (times per shift) Period (hrs)*    Gloves hands 4† (at breaks/lunch) 2  Shirt upper torso/arms 2 (at lunch) 4  Hood head 2 (at lunch) 4  Socks feet 1 8  Pants lower torso/legs 1 8   * assumes 8-hr workshift  † if hands are washed on other occasions than breaks and lunch, sampling gloves should be  changed on these occasions as well  2.2.4 Sampling garments should be removed from the worker at the end of the designated sample collection period by technical staff wearing surgical gloves.  The surgical gloves should be changed for each worker sampled.  Each sampled worker's set of sampling garments should be placed in an appropriate container for transport to the laboratory. 2.2.5 Each worker's set of garments may be analyzed combined, with the following exceptions: 1) Gloves must always be analyzed separately from the remainder of the clothing. 2) A minimum of fifty sets of sampling garments should be analyzed in sections by body region (as designated in Table 3 above).  In addition, sleeves should be sectioned from the shirt before removal, for separate analysis. The fifty workers whose garments are to be analyzed in sections should be selected at random, 10 workers from each of the 5 worker strata defined in section 2.6.3 of these guidelines.  Each type of sampling garment which will be analyzed separately should be placed in a separate container for transport to the laboratory.  Shirt sleeves which will be analyzed separately should be placed in a container separate from the shirt.  Multiples of sampling garments from the same body region for the same worker can be put in one container for pooled analysis (e.g., all 8 gloves can be put in the same container).  Surgical gloves used for removing the garments should be changed  Guidelines for Antisapstain Exposure Assessment  7 for each type of sampling garment, and for shirt sleeves, whenever these items will be analyzed separately. 2.2.6 The specific analytical procedure employed will depend on the compound under study.  A reasonable analytical goal is detection of at least one microgram per sample, with a coefficient of variation of less than 10%. 2.2.7 Appropriate quality assurance samples should be collected during field sampling, as detailed in section 2.5. 2.2.8 Data should be reported as mass of active ingredient per sample, or, for sectioned samples, in terms of mass of active ingredient per body region, (i.e., head, upper torso, arms, hands, feet, and lower torso and legs ).  Each sample should be listed with a unique identification number, sampling times (start and end), sampling period (length of the measurement period), and mass of active ingredient detected.  2.3  Measurement Techniques:  Respiratory Exposure  2.3.1 Personal air samples for vapours and inspirable aerosols (i.e., all aerosols capable of entering the respiratory system; see definition on page 45) should be collected in the worker's breathing zone. Sampling should be conducted throughout the workshift, including clean-up of equipment and materials at the end of the shift. 2.3.2 The air sampling pump should be calibrated with the sampling device and collection medium in line prior to and following the sampling period.  The mean flow rate should be employed to determine total sample volume. 2.3.3 The sample collection medium selected will be dependent on the physical and chemical nature of the compound under study (e.g., volatility, solubility, physical state).  The appropriateness of the collection medium must be demonstrated as part of the quality assurance procedures.  It also must be demonstrated, prior to field sampling, that the collection medium can retain the compound under study throughout the measurement period (i.e., no breakthrough or other losses) across the range of expected exposures.  Alternatively, the collection medium may be replaced periodically during the measurement period.  (Note that, if separate collection media are required to capture aerosols and vapours, they may be analyzed together, provided that the requirements of the quality assurance procedures are met.) 2.3.4 The collection medium should be removed from the sampling device at the end of the workshift and placed in storage for transport to the laboratory. 2.3.5 The specific analytical procedure employed will depend on the compound under study.  A reasonable analytical goal is detection of at least one microgram per sample, with a coefficient of variation less than 10%.  Guidelines for Antisapstain Exposure Assessment  8 2.3.6 Appropriate quality assurance samples should be collected during field sampling, as detailed in section 2.5. 2.3.7 Data should be reported as micrograms of active ingredient per cubic meter of sampled air.  A table should be prepared wherein each sample is listed with a unique identification number, mean pump flow rate, sampling times (start and end), sampling period (length of measurement period), mass of active ingredient detected, total air volume sampled, and calculated air concentration.  2.4  Measurement Techniques:  Laboratory Quality Assurance  2.4.1 An extraction efficiency study should be conducted for each sampling medium to be used in the exposure study.  High efficiency and precision (> 90% ± 10%) should be demonstrated in the laboratory prior to commencement of field sampling.  Three spiking levels should be selected to represent the range of exposure values expected during the study.  Seven replicate samples should be produced for each spiking level.  For dermal sampling garments, extraction efficiency studies should be conducted on each type of sampling garment using the same number of sampling garments as are likely to be combined for analyses of field samples (e.g., 8 gloves;  and 2 socks, 2 shirts, 1 pair of pants and 2 hoods). 2.4.2 A storage stability study should be conducted prior to commencement of field sampling to indicate that samples can be maintained without significant losses over time.  Sets of each sampling medium should be spiked at three levels representing the range of exposure values expected during the exposure study.  The number of replicates per spiking level will depend on the length of the storage study, and should allow triplicate samples at each spiking level to be removed from storage at specified time intervals (e.g., weekly) for analysis.  The final set of sampling media should be stored for the maximum period of time anticipated for field sample storage.  If losses during storage exceed 5%, the storage procedures should be revised and the study repeated.  If the goal of less than 5% loss cannot be attained, then sample decay may be determined quantitatively as a function of storage time, and appropriate correction factors derived from these data.  For dermal sampling garments, storage stability studies should be conducted for each type of garment (e.g., gloves, socks), unless all the dermal sampling garments are made of the same fabric.  In this case, the fabric is considered the sampling medium, and storage stability studies for each type of garment are not required. 2.4.3 The formulated material that will be applied in the workplace during the exposure study, rather than an analytical standard, should be used to spike sampling media in both the extraction efficiency and storage stability studies.  Guidelines for Antisapstain Exposure Assessment  9 2.4.4 If respiratory sampling media for aerosols and vapours will be pooled for laboratory analysis of field samples, extraction efficiency and storage stability studies should also be done on the pooled media. 2.4.5 Blank sampling media should be handled in the laboratory during sample preparation, and then extracted and analyzed in a manner identical to the other samples.  A minimum of 1 laboratory blank should be analyzed for every 20 laboratory quality assurance samples.  2.5  Measurement Techniques:  Field Quality Assurance  2.5.1 Two field loss studies should be conducted prior to commencement of field sampling:  one to evaluate transport, handling, and storage losses; and the other to evaluate these losses, and in addition, the effect of the environmental conditions of the sampling period.  Two sets of spiked samples of each sampling medium should be prepared in the laboratory following procedures detailed in section 2.4.1 (21 samples per sampling medium per set), and taken to a field site. Environmental conditions at the field site should approximate worst-case conditions anticipated for the field study in regard to sample loss (e.g., elevated temperatures). 2.5.2 The first set of spiked samples (transport loss study) should be handled in a manner identical to field samples (e.g., air sampling cassettes should be taken out of storage containers, opened, closed, and returned to storage containers; garments should be taken out of storage bags and returned to the bags), and then transported to the laboratory for extraction and analysis.  This set will evaluate the effect of handling, temporary storage, and transport on sample loss.  If losses are greater than 10%, handling and transport procedures should be revised and the field loss study repeated. 2.5.3 The second set of spiked samples (environmental loss study) should be exposed to environmental conditions (i.e., at the worksite, but outside of the area where antisapstain chemicals are used) for the length of a typical workshift, and then transported to the laboratory for extraction and analysis.  This set will evaluate both the effect of environmental conditions over the sampling period (workshift) and the effect of handling, temporary storage, and transport on sample loss.  If losses from this set exceed those of the transport loss study by 20%, further studies may be required to characterize the environmental conditions responsible for losses during the sampling period. 2.5.4 During subsequent field sampling, two field spike samples per workshift for each sampling medium should be taken to the worksite.  Both samples should be spiked at the same spiking level, selected randomly from the three spiking levels used in the laboratory quality assurance studies described in section 2.4.1.  The field spike samples should be exposed to environmental conditions  Guidelines for Antisapstain Exposure Assessment  10 (i.e., at the worksite, but outside of the area where antisapstain chemicals are used) for the sampling period (workshift), and then handled in a manner identical to field samples. 2.5.5 One field blank for each sampling medium should be taken to the worksite for every 10 samples, selected randomly.  The field blank sample should be exposed to environmental conditions (i.e., at the worksite, but outside of the area where antisapstain chemicals are used) for the sampling period (workshift), and then handled in a manner identical to field samples. 2.5.6 If all dermal sampling garments are made of identical fabric, then the fabric is considered the sampling medium, and field loss studies, field spikes, and field blanks for each type of garment are not required. 2.5.7 If respiratory sampling media for aerosols and vapours will be pooled for laboratory analysis of samples, field loss studies, field spikes, and field blanks should also be analyzed with the media pooled. 2.5.8 Blank sampling media should be handled in the laboratory during sample preparation, and then extracted and analyzed in a manner identical to the other samples.  A minimum of 1 laboratory blank should be analyzed for every 20 field samples, including field quality assurance samples.  2.6  Representative Sampling:  Where, When, and Who  2.6.1 In order to ensure representativeness, all work sites where the antisapstain agent of interest is being used should be considered eligible for sampling. 2.6.2 Exposure measurements should be made after the antisapstain agent of interest has been in use for a minimum of 3 months at the sites included in the study.  (To allow application of new end use products during this initial period and throughout the subsequent exposure study period, Agriculture Canada's Pesticides Directorate is prepared to consider granting registration for a limited time period.) 2.6.3 Workers employed in the work sites of interest should be stratified into the following groups according to the jobs they perform, then selected randomly from each group: - graders or lumber pullers who handle wet wood; - individuals who operate elevator dip tanks, or drive forklifts or carrier trucks to dip lumber; - others who handle wet wood; - individuals who operate the antisapstain solution supply system or perform maintenance work on any part of the treatment system or machinery downstream of treatment (including vehicles used in the treatment process); and - workers who handle treated wood after it is dry.  If a worker appears to fit into more than one of these strata, he should be included in the stratum highest on this list which best describes his usual work tasks.  Guidelines for Antisapstain Exposure Assessment  11  These strata were established using existing antisapstain measurement data and observations of the treatment facilities.  They are expected to include workers with similar exposure potentials, and should include those workers with the highest exposures.  The strata should be re-evaluated after each exposure study has been completed.  2.7  Representative Sampling:  How Many  2.7.1 In order to achieve a reliable estimate of both the mean and variance of the exposures, at least 30 workers should be selected from each stratum, as defined in section 2.6.3 above.  Exposures of each selected worker should be measured on two occasions to allow an estimate of within-worker (i.e., shift-to-shift) variability.  This would give a total of 60 measured shifts per stratum.  Data analysis recommendations for this approach are described in Appendix A as Method A. 2.7.2 An alternative approach would be to measure each worker's exposure only once for a total of 30 sampled shifts per stratum.  Data analysis recommendations for this approach are described in Appendix A as Method B.  Although this approach will reduce the cost of the exposure measurement study, it is likely to result in higher estimates of exposure at the 95th percentile of each worker group because a conservative assumption about the between-worker (i.e., person-to- person) component of variance would have to be used. 2.7.3 Where there are fewer than 30 workers in a stratum, 30 worker-shifts should be selected from the pool of workers in the stratum.  If Method A (Appendix A) is to be used for the data analysis, each worker randomly selected for measurement would need to be measured twice, as described in section 2.7.1.  2.8  Representative Sampling:  Selecting the Random Sample  2.8.1 To identify the pool of workers of interest, the managers of each work site using the antisapstain agent in question should first be asked to tally the total number of workers they employ in each of the five worker strata identified in section 2.6.3 above. 2.8.2 Thirty workers (at this point unnamed) should then be selected randomly from each stratum from the pool of all mills which have workers in that stratum. 2.8.3 Each work site for which one or more workers was selected in step 2.8.2 above, should then be approached to provide lists of workers in the strata selected.  From each list, the previously determined number of workers should be selected at random. 2.8.4 In order to ensure that both cold and warm weather are included in the exposure measurement study, 15 of the workers in each stratum should have one measurement shift randomly selected from the period April to September inclusive, and 15 should have one measurement shift randomly  Guidelines for Antisapstain Exposure Assessment  12 selected from the period October to May inclusive (this applies to both Method A and Method B, as described in sections 2.7.1 and 2.7.2 above). 2.8.5 If two shifts are to be measured for each selected worker (i.e., Method A only, as described in section 2.7.1), the additional measurement shift should be randomly selected from all shifts over the one-year study period. 2.8.6 If, at the time of a worker's initial measurement shift, the selected worker is working in a job in a different stratum, or refuses to participate in the study, an alternate worker from the selected stratum should be chosen at random for measurement.  A record should be kept of this change and the reason for it. 2.8.7 If, at the time of the second measurement shift, the selected worker is working in a new job in the same or a different stratum, the worker should be monitored in this new job, and data about the new job should be recorded. 2.8.8 If, at the time of any measurement shift, the selected worker is absent, this should be recorded, and no replacement worker should be selected.  2.9  Representative Sampling:  Pre-Registration Option   An additional consideration is whether sampling should be done prior to product registration or after it has been in use for a period of time.  It is unlikely that measurements taken prior to registration under test conditions will be representative of those which will occur when the product is registered and in normal use.  Arguments can be made that in test conditions the situation is under tight control so that exposures are lower, or that because of inexperience with the product, exposures are higher.  Therefore, the only way to properly assess exposures would be to take measurements after a period of normal use, as described above.   To allow a preliminary determination of the likely character and magnitude of exposure, a pre- registration study to estimate exposures to unregistered compounds could be conducted under federal research permit.  Federal regulatory agencies have indicated that a full toxicological data package would have to be submitted to Health and Welfare Canada and be under review prior to the initiation of such a pre-registration study.  To conduct a pre-registration exposure measurement study, we recommend the following procedures.  2.9.1 A random sample of at least 10 workers should be selected from each of the five strata of workers defined in 2.6.3 above.  Where there are fewer than 10 workers in a stratum, at least 10 worker- shifts should be selected from the pool of worker-shifts in the stratum during the pre-registration exposure study period.  Guidelines for Antisapstain Exposure Assessment  13 2.9.2 This small number of measurements per group should allow a reasonable estimate of the mean but not the variance of exposures.  We therefore recommend that the variance not be calculated from the data;  instead a predetermined conservative estimate of the variance should be selected for calculations of such measures as the proportion of workers in each group exceeding a given exposure concentration, exposure levels at the 95th percentile of the worker group, and confidence intervals around the mean.  As described in the data analysis recommendations for Method C in Appendix A, we recommend that a logarithmic variance of 3.21, corresponding to a geometric standard deviation of 6, be used.  This estimate should provide a safety factor to account for variability not observed during the limited setting of an efficacy trial.  The variance estimate may be re-evaluated after full exposure studies and pre-registration studies have both been conducted for similar or identical antisapstain formulations. 2.9.3 On each work shift selected for sampling, 15 field spikes should be prepared for each sampling medium, i.e., 3 spiking levels, with 5 replicates for each spiking level.  The field spikes should be exposed to environmental conditions (i.e., at the worksite, but outside of the area where antisapstain chemicals are used) for the sampling period (workshift), and then handled in a manner identical to field samples. 2.9.4 All other monitoring issues, including other quality assurance protocols, should be handled in exactly the same way as a full-scale exposure assessment described in these guidelines.  2.10 Supplementary Data to be Collected at the Time of Measurement   In order to ensure that the exposure assessment programs will provide data which can improve decision-making for future sampling and for recommendation of control measures, it is important that supplementary information be collected at the time of measurement to help ascertain the determinants of exposure.  Such information should include the following: - day of week, date, and time of measurement; - temperature, relative humidity, type of lighting (incandescent, fluorescent, halogen, sunlight, etc.), and location of each sampled worker, each recorded three times daily:  at the beginning, middle, and end of the shift; - temperature, relative humidity, type of lighting (incandescent, fluorescent, halogen, sunlight, etc.), and location of the field spike samples, each recorded three times daily:  at the beginning, middle, and end of the shift; - job title of each sampled worker (usual job title and job during the selected shift, if they differ); - tasks performed by each sampled worker during the measurement period and their durations;  Guidelines for Antisapstain Exposure Assessment  14 - description of the clothing and protective equipment worn by each sampled worker and the duration of use; - description of the types of contact each sampled worker has with the antisapstain agent or treated lumber. - composition of the antisapstain treatment mixture, both as it is received by the work site and as it is applied to the lumber (including information about additives such as defoamers, colorants, etc.); - description of the treatment process (including temperature of application); - process flow diagram showing the treatment process and the locations of the sampled employees (if a sampled individual changes locations throughout the selected shift, this should be indicated;  there should also be an indication of whether the locations of interest are indoors or outdoors); - a description of any control measures in place (e.g., ventilation, demisters, self-cleaning nozzles, wood drying areas, spill containment); and - face velocities for local exhaust ventilation systems used in the mixing or treatment areas, and dimensions of the face of the exhaust hood; and - weather conditions (extent of cloud cover, rainfall, wind velocity) during the selected shift (these data may be obtained from the local Environment Canada weather station).  Guidelines for Antisapstain Exposure Assessment  15  Guidelines for Antisapstain Exposure Assessment  16  Guidelines for Antisapstain Exposure Assessment  17 3.  Background   The British Columbia coastal lumber industry has used antisapstain agents to prevent fungal growth on export lumber for 50 years.  The formulations used during most of this period contained chlorophenols, but in the late 1970s and early 1980s, studies reported that these chemicals were contaminated with dioxins, and that they might be associated with certain human cancers, including soft- tissue sarcoma, non-Hodgkin's lymphoma, and nasal cancers (1-6).  In response to these concerns, the sawmill industry implemented various control strategies and began to investigate alternative antisapstain chemicals.  In the late 1980s, importing nations, including Japan and Germany, suggested that they would prohibit importation of wood treated with this class of chemicals by the end of the decade.  The B.C. lumber industry was therefore required to find substitute fungicides that were both effective at preventing fungal growth on the lumber surfaces, and able to meet the safety interests of workers, environmentalists, and consumers.   In order to monitor the introduction of new fungicides, a committee of interested members of industry, labour unions, environmental groups, and government agencies was formed.  The B.C. Stakeholder Forum on Sapstain Control includes representatives from the following organizations:  the B.C. Ministry of Environment, Lands, and Parks;  the B.C. Ministry of Forests;  the Canadian Paperworkers Union; the Council of Forest Industries of B.C.;  Earthcare;  the International Longshoremen's and Warehousemen's Union;  IWA - Canada;  the Pulp, Paper and Woodworkers of Canada;  the Sawmill Industry of British Columbia;  the Westcoast Environmental Law Association;  and the Wharf Operators of British Columbia. The Forum has taken the initiative to oversee occupational and environmental safety issues for these new chemicals, and may require manufacturers to follow certain procedures either prior to or during the introduction of their products into the industry.   In order for a pesticide to be registered for use in Canada, the manufacturer must follow the Registration Guidelines published by Agriculture Canada (7).  These include requirements for testing of the product's toxicity and metabolism in animals, its environmental fate, and the potential for human exposure. Health and Welfare Canada uses human exposure data in conjunction with animal toxicity data to qualitatively describe the potential risk the pesticide might pose to humans, and to advise Agriculture Canada about the health aspects of the chemical being proposed for registration.  The Canadian registration process has no formal criteria for the performance of exposure studies, but applicants are referred to documents published by the Environmental Protection Agency in the United States:  Subdivision U, Applicator Exposure Monitoring; and Subdivision K, Exposure Reentry Protection (8,9).  These documents were written with agricultural pesticide application in mind, and members of the B.C.  Guidelines for Antisapstain Exposure Assessment  18 Stakeholder Forum on Sapstain Control agreed that sawmill application of antisapstain fungicides differs substantially.  The lumber industry uses methods of application not used in agriculture, the process is continuous rather than sporadic, and there are more workers at potential risk of exposure.  Also in contrast to agriculture, the lumber industry is not treating a product which will ultimately be ingested by humans. Members of the Forum therefore decided that exposure monitoring guidelines should be developed specifically for antisapstain applications in B.C.  These guidelines would then be followed by any manufacturer wishing to have its pesticide used by the industry.   The Stakeholder Forum contracted an independent group of university-based scientists, designated the Scientific Advisory Panel, to design the exposure monitoring guidelines.  Its members include Dr. Richard Fenske, an industrial hygienist whose research focuses on pesticide exposure monitoring, Dr. Chris van Netten, a toxicologist, and Kay Teschke, also an occupational hygienist.  Dr. Andrew Jin, an occupational physician, carried out the basic data collection for the Panel.  Dr. Stephen Marion, a community health physician with a special interest in biostatistics, provided advice on the statistical issues.   The funding of the Scientific Advisory Panel itself is unique and worthy of mention.  It was decided that the cost of the exposure monitoring guidelines should be shared amongst the fungicide manufacturers.   Dr. Bill Leiss of Simon Fraser University, who chairs the B.C. Stakeholder Forum, wrote to all pesticide manufacturers and formulators who might have an interest in marketing antisapstain agents in British Columbia, asking them to contribute a small amount (on the order of several thousand dollars) to fund the design of the generic exposure assessment guidelines.  Seventeen companies contributed (listed in Appendix B);  they include firms from several countries in North America and Europe.      Guidelines for Antisapstain Exposure Assessment  19 4.  Objectives   The overall objective of the Scientific Advisory Panel was to develop guidelines for assessing occupational exposure to antisapstain products which take into account the unique conditions of fungicide application in the lumber industry.   Exposure assessments conducted with these guidelines should provide data that can be used, in conjunction with toxicological information, for risk assessment.  Risk assessments might be performed by government agencies, such as Health and Welfare Canada or the Workers' Compensation Board, or by the lumber industry itself, and could be used to make decisions about pesticide registration, specifying conditions of use, or setting exposure standards.  A secondary use of monitoring data gathered using such guidelines would be to provide the basis for dose-response evaluations of health outcomes in epidemiologic studies of antisapstain fungicides.   Our specific aims were as follows: 1. to conduct a survey of a representative sample of sites in British Columbia which currently operate antisapstain chemical treatment systems; 2. through interviews and direct observation at these sites, to obtain descriptive data on lumber treatment application methods, and the ways in which exposures may occur; 3. to review currently available data on exposure to antisapstain chemicals in the lumber industry; 4. to evaluate measurement techniques which have been used in similar work situations and which will address all relevant routes of exposure; 5. to review the literature on sampling strategies to ensure representative and biologically appropriate sampling; 6. to determine the range of physical, chemical, and biological properties of the antisapstain formulations proposed for use in the lumber industry; 7. to prepare a preliminary report proposing generic exposure assessment guidelines which addresses the concerns of the Stakeholder Forum; 8. to have the guidelines reviewed by interested parties in a venue providing an opportunity for interaction; and 9. to prepare a final report.   Guidelines for Antisapstain Exposure Assessment  20 5.  Methods  5.1  Site Visits   In order to determine the range of potential exposure conditions in the antisapstain application process, we conducted on-site observations of treatment processes at representative work sites. Active sites were identified to us by Environment Canada (Forest Products Division, Authorizations Branch, Environmental Protection, Conservation & Protection), and by the Workers' Compensation Board of BC (Occupational Health, Occupational Safety & Health Division).  The list provided by Environment Canada had last been updated in August 1991, by telephone survey.  The WCB of B.C. maintains a voluntary registry of antisapstain chemical users in B.C. and we received a copy listing all users registered as of October 1991.  A total of 67 currently active use sites were thus identified.   This list of sites served as our sampling frame.  Since there were only three shipping terminals, all were contacted for on-site observation.  Since many lumber mills were listed as using antisapstain agents, we needed to select a subset for observation in such a way that as complete as possible a range of application methods would be observed.  Because our list did not include complete data on treatment method, we attempted to retrieve data on other factors which might be related.  Data on such factors as size of work force and production volume were not readily available for the majority of mills.  Therefore, in consultation with several industry contacts, it was decided to stratify the mills into four regions (Vancouver Island, lower mainland, central interior, and the north), and to randomly select mills in proportion to the numbers in each region.    We wrote to the manager at each selected sawmill and shipping terminal to introduce the project and request a site visit (Appendix C).  Data collection procedures during the site visit consisted of the following:  interview of a manager knowledgeable about the antisapstain treatment process;  interviews with workers and supervisors involved in the operation, clean-up, and maintenance of the treatment system;  a site tour to observe the treatment devices and their chemical supply and mixing systems in operation;  and observation of the activities of production workers downstream from the antisapstain treatment system.  Data were collected in a semi-structured manner, and recorded on the form attached as Appendix D.  The description of each site included:  general characteristics of the site, its production activities and work force;  details about the antisapstain treatment process including the type of equipment used, how it works, and its location in the production process;  and opportunities for worker exposure including the job activities required to operate and maintain the system and those involving handling or processing of treated lumber.  Guidelines for Antisapstain Exposure Assessment  21 5.2  Available Data on Exposure   In order to determine the factors which are related to exposure and might therefore help determine how individuals would be selected for exposure sampling, we reviewed the scientific literature on antisapstain exposure measurements in the lumber industry.   In addition, we gathered unpublished data from a variety of sources on exposure measurements conducted within British Columbia and Washington state.  Where possible, information about jobs, chemical contact, and fungicide application processes was examined in conjunction with exposure levels to determine how worker groups might be selected for monitoring.  5.3  Literature Review on Measurement Techniques and Sampling Strategies   In order to address the issues of measurement techniques and sampling strategies, we conducted searches of the scientific and government literature related to pesticide exposure monitoring techniques and occupational hygiene sampling strategies.  The literature on exposure monitoring for pesticide applicators focuses mostly on techniques to ensure that all pertinent routes of exposure are assessed (11- 14).  This is a particularly important question since many pesticides may be dermally absorbed as well as inhaled, and many have direct irritant effects at the point of body contact (7,8,15-18).  The pesticide literature also addresses the issue of quality assurance, for example, recovery and replicate samples.  The hygiene sampling strategy literature points out additional issues which have received less attention in the pesticide registration process (19-21).  For example, it addresses the issue of how to select the individuals to be sampled, and the question of how many workers should be sampled.  It also asks whether the samples should be of long or short duration to reflect the way each specific chemical acts on the body.  5.4  Physical and Biological Properties of Antisapstain Chemicals   Because the exposure assessment guidelines are generic, they must be able to respond to any characteristics of the fungicides which affect their ability to enter the body, to be distributed to different body sites, or to cause harm at a particular body location.  We examined the range of these characteristics which might be present in antisapstain fungicide formulations, using the information made available to us. This included a Canadian government review document (9), but also data provided by the fungicide manufacturers contributing to development of these guidelines.  Each company was contacted by letter (Appendix B) and asked for the following information:  antisapstain product trade names;  Material Safety Data Sheets;  active ingredients, concentrations, and physical form;  instructions to users about safe handling;  industrial hygiene measurement methods;  and results of any worker exposure studies or toxicological studies.  Guidelines for Antisapstain Exposure Assessment  22  5.5  Design of the Exposure Assessment Guidelines   Using the data from the elements discussed above, we designed generic exposure assessment guidelines based on a consensus among the Scientific Advisory Panel members.  The following questions in particular were addressed: 1. What chemicals should be monitored? 2. Which monitoring techniques should be used? 3. How should measurement quality be assured? 4. What should the duration of the measurements be? 5. How many samples should be taken? 6. Which mill and shipping terminal locations, which workers, and which times should be monitored to ensure that the exposures measured are representative of the industry? 7. What supplementary information should be collected at the time of sampling to allow appropriate decisions to be made about future sampling and exposure control?   As part of the process of designing the exposure assessment guidelines, a draft of this document was sent out for review to the members of the Stakeholder Forum, the participating fungicide manufacturers, Health and Welfare Canada, Agriculture Canada, the Workers' Compensation Board of B.C., B.C. Environment, and other concerned groups.  A workshop to discuss the draft was held on April 23rd, 1992 at Simon Fraser University Harbour Centre Campus and was attended by manufacturer's representatives, forest industry and labour groups, as well as provincial government agency representatives.  The Scientific Advisory Panel considered the written and oral comments received, and incorporated changes into the revised final guidelines wherever appropriate.   Guidelines for Antisapstain Exposure Assessment  23 6.0  Results  6.1  Site Visits  6.1.1  Representativeness of  the Sample   Figure 1 shows the locations of the sawmills in B.C. using antisapstain agents in late 1991, and also indicates the locations of the mills visited.  Table 4 shows the numbers of work sites visited according to type of operation, region, chemical in use, and application method.  It indicates the numbers selected from each stratum, and the numbers actually surveyed by site visit.  It also shows the numbers ineligible because they no longer used antisapstain agents, and the numbers who refused or who did not give a clear yes/no answer to participation before our site visits were finished.  Although 13 sites agreed to participate, 14 were actually visited, because one additional site which was not randomly selected came forth and volunteered to be studied.  The number of northern B.C. lumber mills which were selected appears disproportionately large because this number includes extra names drawn to compensate for the high frequency of sites selected from this stratum which were subsequently found ineligible.  It is interesting to note that of the 23 work sites contacted, 5 had stopped treating lumber in the few months since the lists were updated and 4 had plans to change either their treatment method or their antisapstain chemical in the near future.   Users of one major formulation (borax and sodium carbonate) were underrepresented among the sites visited.  This occurred largely by chance;  the random selection of 24 sites drew only 2 users of this formulation (9%), even though they were 26% of the sampling frame and were present in every geographic region except Vancouver Island.  Comparison of the 14 sites visited to the 67 known use sites in the province in terms of treatment method is difficult because for most of the sites, we had no information on the method of chemical application in use.  However, the application methods of the sites visited were quite similar to those of the remaining mills whose method of treatment was known.   The stratified random sampling method resulted in inclusion of work sites represented by each of the labour unions belonging to the B.C. Stakeholder Forum on Sapstain Control, as well as several sites which were not unionized.  The sites belonged to 8 different forest companies and 3 shipping companies; no company had more than two of its sites included.  There were two large multi-site forest companies not represented;  one company was not selected in the random sampling, the other declined to participate. Some general characteristics of the 14 sites visited and their 29 antisapstain treatment systems are summarized in Table 5.  Guidelines for Antisapstain Exposure Assessment  24 Figure 1:  Map showing locations of sawmill using antisapstain agents in BC   Guidelines for Antisapstain Exposure Assessment  25 Table 4: Characteristics of selected site-visit locations in comparison to characteristics of all mills  and shipping terminals where antisapstain treatment is currently done in British Columbia     Number Number Number of  Total Number Not  Not Agreeing Sites    Number* Selected† Eligible¥ to Site Visit Visited   Current Use Sites 67 23 5 5 ††14  Type of Operation and Region  Shipping Terminal   Lower Mainland 3 3 0 0 3  Lumber Mill   Interior 5 2 1 0 1   North 16 7 4 1 2   Lower Mainland 23 6 0 2 4   Vancouver Island 20 5 0 2 ††4  Chemical In Use**  Unknown 1 0 0 0 0  Azaconazole 1 0 0 0 0  Borax & sodium carbonate 13 2 0 1 1  Borax & DDAC & IPBC 1 0 0 0 0  Copper 8 2 0 0 0 ††1  DDAC 12 7 0 2 5  DDAC & IPBC 27 8 0 2 6  DDAC & Borax 1 1 0 0 1  TCMTB 4 0 0 0 0  None 5 5 5 0 0  Application Method¥¥  Diptank 14 9 2 1 ††7  Spraybox 29 12 2 3 ††8  Unknown 26 3 2 1 0  * Number of sites currently using anti-sapstain treatments, identified from Environment Canada and Workers' Compensation Board of B.C. data. † Sites randomly selected, including replacements drawn for sites later found to be ineligible. ¥ Sites selected found to be no longer using antisapstain agents. †† Includes one additional site which volunteered to be studied. ** Abbreviations for treatment chemicals have the following meanings:  DDAC = didecyl dimethyl ammonium chloride;  IPBC = 3-iodo-2-propynyl butyl carbamate;  Copper 8 = copper-8- quinolinolate;  TCMTB = 2-(thiocyanomethylthio) benzothiazole. ¥¥ Numbers in this section may add up to more than the expected totals because some mills use more than one type of treatment system.    Guidelines for Antisapstain Exposure Assessment  26 Table 5: Characteristics of sites studied     Lumber   Treatment Method*  Type of Produced Treated Chemical Dip Spray Region Operation (m3/yr) (m3/yr) in Use FL FE AE SC LS XC CW   Vanc. Is. sawmill 260,000 260,000 DDAC+IPBC . . . . 2 . .  planer 130,000 130,000 DDAC+IPBC . . . . 1 . .  Vanc. Is. sawmill   Copper 8 . . 1 . 1 . .  planer   Copper 8 . . . . . 1 .  Vanc. Is. sawmill 200,000 200,000 DDAC . . . . 1 . .  planer 54,000 54,000 DDAC . . . . . 1 .  Vanc. Is. sawmill 346,000 346,000 DDAC+IPBC . . . . 2 1 .  planer 242,000 242,000 DDAC+IPBC . . . . 1 . .  L.  Main. shipping  †52,333 DDAC+IPBC 1 . . . . . .  L.  Main. shipping  †52,333 DDAC 1 . . . . . .  L.  Main. shipping  †52,333 DDAC+IPBC 1 . . . . . .  L.  Main. sawmill 87,000 87,000 DDAC+IPBC . . . . 2 1 .   L.  Main. sawmill 342,000 144,900 DDAC+IPBC . . . . . . 3  planer 47,000 9,400 DDAC+IPBC . . . . 1 . .  L.  Main. planer 85,000 67,000 DDAC . . . 1 . . .  L.  Main. sawmill   DDAC+borax . . 1 . . . .  Interior reman 113,000 62,150 borax . 1 . . . . .  North sawmill 250,000 247,500 DDAC . . . . . 2 .  North sawmill  nil none . . . . . . .  planer 300,000 270,000 DDAC . . . . 2 . .  * Abbreviations for treatment methods have the following meanings:  FL = forklift diptank;  FE = forklift with elevator diptank;  AE = automated elevator diptank;  SC = sorting chain diptank;  LS = linear spraybox;  XC = crosschain spraybox;  CW = carwash spraybox. † For shipping terminals, reported volume of lumber treated per year is the average of all 3 shipping terminal sites   Guidelines for Antisapstain Exposure Assessment  27 6.1.2  Types of Treatment Systems   The 14 sites visited included examples of 7 different types of antisapstain chemical application systems, namely: 1. forklift diptank 2. forklift and elevator diptank 3. automated elevator diptank 4. sorting chain (trough) diptank 5. linear spraybox 6. crosschain (transverse) spraybox 7. carwash spraybox Descriptions of the above seven categories of treatment systems follows.  Lists of the job tasks associated with the operation and maintenance of the various types of systems appear in Appendix E. There are other systems besides those listed above, for example, drive-in (carrier) diptanks, where a carrier vehicle picks up a stack of lumber and then drives down a ramp into a tank of chemical solution, immersing the lumber and dragging it through the liquid as it goes.  No representative of this category was among the 29 systems which we observed.  No one we contacted was aware of any example of this type of system still in operation in British Columbia.  Forklift Diptank   This system uses forklifts equipped with special extension devices which clamp down on a package of lumber, holding it below the level of the original forks and slightly away from the body of the vehicle. The forklift picks up a bundle of lumber, carries it to the edge of a diptank (at ground level or below), immerses the lumber, raises it and holds it briefly over the diptank to drain, and then carries the treated package to the designated drip dry area.   The essential components of this system are the diptank itself, the forklift fitted with extension forks, a holding tank for the liquid chemical concentrate, a supply of make-up water, and plumbing to connect the latter two components with the diptank. For most antisapstain agents, it is desirable to contain runoff, splashing, and tank overflow, both for economy and for purposes of environmental protection, therefore the entire area around the tank (including the drip dry area) is usually paved, sealed, dammed around the perimeter, and sloped towards a drain system and sump. Sump liquid may be either returned directly to the diptank, or used for make-up water. The diptank area is usually roofed but not enclosed, so rain may blow into the area and add to the drainage returning to the sump. Some systems  Guidelines for Antisapstain Exposure Assessment  28 therefore have additional holding tanks for the extra fluid when the sump threatens to overflow, or if the dilution of the sump contents is excessive.   The degree of automation of diptank mixing and supply systems vary. Some require a variety of manual operations with potential for chemical contact, for example, manually placing chemical concentrate supply hoses, opening and closing valves, testing fluid levels in tanks or sumps, and agitating the tank with a paddle.   The forklift diptank is one of the few systems where a worker (in this case, the forklift driver) continuously operates and controls the actual chemical application. Workers other than the forklift operator are needed in the diptank area while dipping is in progress, and task assignments may require the forklift driver to leave the cab. This is of note because lumber dipping produces perhaps the most dramatic immediate workplace contamination of any of the treatment methods we observed. Splashing and overflow from the tank occurs with each package of lumber immersed, and free fluid runoff is copious as packages are carried to the drip dry area.   Clean-up requirements associated with forklift diptank systems tend to be minimal, except for periodic major tasks. Drainage from the diptank area carries silt and debris into the system's circulation, but unlike spraybox systems, there is no need to keep the chemical solution clean and filtered to prevent nozzle plugging. Debris and sludge may be allowed to accumulate in large quantities in the sump and diptank, over long periods of time (up to a year or more), and then cleaned out all at once in an effort requiring several days, and a crew of labourers.  Forklift and Elevator Diptank   A forklift with ordinary forks moves packages of lumber one at a time on and off a hydraulic lift platform over the diptank. The elevator platform is lowered to immerse the lumber, and bars above the package prevent the lumber from floating to the surface of the tank. The system is otherwise the same as a forklift diptank system. Worker exposure considerations would also be the same, except that the elevator platform might present an additional piece of contaminated equipment to maintain and repair.  Automated Elevator Diptank   Bundles or stacks of lumber are moved by conveyor belts or chains to an elevator platform over a diptank. After dipping, the lumber stack is moved by conveyor again, this time off the platform and further down the line to a covered drip dry area. There may be a tilting platform immediately after the diptank  Guidelines for Antisapstain Exposure Assessment  29 which holds each freshly dipped package tilted at an angle for several minutes to enhance draining. After remaining in the drip dry area for a prescribed period (usually 4 hours), the lumber stacks are moved by conveyor to the end of the line for pick-up by forklift. The entire sequence, from the conveyors before the diptank, to the forklift dock at the end of the line, is usually completely automatic, requiring no operation other than setting controls at the beginning of each shift.   The floor of the drip dry area is diked and sloped to force drips and runoff to drain back to the sump. A pump continuously circulates fluid between the sump and the diptank.  The sump fluid is poured through a filter before returning to the diptank to protect mechanical components which are immersed in the tank (lift components, conveyor chains and sprockets).   The automated elevator system does not reduce the inherent messiness of a diptank system, but reduces its impact by removing people from the dipping procedure.  Sorting Chain (Trough) Diptank   This type of system is located in the production line where individual pieces of lumber are laid side- by-side perpendicular to the line's direction of movement, as is the case in the sorting/grading areas. Individual pieces of lumber moving along the line are carried by sprocket-driven transport chains down into, through, and then out of a shallow immersion tray, and on down the production line. Lumber pieces are held down by weighted wheels as they pass through the diptank.   Sorting chain diptanks have a full mixing and supply system, usually in a room or enclosure away from the diptank itself. A pump system constantly circulates fluid between the chemical feed tank (also called a daytank or mixing tank) and the diptank, filtering the return from the diptank before allowing it to enter the daytank. The daytank is kept full with water and chemical concentrate delivered by a proportioning pump; alternatively, there may be a batch mixing tank intermediate between the chemical concentrate holding tank and the daytank.   As with all diptanks, the lumber is wet and dripping as it emerges from treatment. However, because of its in-line position, the opportunities for downstream worker contact with freshly treated lumber are greatly increased with a sorting chain diptank, compared to other types of diptank systems.   Guidelines for Antisapstain Exposure Assessment  30  Linear Spraybox   A linear spraybox is an enclosure, open at both ends, positioned longitudinally over the production line. Individual pieces of lumber, moving end to end along the line, approach the spraybox and are picked up by speed-up rollers or belts which shoot them through the boxes at high velocity. As each piece passes through the box, it is sprayed with a fine mist of chemical solution, generated by 4 to 10 small-aperture spray nozzles. To contain the spray, there are brush curtains at each end of the box, and an exhaust system draws air up through a roof stack, maintaining a constant negative pressure in the box. Before venting to the outside atmosphere, the exhaust gases pass through a condenser/mist eliminator, to recover as much chemical solution as possible. Condensation and runoff inside the spray box is collected by a drip pan in the bottom of the box, and drains back to the mixing and supply system.  Linear sprayboxes can treat lumber of any length, as long as the cross-sectional dimensions are smaller than the dimensions of the box (so a typical linear spraybox, with cross-sectional dimensions of 50 cm x 50 cm can handle even enormous timbers).   Linear sprayboxes have full mixing and supply systems, usually in a separate room or enclosure. One mixing and supply system often serves more than one spraybox.  There is constant circulation of fluid between the chemical feed tank and the spraybox, with the drainage from the spraybox being filtered before it is allowed to return to the daytank.  The daytank is kept full with water and chemical concentrate delivered by a proportioning pump.  Drips and spills in the chemical mixing room and from around the spraybox drain to a sump;  sump contents are used as make-up water.  A powerful pump delivers fluid from the mixing tank to the spraybox at high pressure (150 psi or more).  Throughout the mixing and supply system there are multiple filters, strainers and screens to remove sludge and debris from the circulation fluid and thus reduce the frequency of spray nozzles clogging.   The mixing and supply component of a linear spraybox system tends to be more complex and time consuming to operate than that of a diptank system.  Maintaining an exact chemical concentration in the mixing tank is important, and this requires multiple flow calibrations or colorimetric indicator tests during a shift.  Many times throughout a shift, filters will have to be checked and emptied, and nozzles may have to be unplugged.  Flushing and cleaning of the spraybox and its components is done weekly, if not daily, since it is important to keep dirt and debris out of the system.   The combination of the high speed of lumber movement through the spraybox, high-pressure small-aperture nozzles, and fine mist spray usually results in little if any free fluid carryover on treated lumber as it leaves the spraybox. However, the in-line position means that some handling of freshly treated lumber in downstream areas is inevitable.  Guidelines for Antisapstain Exposure Assessment  31  Crosschain (Transverse) Spraybox   As the name suggests, this type of spraybox is positioned across the production line, with its length at right angles to the direction of lumber movement.  Individual pieces of lumber moving side by side are carried by sprocket-driven transport chains through the spraybox and on down the production line. Crosschain sprayboxes tend to be much larger than linear sprayboxes, because they must be as long as the longest piece of lumber. They also must have spray nozzles positioned along the full length of the box in order to completely cover each piece of lumber with spray. This combination of factors means that crosschain sprayboxes 7 m or more in length, with 48 spray nozzles or more, are not uncommon.   More nozzles means more frequent episodes of nozzles getting plugged and needing service. The size and design of some crosschain boxes makes it impossible to reach nozzles without climbing into the spraybox, though in others spray nozzles are accessible from outside the box.  The transport chains and sprockets moving lumber through the box also represent contaminated mechanical components which do from time to time require maintenance and repair. Finally the slow speed at which lumber passes through a crosschain spraybox often (but not always, since other factors such as pressure and nozzle size also play a role) results in more free fluid remaining on lumber as it exits the crosschain box than with a linear box.   The mixing, supply, and recovery systems associated with crosschain sprayboxes are not substantially different than those associated with linear sprayboxes.  Carwash Spraybox   A carwash spraybox is an automated system where bundles of lumber are moved by conveyor (sprocket-driven chains or rollers) through a spraybox, within which 25 to 60 large-aperture nozzles flood the package with antisapstain chemical solution in sufficient quantity and for a sufficient period of time to allow penetration into the inner layers of the package. Drainage after spraying is enhanced by a tilt platform which holds each treated stack at an angle for several minutes. The tilt platform can be located either right in the spraybox, or over a drip dry area just past the box. The large aperture of the spray nozzles means episodes of nozzle clogging are infrequent. The low pressure flood nature of application results in much free fluid carryover on the treated lumber (as much as with a diptank) but again, as with most diptank systems the fact that bundles and not individual pieces of lumber are treated reduces opportunity for contact with freshly treated lumber.  6.1.3  Location of Treatment Systems in the Production Process  Guidelines for Antisapstain Exposure Assessment  32   The positions of the 29 observed antisapstain treatment systems within the production processes are summarized in Figures 2,3, and 4 (overleaf).  Figure 2 is a generic production flow scheme for a sawmill (with or without a timberdeck);  Figure 3 represents planer and remanufacturing mills;  and Figure 4 represents a lumber treatment facility within a shipping terminal. The flow lines show all the pathways for lumber which were identified at the sites visited;  they are not intended to be exhaustive of all the possible pathways, nor do they suggest that all the lines exist at any one site.  For each production step through which freshly treated lumber could flow, the production job categories are listed.  The presence of workers involved in equipment maintenance and repair, and clean-up in each production area is implicit, though they are not listed.   There was wide variation in treatment system placement among the lumber mill sites we visited. One of the few clear patterns to emerge was that treatment methods which processed stacks or bundles of lumber (i.e., forklift diptanks, forklift and elevator diptanks, automated elevator diptanks and carwash sprayboxes) tended to be placed near the end of the production process, with few steps (and few workers) after treatment occurred.  Also, these systems were located beyond the last point at which freshly treated lumber might be diverted back into the mill for further sawing, planing, sorting or grading.  Guidelines for Antisapstain Exposure Assessment  33 Figure 2:  Observed locations of antisapstain treatment in sawmills  Guidelines for Antisapstain Exposure Assessment  34 Figure 3:  Observed locations of antisapstain treatment in planer and remanufacturing mills  Guidelines for Antisapstain Exposure Assessment  35 Figure 4:  Observed locations of antisapstain treatment in shipping terminals  Guidelines for Antisapstain Exposure Assessment  36 6.2  Available Data on Exposure   Existing data from earlier exposure studies, if available, are often very helpful in designing sampling strategies for further monitoring, especially if factors which influence exposure can be identified and allow workers to be grouped for sampling.  Such data may indicate certain jobs or activities which have high potential for exposure, and may indicate others which are unlikely to have exposures beyond background levels.   Studies referring to antisapstain fungicide exposures in sawmills mainly report exposures to chlorophenates (10-19), though two B.C. reports also give results for other agents (19,20).  Although chlorophenate exposures are not likely to be the focus of exposure studies pursuant to these guidelines, the results of these earlier studies are useful to the extent that they provide data about how such factors as season, job, proximity to wood treatment areas, skin contact with wet lumber, airborne concentrations, and protective equipment influence exposure.  The following is a summary of the results of the earlier studies which may guide the design of generic exposure monitoring guidelines.  Unless otherwise indicated, the results reported refer to chlorophenate measurements, since these predominate in the literature.  6.2.1  Range of Exposures   In all studies which monitored exposure at a variety of jobs and locations throughout lumber mills, variability in exposure between individuals was great, with differences between the highest and lowest measured values being more than one order of magnitude in most studies (11,13-16,18,19).  In one study, differences as high as three orders of magnitude were observed (16).  Chlorophenates measured in urine ranged from a low of 2 ppb (13) to a high of 48,800 ppb (16), with measurements often ranging from 20 to 2,000 ppb within a single mill (18,19).  The lowest levels measured in these lumber mills correspond to background levels found in populations not known to be occupationally exposed who have means of between 20 and  50 ppb (21).  6.2.2  Influence of Time of Sampling   The two studies (14,18) which examined the influence of season of sampling both showed higher exposures in warmer summer weather, with mean levels more than twice as high in the summer than the fall or winter.  Chlorophenates have excretion half-lives on the order of 4 days to 10 weeks (22), therefore fluctuations in exposure over periods shorter than seasons, for example hours or days, have little influence on the accumulated body burden.  Therefore variations in exposure due to systematic differences by day of the week or shift cannot be distinguished in urinary chlorophenate data (14).  Guidelines for Antisapstain Exposure Assessment  37  6.2.3  Influence of Type of Contact   Several studies attempted to separate workers into categories according to factors judged likely to influence exposure.  Kleinman and Horstman's study (18) of a Washington state sawmill examined several factors:  skin contact with wet or dry treated wood;  proximity to the spray box;  and adequacy of protective equipment.  Workers who handled wet treated lumber were found to have mean urinary chlorophenate levels 2.5 to 7 times higher than those who had contact with dry treated lumber or little contact with treated lumber.  Individuals working near the spray box or downstream of it had mean exposures 1.5 to 4.5 times higher than those upstream of the box, and 3 to 10 times higher than outside workers.  There was also evidence that maintenance workers who intermittently worked on the spray system had lower overall urine levels than workers handling treated wood on a continuous basis.  Adequacy of protection (not defined in detail by the authors) did not have a consistent relationship to exposure.   Embree et al (15) categorized workers into three groups by type of exposure:  those with manual contact with treated wood (pullers, graders, and stencilmen);  those with inhalation exposures only (slasher saw, trimmer saws, and day cleaners);  and those with no exposure (office and log pond workers). Mean serum levels of chlorophenates were the highest in dermally exposed workers, 3 times lower in those with airborne exposures, and more than 10 times lower in the unexposed group.  Lindroos et al (16) used similar groupings and found that workers with mainly skin contact had the highest median urinary chlorophenate levels (stackers, loaders, trimmers, graders, and packers).  Those with mainly inhalation exposures had urinary levels 8 times lower;  these were kiln workers at a Finnish mill where lumber was sometimes kiln-dried after antisapstain treatment.  Workers with equal contributions from skin and inhalation exposure routes had intermediate levels (outdoor dip tank workers and mixing room attendants).   Several studies (11,15,18,19) have measured both urinary chlorophenate levels and airborne exposures to more directly assess the contributions of inhalation and dermal absorption.  Consistent with the results reported above, inhalation exposures were shown to contribute only a fraction of the chlorophenate burden in those workers who had opportunity for skin contact.  Similar conclusions were reached by Fenske et al (10) in a study evaluating the contribution of dermal exposure in sawmill graders and lumber pullers.  6.2.4  Influence of Job or Task   Many studies provide exposure data according to job or task (11,16,19,20).  Most of these have selected only a portion of the work force for sampling, usually those workers assumed to be the most  Guidelines for Antisapstain Exposure Assessment  38 exposed.  Because the locations of antisapstain treatment (pre- and post- grading, sorting, bundling, etc.) vary from mill to mill, exposures by job title must be considered with treatment location in mind.   The highest exposures reported in the literature were from two Finnish studies (11,16);  the exposed workers held jobs as loaders in sawmills where trough-dipping of single boards was done.  In the study by Kauppinen and Lindroos (11), the median urinary chlorophenol levels for the remaining job categories were ordered from high to low as follows:  jobs nearest spray boxes;  preparation of the treatment solution;  machine stacking or truck transport of treated lumber;  trim grading and packaging, and kiln operation;  and jobs near outside dip tanks.  Several cautions should be borne in mind, as indicated by the authors.  Few samples were taken for some job categories, and although the median levels may differ, the ranges of exposures experienced in many of the job groups overlapped.   McDonald (19) summarizes the results of studies conducted in three B.C. sawmills in 1985 and 1986.  All mills had spray boxes directly after the planer, with graders and chain pullers stationed after treatment.  One mill also used a drive-in diptank with opportunities for the carrier truck driver to be exposed.  Another mill had spray boxes in the sawmill, again with grading and sorting downstream.  The diptank carrier driver had the highest mean exposures, then graders and chain pullers, mix room attendants, and stacker operators.  Spray box maintenance staff and yard tally/stencil crew had the lowest exposures of those measured.   In the only study (20) reporting results of TCMTB measurements, airborne exposures were highest in the mixing areas, lower in grading areas near spraybox outflow, and below detection limits in most other areas.  Opportunities for skin exposures were only assessed for graders and lumber pullers;  results showed similar ranges of TCMTB contamination in gloves for these two job categories.  In a study examining dermal and inhalation exposures to IPBC (19), all measurements were below detection limits except one dermal exposure patch on a stacker operator.   The University of British Columbia Department of Health Care and Epidemiology has unpublished data showing total urinary chlorophenate levels for workers in all jobs in two B.C. sawmills where exposure studies (13,14) were conducted in the mid-1980s.  One of the mills sprayed large timbers at the sawmill timberdeck, and sprayed dimension lumber in the planermill directly after load breakdown and planing, prior to all other planer jobs.  The other mill treated lumber at an outside drive-through dip tank, and in a spray box also immediately after the planer.  Urinary chlorophenate levels in boom area, kiln, and power plant workers, and in carpenters, machinists, oilers, pipefitters, sawfilers, and welders were consistently less than 100 ppb, near background levels.  In the maintenance trades, some electricians, millwrights, and mechanics had higher levels.  In the mill with the diptank, about half the yard workers had exposures over 100 ppb  Guidelines for Antisapstain Exposure Assessment  39 (112-449 ppb), and the carrier drivers were second highest exposure job in the mill (mean = 396 ppb).  In the other mill, only three yard workers had exposures higher than background:  two swampers and a tallyman.  Most of the sawmill workers in both mills had low exposures, however 24 of 143 measurements were over 100 ppb (113-772 ppb).  Nine of these individuals had recently moved from jobs with more opportunity for exposure, but for the remainder there was no explanation for the higher levels.  Of the 120 measurements from planermill workers, 79 had measurements in excess of 100 ppb (109-1175 ppb). Greenchain pullers had the highest exposures in both mills, with graders and others near the spray box or handling wet wood also showing high levels.  Of the planer jobs, none had chlorophenate levels consistently below 100 ppb in both mills.  6.2.5  Rationale for stratif ication by season and by work group   Exposures in the summer appear to be higher than those in the fall and winter, therefore it is important that measurements in both warm and cold weather are included to appropriately estimate mean exposures over time.  Data from Environment Canada over the last 20 years indicate that, in British Columbia, the optimum division of the year into two equal parts based on temperature and rainfall is the following:  April to September inclusive, and October to March inclusive.  To maximize the chance of including measurements in both these periods, we recommend selection of one measurement per worker after stratification of worker-shifts into these two periods.  However, where an additional measurement is made for each worker to estimate within-worker (shift-to-shift) variability, it is important that the additional shift is selected at random from throughout the entire year so that shift-to-shift, not season-to- season, variability is measured.   Stratification of workers into groups according to exposure potential should maximize the precision of estimated mean exposures for a given sample size.  It also ensures that exposure data is available on each stratum included in the study.  Based on the antisapstain measurement data described above, several strata can be recommended.  Two strata are proposed for the groups which appear to have the highest exposure potentials:  graders and lumber pullers who handle wet wood; and carrier drivers dipping lumber in drive-through tanks.  Because drive-through dip tanks are not in use in British Columbia at the present time, workers who may have similar exposure potentials, i.e., forklift drivers who dip lumber and elevator dip tank operators, are included in the latter group.  Another stratum is proposed for maintenance workers who work on the treatment systems;  these individuals have shown elevated exposures in studies to date, but levels were not as high as continuously exposed workers.  Fewer mills currently have graders and lumber pullers downstream of treatment systems, therefore a stratum was created to ensure that other jobs which require less frequent handling of wet wood, e.g., bin sorters, are  Guidelines for Antisapstain Exposure Assessment  40 monitored.  A final stratum includes workers who handle treated lumber which is dry, so that some exposure monitoring is conducted in mills where controls minimize or prevent exposure to wet wood.   An additional stratum could be created to include all other mill employees, including those upstream of the treatment process and, perhaps, office personnel.  Because exposures measured in such workers to date confirm that they have low probabilities of exposure, we have not recommended that this group be included in an exposure study.  Such a stratum would be useful, however, where there is a need for data on background exposures in the treatment environment.  It should be noted that foremen and supervisors may not belong in this background exposure group.  In some mills, foremen may maintain the antisapstain supply system or they may be included in the groups handling wet or dry treated lumber.   Although there is a considerable amount of data describing differences in exposure according to job, wet lumber contact, and proximity to treatment equipment, there is little information which compares exposures between mills with different treatment methods, therefore stratification based on treatment system was not attempted.   Guidelines for Antisapstain Exposure Assessment  41 6.3  Review of Measurement Techniques   Occupational exposures to chemical hazards may be assessed by two complementary approaches: environmental monitoring of respiratory and dermal exposures, and biological monitoring of chemical levels in body fluids or tissue.  Data derived from either of these approaches require interpretation by means of laboratory-based models to estimate doses received by humans during particular work activities.  Studies which employ both environmental and biological monitoring are most useful in elucidating absorbed dose and health risks resulting from chemical exposures.   Numerous reports have indicated that the dermal route of exposure is the major contributor to total absorbed dose during occupational exposure to antisapstain agents (10,11,20).  Consequently, recent industrial hygiene studies have focused on biological monitoring (13,14,23) and to a lesser degree on direct assessment of dermal and respiratory exposures (10,19,20).  6.3.1  Biological Monitoring   Regulatory agencies responsible for pesticide registration have recognized the complexities inherent in biological monitoring.  Although such an approach to exposure assessment is strongly encouraged, current guidelines rely on environmental monitoring (8,9,24).   Widespread use of chlorophenolic compounds as antisapstain agents created almost ideal conditions for use of biological monitoring;  i.e., substantial animal and human pharmacokinetic data bases have been developed, and these compounds are excreted primarily in urine with moderately long half-lives (22).  However, with the introduction of new antisapstain agents as substitutes for the chlorophenolic compounds, biological monitoring becomes problematic.  Unless, and until, adequate human pharmacokinetic data bases are established for these agents, interpretation of biological monitoring values would be difficult.   Biological monitoring has therefore not been incorporated in these generic exposure guidelines, but this omission does not preclude biological monitoring becoming a feasible or even preferred exposure monitoring option in the future.  If guidelines for biological monitoring are developed at some future date, certain elements of the proposed exposure assessment are likely to remain the same, e.g., the representative sampling methods.  Other elements may change, e.g., biological sampling could not be conducted when workers are wearing an additional layer of sampling garments under their regular clothing for dermal exposure assessment.   Guidelines for Antisapstain Exposure Assessment  42 6.3.2  Dermal Exposure Measurement   The largest body of published material addressing dermal exposure has been developed in the area of exposure to agricultural chemicals (25-27).  Three techniques currently employed to estimate pesticide skin deposition are discussed here briefly.  Patch Technique    The patch technique involves attachment of small, passive collection devices ("patches") to portions of the body.  Deposition rates to the patches are then extrapolated to larger body surface areas (28,29).  Use of patches is normally accompanied by handwashing or glove monitors to assess hand exposure.  The patch technique has been employed extensively to monitor pesticide applicators and is currently the procedure recommended by both the U.S. and Canada (8,9,24).   The validity of the patch technique as an exposure assessment method rests on one of two critical assumptions: 1) uniform exposure; i.e., the deposition rate on the patch is representative of deposition over the body region ; or 2) worst-case exposure; i.e., the patch has been located at the point of highest exposure potential for the body region.  These assumptions have not been investigated systematically for pesticide mixers and applicators, and thus the exposure estimates derived from this technique are open to question (30-32).   The use of the patch technique for antisapstain agent exposure appears to be unsuitable for two reasons:  dermal exposure is not uniform in most cases, since exposure results from contact with equipment or treated lumber rather than through spray deposition;  and the workplace includes a wide variety of work activities, each with unique skin exposure patterns.  Whole Body Technique   This approach requires that workers wear garments which cover virtually all skin surfaces.  The garment is later sectioned and extracted to account for total deposition (33,34).  This approach was first recommended for use among pesticide applicators (28), and has been employed in timber mills to estimate exposure to the arms and chest (20).  Most recently, the California Department of Food and Agriculture (CDFA) used garments covering nearly all of the body to estimate pesticide transfer from treated carpets during an exercise session (35,36).   Guidelines for Antisapstain Exposure Assessment  43  This technique has several advantages:  it does not require the assumptions of the patch technique, since the entire body region is sampled;  a standard measurement approach can be applied to virtually all body regions (e.g., hands, feet, arms, legs);  measurements across work activities with different dermal exposure patterns are comparable.   Several disadvantages can also be noted:  whole body garments are more cumbersome than patches in the field;  removal of samples from workers, and sample storage/transport are more complicated;  extraction of whole body garment samples requires use of large volumes of solvents;  cloth garments are susceptible to breakthrough.  Thus, garments may have to be changed over the course of the work shift to ensure accurate exposure values.   A major impediment to widespread use of this approach is the lack of standard garments to serve as collection media.  In studies to date, garments have been selected for convenience rather than according to scientific criteria, and garment characteristics have not been well described.  For example, a BC Research study described the dermal sampling shirts by brand name and simply as "extra-large, long- sleeved, 100% cotton" (20).  Similarly, the CDFA study noted brand names, and described the garments as "54% cotton/36% polyester/10% spandex fabric tights, medium long-sleeved 'T-shirt' of 100% cotton, thin 100% cotton gloves, white 'athletic' socks of 100% cotton" (35).   At a minimum, garments should be characterized by fiber content, construction, finish, weight per unit area, and total weight (37).  Additional parameters which may be useful include thickness, air permeability, spray rating, yarn warp, count fill, and surface energy.  Ideally, garments employed as dermal collection media would be categorized according to their ability to adsorb, absorb, and retain the particular chemical under study.  Fluorescent Tracer Technique   This approach involves introduction of a fluorescent compound into the pesticide mix.  Tracer deposition is quantified under long-wave ultraviolet light with a video imaging system, and dermal fluorescence is related to pesticide exposure by an empirical transfer factor (38,39).   The fluorescent tracer/video imaging technique has been used in timber mills (10) and shows promise for quantifying dermal exposure, but at present the method is technically complex.  In addition to video imaging sampling, extensive chemical residue sampling is required to extrapolate dermal fluorescence to chemical exposure.  In cases where protective clothing is worn, separate studies must be conducted to determine the relative fabric penetration of the tracer and the chemical of interest.  Guidelines for Antisapstain Exposure Assessment  44   Fluorescent tracers can be used qualitatively to indicate specific dermal exposure patterns and to identify dermal exposure among worker groups expected a priori to have little or no exposure potential (40).  This approach can also prove valuable for worker education and training (41). Rationale for Dermal Exposure Measurement Technique   Any material employed as a sampling device for exposure assessment studies should be well characterized and, ideally, standardized.  Alpha-cellulose (a commercially available filter paper) has been used as a standard material for patches for nearly 40 years.  Similarly, surgical gauze pads serve as a reasonably standardized sampling medium when their size and thickness are characterized.  As yet no standardized garments have been developed for whole body dermal sampling, as noted in section The parameters for such garments outlined in these guidelines serve as a starting point for standardizing this technique.  Cotton (100%) is selected for simplicity and because it is widely used in the manufacture of garments such as T-shirts.  The selection is not based on an analysis of physical or chemical properties related to absorption/adsorption processes.  A minimum weight per unit area (200 g/m2) is assigned since lighter fabrics will likely exhibit substantial breakthrough over the course of the proposed sampling periods. Fabrics treated for water repellence are excluded from use since antisapstain products are commonly used with a water vehicle.   The sampling approach in these guidelines aims to measure the amount of material reaching the skin of the worker.  This approach is in contrast to that of the U.S. EPA's Subdivision U guidelines (8), where patches are placed on the outside of work clothing or protective clothing, and penetration factors are assigned subsequently to estimate dermal exposure.  With EPA's approach, the exposure values are independent of the particular clothing worn at the time of sampling, whereas the use of whole body sampling garments next to the skin will be highly influenced by whatever clothing is used.  This latter approach should achieve the best estimate of workers' exposure, since it includes, by definition, the effect of penetration of workers' clothing, including the variability in that clothing as it occurs in the workplace. A drawback of the approach proposed here is a possible bias related to protective clothing use;  i.e., workers may wear clothing on the study day that they would not wear routinely, or may use such clothing in a non-routine manner.  It is therefore essential that the study staff instruct study participants regarding normal usage and record any deviations from routine practices.  It will also be necessary to have sampling garments available on-site in a variety of sizes to insure that the garments fit the worker snugly and that they cover the entire body region.   Guidelines for Antisapstain Exposure Assessment  45  The sampling schedule is designed to collect valid total workshift samples.  Changing of sampling garments periodically during the workshift is suggested both for practical purposes (e.g., workers will take protective gloves off at breaks) and to avoid sampling garment saturation and breakthrough.  Thus, the total workshift sample for the hands, for example, will consist of 4 pairs of sampling gloves, and all of these sampling gloves can be pooled for extraction and analysis to reduce the analytical load.  The five sampling garments will cover virtually the entire body surface area.  The only notable exception is the face. It is deemed impractical to ask workers to wear the equivalent of a ski mask for an entire workshift.  The absence of a sampling garment for the face may result in an underestimation of total dermal exposure, but the contribution of facial exposure to total body exposure is likely to be very small.  6.3.3  Respiratory Exposure Measurement   Respiratory exposure has been shown to be a relatively small contributor to total absorbed dose among most workers exposed to chlorophenolic compounds (11,15,18,19).  However, there may be some work activities for which respiratory exposure is very important.  Furthermore, absorption is generally much more efficient through the respiratory route than through the skin.  Thus, even in the case of chlorophenolic compounds, air sampling is an essential component of a comprehensive exposure assessment.   It would be inappropriate to assume a priori that respiratory exposure is of minor importance for other antisapstain agents. The relative contribution of respiratory and dermal exposures to dose is highly dependent on the physical and chemical properties of the agents; e.g., highly volatile compounds are likely to result in greater respiratory exposures.  In the case of compounds which are not well absorbed through the skin, respiratory exposure will likely be the major route.  Finally, some compounds may have their most significant toxicological effects on the lungs, requiring particular attention to the inhalation route.  Thus, for a variety of reasons air sampling should be included in any exposure assessment study.   Personal air sampling is the only method considered appropriate for determining respiratory exposure.  The sampling train should be designed to collect both vapours and inspirable aerosols. Inspirable aerosols are those defined by the American Conference of Governmental Industrial Hygienists (42) as "inspirable particulate mass", i.e., aerosols less than or equal to 100 µm in aerodynamic diameter (da) captured with the following efficiency (E):  E = 50(1+ exp[-0.06 da]) ± 10.  It is important to measure all inspirable aerosols, not just aerosols small enough to reach the gas exchange tissues of the lung, because antisapstain agents which are deposited in the respiratory tract may be absorbed through mucous membranes or swallowed, and thus may contribute to total absorbed dose.   Guidelines for Antisapstain Exposure Assessment  46 6.3.4  Quali ty Assurance   Adherence to good laboratory practices and effective quality control procedures is now commonplace in most analytical laboratories (43), but these principles have been extended only recently to the collection, handling, storage, and transport of samples during pesticide exposure field studies. Laboratory and field quality assurance procedures are needed to ensure that samples are handled in a systematic, well-documented manner, such that the accuracy of resulting data is verifiable.   Laboratory quality assurance procedures documenting adequate extraction efficiency and storage stability are fundamental to studies of this nature.  The quality assurance goals presented in the guidelines (extraction efficiency of 90% ± 10%;  losses during storage of < 5%) may not be attainable for all compounds.  In such cases a thorough effort to attain these goals should be demonstrated, and sample values should be adjusted (increased) by well-documented correction factors.  Laboratory blank samples provide a check on performance when following standard operating procedures in the laboratory.   Field quality assurance procedures are intended to identify losses or contamination which may occur during sampling or between sampling and analysis.  The primary concern regards potential degradation of samples during field collection, temporary storage, and transport.  The quality assurance procedures for field losses recommended here attempt to account for differences between the study design proposed in these guidelines and study designs proposed in traditional guidelines for agricultural chemical exposures (8); i.e., samples in this study will be collected one worker at a time at many sites over an extended period of time, rather than in batches at a few sites in a relatively short time.   Two initial field loss studies are called for in these guidelines to demonstrate that samples can be handled without significant losses.  This approach is recommended in lieu of requiring an extensive field loss study (e.g., multiple replicates at multiple spiking levels) for each field sampling event, since the event will normally only include one or a few workers.  The initial field loss studies are designed to separate possible losses due to environmental conditions during the sampling period from losses due to handling, temporary storage, and transport.  If minimal losses (< 10 or 20%) can be demonstrated in these studies, field sampling can be undertaken with confidence.  It is recognized, however, that circumstances are likely to vary from sampling event to sampling event, such that the initial field loss studies may not be representative of the particular conditions and procedures related to individual field sampling events. Thus, collection of two field spikes (at one of three spiking levels, selected randomly) is proposed for each sampling medium during each field sampling event.  In addition, temperature, relative humidity, and location data will be collected for both the field spikes and the field samples to allow statistical techniques (e.g., multiple regression) to be used to develop a calibration curve to correct for losses under different  Guidelines for Antisapstain Exposure Assessment  47 sampling conditions.  If, however, the data collected in this manner is unable to predict losses, and field spikes collected during a particular field sampling event indicates significant losses, then the field samples collected during this event will be invalidated.  Guidelines for Antisapstain Exposure Assessment  48 6.4  Review of Sampling Strategies   Sampling strategies address the issue of how to ensure that exposure measurements are representative of the risk by answering the following questions:  how long, when, where, and who to sample;  and how many samples to take (44).  Although these questions have been raised for at least 40 years in regards to occupational exposures (45), the theoretical and empirical bases for answering them have recently been addressed more extensively in the occupational hygiene literature (46-55). Representative sampling is a vital issue, because as described above, work place exposure measurements can vary by several orders of magnitude, and inattention to sampling strategy may result in misleading conclusions.  It is interesting to note that exposure variability in work places is usually much greater than measurement variability (50), suggesting that sampling strategies should merit at least as much attention as quality assurance methods used to ensure precise and accurate measurements (8,9).  6.4.1  Shape of  the Exposure Distribution   One issue related to sampling strategy is the likely shape of the exposure distribution and the best summary measures to describe the data.  Most investigators concur that occupational and environmental exposures are frequently positively skewed (45-47,54) (Figure 5 overleaf).  Such distributions are usually well modeled by the log-normal function, such that the logarithms of the measurements will have a distribution with a normal (bell-shaped) curve.  In a log-normal distribution, the geometric mean is equivalent to the median, and is less than the arithmetic mean.  Most investigators (47,54) agree that the arithmetic mean should still be used as the appropriate summary measure since it weights the influence of high and low exposures equally, and is therefore the best predictor of cumulative dose and biologic effect.  The asymmetric variance of a log-normal distribution is well described by the geometric standard deviation and is used to calculate confidence limits around the arithmetic mean (47,54).  6.4.2  Duration of Measurement   In order to address the question of how long to measure an exposure, one must consider the fact that the duration over which each exposure measurement is taken affects the variability of the resulting measurements (53).  Measurements taken over short periods will include the peaks and troughs of work site concentrations, whereas longer term measurements will tend to average out these fluctuations. Therefore all exposure measurements should be made with the same averaging time in order to achieve a stable estimate of the variance of the distribution (47,54).  Roach (55) first Figure 5:  Example of the distribution of chlorophenate exposures in a B.C. sawmill  Guidelines for Antisapstain Exposure Assessment  49 0 5 10 15 20 25 0 100 200 300 400 500 600 700 800 900 1000 N u m b e r o f O b se rv a ti o n s Arithmetic Mean (103 ppb) Geometric Mean (43 ppb) Urinary Chlorophenate Data (in ppb) from a B.C. Sawmill These data illustrate the positively skewed shape of an approximately log-normal distribution.  They also show the relationship of the arithmetic mean and geometric mean.  suggested that, since the body also acts as a time-based collection device, measurement duration should be related to the speed with which the body processes a substance.  Biological sampling has the advantage of reflecting total dose rather than exposures by specific routes, and automatically incorporates the biologically appropriate averaging time into the measurement.  For measuring respiratory and dermal exposures, Roach and Saltzman (52,55) developed models which suggested that measurements should be averaged over periods three to four times the biological half-life of a chemical in order to appropriately estimate the biologically important exposure.  Thus for substances with chronic effects, measurements would focus on long-term average exposures, and for substances with acute effects, measurements would attempt to capture peaks measured over shorter durations.   Guidelines for Antisapstain Exposure Assessment  50  This strategy presented the problem that several different sampling durations might be required for substances with effects on multiple body compartments, each with a different half-life.  Rappaport et al (51) addressed this issue by showing that, for log-normally distributed data, arithmetic means based on long-term averages can be used to estimate the maximum frequency with which peak concentrations would be exceeded.  He therefore recommends that shift-long or longer averages be used as the basic measure for exposure assessment, except where acute exposures are likely to cause permanently disabling effects.  These latter situations (e.g., hydrogen sulfide exposures in the petroleum industry) are best handled by continuous alarm monitors, and are not likely to be the subject of risk assessments for antisapstain fungicides.  EPA's Subdivision U (8) does not specify measurement durations, and suggests that the issue be decided on a case-by-case basis depending on the measurement technology for the pesticide in question.  6.4.3  Who to Sample   Before addressing the question of who to sample, it is vital to ask what the purpose of the exposure measurements is.  Exposure data gathered using these guidelines will be used for risk assessment, that is to help make decisions about whether a chemical is registered, what the conditions of registration should be, and perhaps to set exposure limits.  The data will be used in conjunction with toxicologic data to predict likely health outcomes at the exposure levels measured in user populations.  In most cases the outcomes of interest will be chronic in nature, but for some agents there may also be acute effects of concern.   In such a scenario, it is important that the sampling strategy allow estimation of the mean and variance of worker exposure, especially for those workers in the upper end of the exposure distribution, since they will be the most vulnerable to adverse health outcomes and may provide evidence which will limit the use of a specific agent.  One question to ask then is whether all exposed workers should be sampled or just those who have the highest exposure potential.  The U.S. EPA, in its Subdivision K (9), allows sampling of the "maximum exposure activity" to minimize exposure measurement effort, but demands documentation of the rationale used to select a particular activity as representing the worst case.  Similar suggestions have been made by authors in the occupational exposure compliance literature (44,46,54), though they and others (51) caution that it is easy to make errors in this regard, and recommend monitoring for all groups suspected of having exposures when the worst case cannot be chosen with certainty.  A disadvantage of the worst case sampling scheme is that it limits risk assessment to one group (51), not gathering additional information to aid in finding exposure control methods or in establishing dose-response data for epidemiologic studies.   Guidelines for Antisapstain Exposure Assessment  51  In order to determine who to sample, Corn and Esmen (56) and Hawkins et al (54) have suggested that the initial step is to categorize workers into strata expected to have similar exposures based on jobs, processes, and/or work site conditions.  Others (47,57) have argued, in the context of monitoring for compliance with exposure regulations, that workers should not be grouped for sampling unless previous data show that mean exposures of individuals within a stratum do not vary by a factor of more than 2 to 4.  This approach requires that an initial sampling campaign be conducted in order to establish the groups. In either approach, grouping workers (stratified sampling) increases the precision of the estimated mean exposure, and gives additional information on each of the groups.   Selection of individuals within groups for sampling should be random in order to permit data interpretation using standard statistical methods (45-47,54).  The practical outcome of random selection of individuals is that the probability that a particular pattern of exposure will be sampled is proportional to the frequency with which it occurs, achieving the representativeness desired.  6.4.4  When to Sample   The question of when to sample must allow for sampling representative of the time distribution of an individual's exposures.  For example, exposures may vary with time of day, day of the week, and season. Since it is often most convenient to measure exposures over several consecutive days, it is important to consider whether samples taken in such a limited time frame are likely to be correlated and thus underestimate variability.  Some investigators (49,58) have attempted to address this issue.  Francis et al (58) found little evidence of autocorrelation in an analysis of several within-week data sets from industrial work sites, however Buringh and Lanting (49) noted that the variances of measurements taken within a week were lower than those taken over a longer period.  As with the selection of who to sample, the best method to ensure that sampling is representative of time variations in exposure is to select shifts randomly over the period of interest (45,54).  It may be possible to stratify based on time parameters known to influence antisapstain exposures, such as season.   However, because the effect of time factors such as shift, day of the week, and production schedules is unknown, all shifts should be included in the sampling frame for random sampling.  6.4.5  Where to Sample   Most of the occupational hygiene sampling strategy literature addresses the issue of how to ensure compliance with exposure standards (46,54,59), and makes the implicit assumption that sampling is being conducted within an individual work site.  For these generic exposure assessment guidelines, the major purpose of sampling is to estimate the mean and variance of exposure levels in the antisapstain  Guidelines for Antisapstain Exposure Assessment  52 treatment population to provide data for risk assessments.  This therefore requires that decisions be made about how to select representative lumber mills or shipping terminals for sampling, as well as workers and work shifts.   Based on survey sampling theory (60), there are three main ways in which such a selection could be done:  cluster sampling;  stratified random sampling;  or simple random sampling.  Deciding between these strategies requires consideration of the state of knowledge about differences in exposure between the various sites where antisapstain treatments are applied.   As outlined in the review of the available data on exposure, the variability of exposure measurements among workers within the same work site is very large (11,13,14,18,19).  However, the variability between different work sites has not been quantitatively explored in the research to date. Based on qualitative research, such as our descriptive survey of 14 B.C. sites currently using antisapstain treatment, one might expect that the variability between work sites is very great as well, because of differences in treatment systems, their locations in the production process, work assignments, the production process itself, and so on.   Cluster sampling is a technique which allows random selection of clusters (in this situation each work site would be considered a cluster), followed by sampling only in those clusters selected.  It has the advantage of reducing the cost of sampling by limiting the geographic area included in the sample. Unfortunately, cluster sampling is appropriate only if the clusters (i.e., work sites) themselves are random groupings, with large variability within themselves, but relatively little variation between each other.  This is unlikely to be the case with antisapstain work sites.  The consequence of inappropriate cluster sampling, especially when selecting small numbers of large clusters like entire work sites, is the possibility of missing work sites with a particular type of exposure.  Therefore, despite the convenience of doing so, we recommend against strategies which choose work sites and then select workers as they occur within the sample of work sites.  Though not stated explicitly, elements of this sampling strategy are implied in EPA's Subdivision U (8) which recommends that a minimum of 3 work sites be selected for pesticide monitoring. A difference is that Subdivision U does not require that the work sites be selected at random.   Stratified sampling is another method which may increase the efficiency of sampling.  In selecting where to sample, this strategy would involve first dividing work sites into strata according to estimated level of worker exposure, and then randomly selecting work sites within each stratum.  This strategy is not advised either because not enough is known about the factors which determine differences in exposure levels between work sites to recommend any particular stratification system.  Furthermore, the technology of antisapstain lumber treatment, indeed of most aspects of the lumber industry worldwide, is in a state of  Guidelines for Antisapstain Exposure Assessment  53 rapid evolution, so any stratification design even if it were appropriate now would likely be inappropriate in the near future.   Therefore until additional information about differences in exposures between work sites is known for this industry, it appears that the best means of achieving a representative sample of workers is to select a random sample from the complete pool of all workers from all sites combined.  6.4.6  How Many Samples to Take   Several authors have addressed the issue of how many samples are needed to best estimate exposure within a specified group (8,9,46,47,49,54).   In the pesticide regulation literature, EPA's Subdivision U (8) recommends that at least 3-5 different workers be sampled for each task in pesticide application from each of 3 different work sites.  In Subdivision K (9), the EPA suggests that sampling a minimum of 10 workers for re-entry exposure potential should be "sufficient for statistical validation of exposure", though they do not indicate how this sample size was derived.  In the industrial hygiene literature, Leidel et al (46) recommended a sampling scheme to ensure that at least one of the most highly exposed workers would be selected with a given degree of confidence.  This approach, as well as tests of means and tolerance limits (47,54), have been suggested where the objective of sampling is to ensure compliance with government exposure standards.   However, since the data collected in these guidelines will be used for risk assessments mainly addressing the outcomes of long-term exposures, an adequate sample size is required to allow estimation of the mean exposure over time and its variance.  Hawkins et al (54) have suggested that mean exposures may be reliably estimated with 6-11 samples, with diminishing improvements in precision as increasing numbers of samples are taken.  However, they suggest that additional samples (20 or more) are required to reliably estimate the variance of the sampling distribution and estimate confidence limits around the mean.  A modeling exercise conducted by Buringh and Lanting (49) demonstrates that exposure variance will likely be underestimated if too few samples are taken, especially if samples are taken on consecutive days during which exposure conditions are relatively stable.  They recommend that at least 30 samples be taken over a period of roughly two years, or where such a strategy is impractical, that a high estimate of the sample variance be assumed.   For these exposure assessment guidelines, we recommend that 30 workers be selected per group, to allow estimation of the mean, variance, and percentiles of the exposure distribution.  For example, to protect most of the workers, it would be useful to know the level of exposure of those workers at the 95th percentile of the group.  An additional consideration  is that for chronic hazards, the shift-to-shift  Guidelines for Antisapstain Exposure Assessment  54 variation is less important than a worker's mean exposure over time.  Therefore, for long-term hazards, the shift-to-shift variation should be removed from the total variance of the exposure distribution.  To separately estimate shift-to-shift variation, we recommend that each of the 30 workers have their exposure measured on two randomly selected occasions, for a total of 60 samples per worker group.   Guidelines for Antisapstain Exposure Assessment  55 6.5  Physical, Chemical, and Biological Properties of Antisapstain Chemicals   In order to design exposure assessment guidelines, it is important to know certain physical, chemical, and biological characteristics of the agent to be monitored.  The relevant properties are those which determine how the agent is able to enter the environment, whether it is toxic at points of contact, the extent to which it is absorbed across body membranes into systemic circulation, the degree to which the agent is accumulated in the body, and whether the agent has systemic effects.  Ideally, these properties should be known for the active ingredients, solvents, surfactants, solubilizing agents, and any other ingredients in the end use product of a fungicide to be monitored.  To develop generic exposure assessment guidelines, it is vital to know the range of these characteristics for the agents in question.  If, for example, none of the fungicides had dermal toxicity and all had negligible dermal flux, dermal exposure monitoring might not be necessary.  In the more likely case that data about these physical, chemical, and biological characteristics vary between compounds or is incomplete, exposure assessment plans must encompass all plausible means of exposure and routes of entry into the body.   Information about characteristics of the antisapstain agents was requested from the 17 companies funding the guidelines development;  data were received from 13 (61-75).  A list of the various active ingredients and formulated products is shown in Table 6 and is based on the information supplied by the contributing companies. As can be observed, a wide range of chemical agents are being considered for antisapstain use.  Table 7 lists some selected physical parameters for these agents:  molecular weight;  pH; water solubility;  octanol:water partition coefficient;  and vapour pressure.  Analytical methods are also included.  It is interesting to note that the fungicides cover almost the complete range of possibilities for the two parameters which have the most complete data, pH and water solubility.  Vapour pressure is also known for many of the agents;  in most cases it is very low, but several formulations have volatilities high enough to allow substantial evaporation.  Table 8 lists some comparative toxicological data based on acute animal experiments.  Many of the compounds are irritants, and therefore able to produce effects at the point of contact, suggesting that dermal and airborne exposure assessment will be necessary even if biological monitoring methods are developed.  An overview of both Tables 7 and 8 illustrates that for many of the parameters, data were not available.  This indicates that the generic exposure assessment guidelines must operate under the  conservative assumptions that all means and routes of exposure should be anticipated, as suggested above.   Properly designed exposure guidelines should be capable of responding to, not only those chemicals currently under consideration, but also new chemicals with properties yet to be determined.  Guidelines for Antisapstain Exposure Assessment  56   Guidelines for Antisapstain Exposure Assessment  57  Guidelines for Antisapstain Exposure Assessment  58  Guidelines for Antisapstain Exposure Assessment  59  Guidelines for Antisapstain Exposure Assessment  60  Guidelines for Antisapstain Exposure Assessment  61  Guidelines for Antisapstain Exposure Assessment  62 When the generic exposure guidelines are applied to evaluate the exposure associated with a particular agent it would be extremely helpful for the researcher to have a rough indication of the relative skin and lung absorption characteristics that can be expected from the agent under investigation. Such an assessment would not only help to identify specific areas of concern but also areas of secondary importance as far as exposure is concerned. A list of physical and chemical conditions and properties that have some effect on skin and lung absorption is shown in Table 9.  This list is based on the current understanding of the underlying physiology that determines the uptake of agents into these tissues. In order to appreciate the relevance of these factors, a brief review is warranted.  Table 9: Physical and chemical conditions and properties that affect skin and lung absorption  Skin Lung    lipid removal due to solvents particle size  hydration  octanol:water partition coefficient  concentration concentration  chemical reactivity association with particulate matter  octanol:water partition coefficient exposure frequency  molecular weight respiratory rate  exposure frequency chemical reactivity   6.5.1  Skin Absorption   The penetration of human skin by an applied chemical agent is referred to as dermal transport or percutaneous absorption.  Different types of exposures are possible such as intentional dermal application of drugs, accidental exposure, and sometimes, as in chemical warfare, deliberate exposure. The knowledge accumulated from these areas has been used to develop mathematical models of transdermal flux in order to achieve reliable estimates which can predict the behaviour of agents.  This information can then be used to set standards and consequently reduce cost and time in order to obtain critical information.   The skin is a multi-layer organ.  The outermost layer, the stratum corneum (SC), is comprised of several layers of keratinized dead cells with low water content.  These elongated cells are tightly interlocked giving mechanical strength to this layer.  The chemical inertness of the layer is a result of the cystine disulfide bonds of the matrix and membrane proteins surrounding the keratinized cells.  This layer is the primary barrier against chemical invasion and is approximately 10-20 µm thick.  It is replaced every 28 days by the action of the living cells of the epidermis which support it.  The epidermis is 40-100 µm thick. In addition, the epidermis is very active metabolically and is capable of not only detoxifying, but also increasing the toxicity of chemical agents prior to entering the vascular system of the underlying dermis. The dermis can function as a barrier and retain very lipophilic agents that are resistant to oxidative  Guidelines for Antisapstain Exposure Assessment  63 metabolism.  The dermis is supported by a layer of subcutaneous fat composed of a dense matrix of fibrous collagenous connective tissues embedded in a hydrous mass of mucopolysacharides.  There is some controversy over whether the blood vessels in the dermis act as a perfect transport system preventing the diffusion of chemicals beyond the vascular system.  Temperature can influence blood flow through the dermis.  Blood flow is controlled by sympathetic nervous fibers which regulate vasoconstriction and vasodilation, and consequently affect the rate of chemical clearance from the skin.  The dermis is approximately 3-5 mm thick and provides strength and elasticity to the skin.  Depending upon the part of the body, the physical variation of the skin can be large and may produce highly different rates of uptake of agents.  Dermal absorption of hydrocortisone, for instance, exhibits a 300-fold difference between the less permeable skin of the foot arch and the more permeable skin of the scrotum.   A variety of physical and chemical factors that control skin absorption have been identified and are briefly summarized below.  Lipid Removal due to Solvent Vehicles   Skin lipids are continuously produced by the sebaceous glands in the skin, resulting in a irregular surface film, 0.4-4 µm thick.  The surface of the skin is therefore water repelling or hydrophobic in nature. Of the lipids in the SC, 75% are neutral, including complex hydrocarbons, free sterols, sterolesters, free fatty acids, and tri-glycerides.  Twenty-five percent are polar lipids including phosphatidyl ethanol, phosphatidyl ethanolamine, phosphatidyl choline, lysolecithin, ceramides, and glycolcides.  Both groups of lipids in the SC present a major barrier to transepidermal water loss and are the limiting factors in percutaneous absorption (76).  The removal of the lipid layer from the SC surface, by means of washing with acetone or ether, apparently has no effect on the transepidermal water loss (77).  The layer does, however, act as a barrier to polar molecules.  Removal of the polar lipids from the SC with acetone or ether results in the loss of 30 to 40% of the lipid content, and if this is followed by extraction with hexane, which tends to remove the non-polar lipids, the entire water vapour barrier is lost.  Hydration   Hydration of the SC is important in percutaneous transport since water tends to act as a vehicle and a plasticizer.  Hydration is a slow process, and is still active after three days of continuous submersion (77).  In the occupational setting, hydration is induced by occlusion of the skin by clothing, gloves, wraps, or patches.  Occlusion also tends to prevent evaporation and consequently increases local temperatures and transport (77).   Guidelines for Antisapstain Exposure Assessment  64  Chemical Reactivity   Because the skin is composed of a variety of proteins, many different types of chemical reactions are possible (78).  Strong acids and bases will destroy the peptide bonds which link amino acids together to form proteins.  Weak acids and bases will modify proteins by ion pairing, as exemplified by the formation of Shiff bases as a result of formaldehyde exposure.  The most reactive groups in the skin are those associated with sulfhydro groups.  These can be cross-linked to form cystine disulfide bonds on exposure to agents such as bromine, hydrogen peroxide, and benzoyl peroxide.  Strong oxidants and cyanide ions can break these disulfide linkages.  Skin Metabolism   Although there is some enzymatic activity in the SC from the keratinized cells, from the sebum, and from the microbial growth on the skin, most metabolic activity resides in the proliferating cells of the epidermis (79).  The effect of these enzymes have been linked to a number of observations regarding toxicity, for instance, contact allergic dermatitis is the result of irreversible chemical binding between proteins and foreign compounds, i.e., acrylites, quinones, picryl chloride, which act like haptens and consequently elicit an immune response.  For this reason, the incidence of contact allergy, especially at the lower levels of exposure, appears to be better related to dose based on concentration per surface area rather than systemic dose (80).  Agents such as urushiol (the active ingredient in poison ivy) need to be metabolically activated (81).  Similarly, ethylparathion is readily absorbed by the skin and metabolically activated to ethylparoxen before it is transported by the vascular system (78).  Conversely, skin metabolism is responsible for the efficient degradation of aldicarb and other toxic N-methylcarbamates (78).  It appears that a long residence time in the epidermis will lead to enzyme induction and greater metabolism.   There have been many attempts to use measurable and calculable properties of agents in order to predict their transdermal transport.  The simplest is perhaps molecular weight.   Guidelines for Antisapstain Exposure Assessment  65  Molecular Weight   The upper boundary for percutaneous absorption has not been found and might, in special cases, be larger than the molecular weight of heparin (MW=17000 Daltons) (82).  The practical limit for most common agents however has been estimated to be between 300 and 400 Daltons (83).  Molecular weight by itself is not a good predictor of absorption behaviour because it is not unique and also implies a passive type of diffusion.  Because percutaneous absorption involves lipid as well as water barriers, there is a much better correlation between uptake and the octanol:water partition coefficient, P.  Octanol:Water Partition Coefficient   The factor P is defined as the ratio of the solubilities of an agent in octanol and in water.  For convenience, log P is often used for comparison.  A log P of 0 means equal distribution between the two phases.  A negative log P indicates high water solubility, whereas a positive log P indicates high lipid solubility.  It appears that most systemic effects are associated with compounds in the log P range of -0.5 to 3.5, with optimum percutaneous transport at around 2.0.  A log P below -.05 indicates entrapment of these molecules in the aqueous phase, and a log P greater than 3.5 indicates slow crossing of aqueous barriers.  Log P is considered to be a good descriptor for transport of molecules that do not specifically bind, and where organic and aqueous phases are involved.  There is a large data base of log P values which is used as a world wide standard (84).  It should not be forgotten that some immiscible phases encountered by a chemical when moving across the skin do not involve water and consequently the octanol:water coefficients do not always apply.  6.5.2  Lung Absorption  With respect to our current interest, there are two main areas in the lung.  The first is the tracheo- bronchial area which is covered with ciliated cells that form the basis of the tracheo-bronchial clearance system.  This clearance system is capable of removing a variety of particulate matter and mucus within 24 - 48 hours after exposure.  Clearance can be affected by a variety of agents including smoke and lead. The other region of interest is in the lower lung, the alveoli, where gas exchange takes place.  For efficient gas exchange, membranes which separate the blood from the alveolar spaces must be very thin.  In order for the alveoli to stay functional and remain inflated certain cells of the alveoli, the surfactant-secreting epithelial cells, release surface active agents which reduce the surface tension of the aqueous layer covering the surface of the alveoli.  Without these surfactants the alveoli would collapse under the pressure generated by the surface tension of water.  These surfactants are a mixture of phospholipids (dipalmitoyl lecithin) as well as proteins and ions such as calcium (85).  The proteins and calcium ions are  Guidelines for Antisapstain Exposure Assessment  66 required to allow the proper spread of the dipalmitoyl lecithin over the surface of the alveoli.  A reduction in surfactant synthesis has been observed after toxic exposures to agents such as gasoline vapours, trichloroethylene, carbon tetrachloride, cigarette smoke and paraquat.  There are a variety of factors that can modify the respiratory uptake of chemical agents by the lung, as described below.  Particle Size  The size of particles will influence the location of deposition in the lung and will control its metabolic fate.  Different parts of the respiratory tract differ in their abilities to metabolize chemicals. Consequently, when larger molecules end up in the naso-pharyngeal region, they will be subjected to an entirely different set of enzymatic activities than a particle small enough to reach the alveolar spaces of the lung (85).  Solubility  Lipid solubility, as expressed by the octanol:water partition coefficient, affects transport of agents across the alveolar membrane.  Results from experiments on polyaromatic hydrocarbons indicate that the more lipophilic the agent, the slower the long term clearance from the lung into the bloodstream (86). Predictions of clearance time can be made using log P.  At a log P of 6, for instance, lung clearance would take about sixty hours (87).  Other experiments indicate that greater lipophilicity enhances the solubility of an agent in lung cell membranes or in lung surfactant, decreasing the clearance time into the less lipophilic blood.  A knowledge of the solubility of the chemicals can therefore be an important criteria in making predictions regarding their biological fate.  Concentration   Factors such as the volatility of an agent and its method of application will affect its concentration in air.  The concentration of inhaled chemicals in turn affects lung absorption.  Studies on benzo(a)pyrene indicate a slow clearance at lower doses with half-lives greater than 1 day.  At higher doses, the system saturates and clearance is much more rapid.  Linear extrapolation from high doses to environmental concentrations must be done with caution and would, in this case, underestimate lung burdens for benzo(a)pyrene by a factor of 30 (88).  Association with Particulate Matter  Guidelines for Antisapstain Exposure Assessment  67   Studies have shown (88) a prolonged retention of benzo(a)pyrene in rodent lungs when it was combined with rust particles (Fe203).  The slower clearance was also associated with an increased incidence of lung cancer in these animals.  When benzo(a)pyrene is inhaled in the pure form, over 99% is readily cleared with a half-life of less than one day.  When the same agent is associated with a carrier particle, such as soot from diesel exhaust, clearance is much slower.  Exposure Frequency   Tissues which produce metabolically active enzymes (such as liver and skin), may produce  larger amounts in response to exposure.  For example, it has been shown that individuals who have had previous exposure to a variety of agents metabolize these agents much more readily than individuals with only a one time exposure (89).  Respiration Rate   In addition to physical and chemical factors, there are a variety of biological factors that also affect the exposure of the individual.  The rate of inhalation can vary between workers depending on the type of work done, and also varies between different species.  Smaller species generally have  higher metabolic rates, requiring more oxygen per cell and consequently will be exposed to larger amounts of chemicals under the same environmental conditions as larger species. These factors should be kept in mind, for instance, when attempting to relate animal toxicological data, as illustrated in Table 8, to humans.  6.5.3  Additives and Altered Formulation Products   Toxicological studies for risk assessment are often carried out on the active ingredient as well as the formulated mixture.  Exposure measurements are carried out on the end use mixture, monitoring for the active ingredient(s).  The possibility exists that the end use mixture has been modified, e.g., solvent vehicles or emulsifiers may be added to increase solubility.  Depending upon the degree of modification, it is possible that exposure data gathered on the unmodified formulation may no longer be applicable.  In this situation, a decision about whether exposure levels should be re-evaluated needs to be made by the regulatory agencies, keeping in mind the extent of the modification, the type and concentration of the agent(s) added, and any established exposure levels for those compounds.  Potential synergistic effects may also be considered based on the current understanding of the mechanisms of lung and dermal absorption.  Guidelines for Antisapstain Exposure Assessment  68   There is also a possibility that certain conditions of use may unintentionally result in exposures which are different from those normally associated with the end use mixture.  For example, compounds coming into contact with hot machinery components such as engine exhaust systems, brake drums, and brake linings of fork lift trucks could be pyrolytically modified, generating a wide spectrum of unknown agents.  Without identifying these agents, it would be impossible to evaluate their potential toxicity. Similarly, it has been observed that aerosols may mix with diesel exhaust gases creating another type of exposure about which little information is known.  In these situations, as a first option, means should be implemented to prevent or control the exposures.  If this is not possible, additional exposure and toxicological evaluations may be required.  6.5.4  Medical  Surveil lance   In order to protect the worker from potential toxicological effects associated with a new agent, allowable exposure levels are calculated using toxicological data from animal studies.  These incorporate a large margin of safety (usually 10 times for intra-specie variability and an additional 10 times for animal to human extrapolation).  Nevertheless, the ultimate data regarding the safety of an agent comes from human exposure.  It is therefore essential, when a new agent is introduced into the work place, that this is accompanied with careful medical surveillance and monitoring for potential health problems.  Guidelines for Antisapstain Exposure Assessment  69 7.  Acknowledgments   We would like to express our appreciation to the mill and shipping terminal employees for their co- operation and helpfulness during our site visits.  We would also like to thank the participating companies for supplying data about their antisapstain products, and all those who provided helpful comments in response to drafts of this document.   8.  References  1. Eriksson M.  1979.  A case-control study of malignant mesenchymal soft-tissue tumors and exposure to chemical substances.  Lakartidningen  76:3872-3875  2. Hardell L, Sandstrom A.  1979.  Case-control study:  Soft-tissue sarcomas and exposure to phenoxyacetic acids and chlorophenols.  Brit J Cancer  39:711-717  3. Hardell L, Eriksson M, Lenner P.  1980.  A case-control study: Malignant lymphoma and exposure to chemical substances, particularly organic solvents, chlorophenols, and phenoxy acids. Lakartidningen    77:208-210  4. Hardell L, Johansson B, Axelson O.  1982.  Epidemiological study of nasal and nasopharyngeal cancer and their relation to phenoxy acid or chlorophenol exposure.  Am J Ind Med  3:247-257.  5. Rappe C, Gara A, Buser HR.  1978.  Identification of polychlorinated dibenzofurans (PCDFs) in commercial chlorophenol formulations.  Chemosphere  12: 981-91.  6. Hagenmaier H, Brunner H.  1987.  Isomer specific Analysis of Pentachlorophenol and Sodium Pentachlorophenate for 2,3,7,8-Substituted PCDD and PCDF at Sub-ppb Levels.  Chemosphere 16(8-9):1759-64.  7. Agriculture Canada.  1984.  Registration Guidelines:  Guidelines for Registering Pesticides and Other Control Products Under the Pest Control Products Act in Canada.  Ottawa, Ontario.  8. U.S. Environmental Protection Agency.  1986.  Pesticide Assessment Guidelines, Subdivision U, Applicator Exposure Monitoring.  U.S. Department of Commerce National Technical Information Service, Washington, D.C.  9. U.S. Environmental Protection Agency.  1986.  Pesticide Assessment Guidelines, Subdivision K, Exposure: Reentry Protection.  U.S. Department of Commerce National Technical Information Service, Washington, D.C.  10. Fenske RA, Horstman SW, Bentley RK.  1987.  Assessment of dermal exposure to chlorophenols in timber mills.  Appl Ind Hyg  2:143-147  11. Kauppinen T, Lindroos L.  1985.  Chlorophenol exposure in sawmills.  Am Ind Hyg Assoc J  46:34- 38  12. Todd AS, Timble CV.  1983.  Industrial Hygiene Surveys of Occupational Exposure to Wood Preservative Chemicals.  DHHS (NIOSH) Pub. No. 83-106, Cincinnati, Ohio.   Guidelines for Antisapstain Exposure Assessment  70 13. Teschke K, Hertzman C, Dimich-Ward H, Ostry A, Blair J, Hershler R.  1989.  A comparison of exposure estimates by worker raters and industrial hygienists.  Scand J Work Environ Health 15:424-9  14. Hertzman C, Teschke K, Dimich-Ward H, Ostry A.  1988.  Validity and reliability of a method for retrospective evaluation of chlorophenate exposure in the lumber industry.  Am J Ind Med 14:703-13  15. Embree V, Enarson DA, Chan-Yeung M, DyBuncio A, Dennis R, Leach J.  1984.  Occupational exposure to chlorophenates:  Toxicology and respiratory effects.  Clin Toxicol  22(4):317-29  16. Lindroos L, Koskinen H, Mutanen P, Jarvisalo J.  1987.  Urinary chlorophenates in sawmill workers. Int Arch Occup Environ Health  59:463-467  17. Levin J-O, Rappe C, Nilsson C-A.  1976.  Use of chlorophenols as fungicides in sawmills.  Scand J Work Environ Health  2:71-81  18. Kleinman G, Horstman SW.  1984.  Exposure Assessment of Permatox 100.  Report to the Simpson Timber Company, Shelton, Washington.  19. McDonald E.  1989.  Documentation of Worker Exposure to Anti-sapstain Chemicals in British Columbia Wood Products Operations.  Report to the Council of Forest Industries of British Columbia, Vancouver, B.C.  20. Occupational Health Group, B.C. Research Corporation.  1989.  Industrial Hygiene Survey of Worker Exposure to TCMTB Anti-sapstain Solutions.  Report to the Forest Industry Health Research Program (FIHRP Project No. WP88/6), Vancouver, B.C.  21. Detrick RS.  1977.  Pentachlorophenol:  Possible sources of human exposure.  Forest Products J 27:13  22. Kalman DA, Horstman SW.  1983.  Persistence of tetrachlorophenol and pentachlorophenol in exposed woodworkers.  J Toxicol-Clin Toxicol  20:343-352  23. BC Research.  1986.  Analysis of Chlorophenols and Chlorophenol Metabolites in Urine Samples from Occupationally Exposed Persons (Sawmill Workers).  Prepared for Pesticides Division, Health and Welfare Canada.  24. Maloney PA, Curry PB, Lyengar S, Linke BJ, Worgan JP, Bell RDL.  1991.  Canadian pesticide exposure assessment guidelines.  ABSTRACT, International Conference on Measuring, Understanding and Predicting Exposures in the 21st Century, Nov 18-21, Atlanta, GA.  25. Gunther FA, and Gunther JD.  1980.  Minimizing occupational exposure to pesticides.  Residue Rev 75.  26. Honeycutt RC, Zweig G and Ragsdale NN.  1985.  Dermal Exposure Related to Pesticide Use, ACS Symposium Series 273.  27. Wang RG, Franklin CA, Honeycutt RC and Reinert JC.  1989.  Biological Monitoring for Pesticide Exposure, ACS Symposium Series 382, American Chemical Society, Washington DC.  28. Durham WF, Wolfe HR.  1962.  Measurement of exposure of workers to pesticides.  Bull World Health Org  26:75-91  29. Davis JE.  1980.  Minimizing occupational exposure to pesticides:  Personnel monitoring.  Residue Reviews  75:33-50   Guidelines for Antisapstain Exposure Assessment  71 30. Franklin CA, Fenske RA, Greenhalgh R, Mathieu L, Denley HV, Leffingwell JT, and Spear RC.  1981. Correlation of urinary pesticide metabolite excretion with estimated dermal contact in the course of occupational exposure to guthion.  J Toxicol Environ Health  7:715-731.  31. Fenske RA.  1989.  Validation of dermal exposure monitoring by biological monitoring:  the fluorescent tracer technique and the patch technique.  In:  Biological Monitoring for Pesticide Exposure, RG Wang, CA Franklin, RC Honeycutt and JC Reinert, eds., ACS Symposium Series 382:70-84, American Chemical Society, Washington DC.  32. Fenske RA.  1990.  Nonuniform dermal deposition patterns during occupational exposure to pesticides.  Arch Environ Contam Toxicol  19:332-337.  33. WHO.  1982.  World Health Organization:  Field Surveys of Exposure to Pesticides.  Standard Protocol, VBC/82.1  Division of Vector Biology and Control, Geneva.  34. Bonsall J.L.  1985.  Measurement of occupational exposure to pesticides, in Occupational Hazards of Pesticide Use.  GJ Turnbull, ed., Taylor and Francis, London.  35. Ross J, Thongsinthusak T, Fong HR, Margetich S, Krieger R.  1990.  Measuring potential dermal transfer of surface pesticide residue generated from indoor fogger use:  an interim report. Chemosphere  20:349-360.  36. Ross J, Fong HR, Thongsinthusak T, Margetich S, Krieger R.  1991.  Measuring potential dermal transfer of surface pesticide residue generated from indoor fogger use:  using the CDFA roller method, interim report II.  Chemosphere  22:975-984.  37. DeJonge JO, Easter E.  1989.  Laboratory Evaluation of Pesticide Spray Penetration and Thermal Comfort of Protective Apparel for Pesticide Application.  Final Report prepared for U.S. Environmental Protection Agency Office of Research and Development.  38. Fenske RA, Leffingwell JT, Spear RC.  1986.  A video imaging technique for assessing dermal exposure - I.  Instrument design and testing.  Am Ind Hyg Assoc J  47:764-770.  39. Fenske RA, Wong SM, Leffingwell JT and Spear RC.  1986.  A video imaging technique for assessing dermal exposure - II.  Fluorescent tracer testing.  Am Ind Hyg Assoc J  47:771-775.  40. Occupational Health Group, B.C. Research Corporation.  1987.  Use of Fluorescent Tracer to Study Skin Contact with Anti-sapstain Chemical.  Report to the Forest Industry Health Research Program (FIHRP Project No. WP/86-1), Vancouver, B.C.  41. Bentley R, Horstman SW.  Morgan MS.  1989.  Reduction of sawmill worker exposure to chlorophenols.  Appl  Ind Hyg 4(3):69-74.  42. ACGIH.  1992.  1991-1992 Threshold Limit Values for Chemical Substances and Physical  Agents. American Conference of Governmental Industrial Hygienists, Cincinnati, Ohio.  43. USEPA.  1983.  Good Laboratory Practice Standards.  40 CFR 53946.  Office of Pesticide Programs.  U.S. Environmental Protection Agency.  November 29.  44. Corn M.  1985.  Strategies of air sampling.  Scand J Work Environ Health  11:173-80  45. Oldham PD, Roach SA.  1952.  A sampling procedure for measuring industrial dust exposure.  Brit J Ind Med  9:112-9  46. Leidel NA, Busch KA, Lynch JR.  1977.  Occupational Exposure Sampling Strategy Manual.  DHEW (NIOSH) Pub. No. 77-173, Cincinnati, Ohio.   Guidelines for Antisapstain Exposure Assessment  72 47. Rappaport SM.  1991.  Assessment of long-term exposures to toxic substances in air.  Ann Occup Hyg  35:61-121  48. Roach SA.  1977.  A most rational basis for air sampling programmes.  Ann Occup Hyg  20:65-84  49. Buringh E, Lanting R.  1991.  Exposure variability in the workplace:  Its implications for the assessment of compliance.  52(1):6-13  50. Nicas M, Simmons BP, Spear RC.  1991.  Environmental versus analytical variability in exposure measurements.  Am Ind Hyg Assoc J  52(12):553-7  51. Rappaport SM, Selvin S, Roach SA.  1988.  A strategy for assessing exposures with reference to multiple limits.  Appl Ind Hyg  3(11):310-5  52. Saltzman B.  1970.  Significance of sampling time in air monitoring.  J Air Pollution Control Assoc 20(10):660-5  53. Spear RC, Selvin S, Francis M.  1986.  The influence of averaging time on the distribution of exposures.  Am Ind Hyg Assoc J  47(6):365-8  54. Hawkins NC, Norwood SK, Rock JC.  1991.  A Strategy for Occupational Exposure Assessment. Akonr, Ohio:  American Industrial Hygiene Association.  55. Roach SA.  1966.  A more rational basis for air-sampling programs.  Am Ind Hyg Assoc J  27:1-12  56. Corn M, Esmen NA.  1979.  Workplace exposure zones for classification of employee exposures to physical and chemical agents.  Am Ind Hyg Assoc J  40:47-57  57. Health and Safety Executive.  1989.  Monitoring strategies for toxic substances.  Guidance Note EH 42.  United Kingdom:  HMSO  58. Francis M, Selvin S, Spear RC, Rappaport SM.  1989.  The effect of autocorrelation on the estimation of workers' daily exposures.  Am Ind Hyg Assoc J  50:37-43  59. Baxter RA, Bonthrone W, Drope E, Jourdan L, Lohwasser H, Sanderson JT, Sartre B, Tassignon JP, De Voogd P.  1991. CEFIC Report on Occupational Exposure Limits and Monitoring Strategy. European Council of Chemical Manufacturers' Federation, Brussels, Belgium.  60. Scheaffer RL, Mendenhall W, Ott L.  1990.  Elementary Survey Sampling.  PWS-Kent Publishing Company, Boston.  61. TCMTB Sapstain Control Information Package.  1989.  Buckman Laboratories.  Montreal Quebec.  62. Fermenta, Plant Protection.  1988.  Material Safety Data Sheet.  63. Woodgard EC.  1988. Sapstain Control Product No. 9979-9006.  Sadolin Paint Products 3950 New Walkertown Road, N.C.  64. Material Safety Data Sheet.  1989.  HH890 8002 Hoechst Holland N.V.  65. Material Safety Data Sheet.  1987.  Sinesto B. F.C.  Finnish Chemicals OY.  66. Bravo, Daconil 2787.  1990. Broad-Spectrum Fungicide ISK Biotech Corporation 5960 Heisley Road PO Box 8000, Mentor, Ohio 44061-8000.  67. Bardac 2250 Bardac 2280.  1988.  Product Information Lonza Inc. Fair Lawn NJ.   Guidelines for Antisapstain Exposure Assessment  73 68. Rosebery, GP. 1991.  Summary of the physical-chemical properties and the toxicology of Oxine, Copper (Copper 8-Hydroxyquinolate) Maag Agrochemicals Inc. PO Box 6430 Vera Beach, Florida.  69. Material Safety Data Sheet:  Nytek GD (ACR3799B).  1990.  Maag Agrochemicals Inc. Vero Beach, Florida.  70. Wood Protection with Mitrol PQ-8.  1988.  Recommendations for Design and Operation.  Chapman Chemical Company Memphis TN.  71. Material Safety Data Sheet:  Sta-Brite, 1991.  Material Safety Data Sheet:  Mitrol PQ-8 1991. Chapman Chemical Company, Memphis TN.  72. Safety Data sheet. Tab 2.  August 1990.  Xylochimie, Boulevard de Courbevoie Neuilly-Sur-Seine Cedex.  73. Preventol O Extra/ON Extra, Preventol OF, Preventol A8.  1990.  Product information. Bayer Organica  74. Technical Bulletin, Troysan Polyphase. 1982.  Anti Mildew.  Troy Chemical Corporation.  Newark NJ.  75. Material Safety Data Sheet.  1991.  NP-1.  Kop-Coat Incorporated.  Pittsburgh, PA.  76. Mencze E, Bucks DAA, Wester RC, Maibach HI.  1985.  Skin binding during percutaneous penetration.  In Percutaneous absorption, eds. RL Bronaugh and HI Maibach, Chapter 3, New York: Marcel Dekker.  77. Scheuplein RJ, Blank IH.  1971.  Permeability of the skin.  Physio. Rev. 51:702.  78. Marzulli FN.  Dermato-toxicology, Fourth Edition.  Hemisphere Publishing Corporation.  79. Stuttgen G.  1982.  Drug absorption by intact and damaged skin.  In Dermal and transdermal absorption, eds.  R Brandau and BH Lippold, Chapter 2.  Stugart: Wissenschaftliche Verlagsgesellechaft.  80.  Jayjock MA, Lewis PG. 1992.  Low-Applied-Dose extrapolation of induction and elicitation of contact allergy in the evaluation and management of sensitization risk from Kathon CG isothiazolone in products.  American Journal of Contact Dermatitis.  Vol. 3;2:86-91.  81. Dupuis G.  1979.  Studies on poison ivy.  In vitro lymphocyte transformation by urushiol-protein conjugates.  Brit J Dermatol  101:617.  82. Wester RC, Maibach HI.  1985.  Structure-activity correlations in percutaneous absorption.  In Percutaneous absorption, eds.  RL Bronaugh and HI Maibach, Chapter 8, New York:  Marcel Dekker.  83. Wester RC, Maibach HI.  1985.  Influence of hydration on percutaneous absorption.  In Percutaneous absorption, eds.  RL Bronaugh and HI Maibach, Chapter 18, New York:  Marcel Dekker.  84. Hansch C, Leo A.  1979.  Substituent constants for correlation analysis in chemistry and biology, Chapter 4.  New York: John Wiley and Sons.  85. Guyton AC.  1991.  Textbook of Medical Physiology.  8th edition.  WB Saunders Company. Philadelphia.  86. Bond JA, Baker SM, Bechtold WE.  1985.  Correlation of the octanol/water partition coefficient with clearance half-times of intratracheally instilled aromatic hydrocarbons in rats.  Toxicology 36:285-295.  Guidelines for Antisapstain Exposure Assessment  74  87. Bond JA, Sun JD, Mitchell CE, Dutcher JS, Wolff RK. McClellan RO.  1986.  Biological fate of inhaled organic compounds associated with particulate matter.  In: Aerosols, edited by SD Lee, T Schreider, LD Grant, PH Verkerk, pp., 579-592.  Chelsea, MI:Lewis Publishers  88. Bond JA.  1988.  Factors Modifying the Disposition of Inhaled Organic Compounds.  In:  Concepts in Inhalation Toxicology, McClellan RE, Henderson RF, editors.  Hemisphere Publishing Corp, New York.  pp 249-70.  89. McKenna MJ, Zempel JA, Braun WH.  1982.  The pharmacokinetics of inhaled methylene chloride rats.  Toxicol Appl Pharmacol 65:1-10.    Guidelines for Antisapstain Exposure Assessment  75 Appendices    Appendix A:  Data Analysis Recommendations .................................................................................... A1   A.1 Introduction ..................................................................................................................... A1  A.2 Notation ........................................................................................................................... A2  A.3 The Statistical Model ....................................................................................................... A3  A.4 Method A ......................................................................................................................... A4   A.4.1   Adjustment for Absences and Job Changes...................................................... A6  A.5 Method B.......................................................................................................................... A7  A.6 Method C ......................................................................................................................... A9  A.7 Example............................................................................................................................ A9   Exhibit A: Method A Results - Stratum Logarithmic Means, and Stratum ................................... A11    Logarithmic Variance Components  Table A: Estimated proportion of workers in 5 exposure strata exceeding various ................. A13   concentrations, Method A (exact)  Exhibit B: Method B Results - Stratum Logarithmic Means and Variances .................................. A14  Table B: Estimated proportion of workers in 5 exposure strata exceeding various ................. A15   concentrations, Method B (conservative)  Exhibit C: Method C Results - Stratum Logarithmic Means and Variances.................................. A16  Table C: Estimated proportion of workers in 5 exposure strata exceeding various ................. A17   concentrations, Method C (conservative)   Appendix B: List of Companies Supporting the Generic Exposure Protocol Project  and Letter to Companies  Appendix C:   Letter to Mills and Shipping Terminals  Appendix D:   Site Visit Data Collection Form  Appendix E:   Job Tasks Associated with Different Treatment Methods  Appendix F: Responses of Canadian Federal Government Agencies to Draft Guidelines  Appendix G: Some Notes and Correspondence on the Idea of a "Surrogate Exposure Study"    Guidelines for Antisapstain Exposure Assessment  76       Appendix A:  Data Analysis Recommendations    Guidelines for Antisapstain Exposure Assessment  A1 Appendix A:  Data Analysis Recommendations  A.1  Introduction  In this appendix, we describe how to analyze the data that would be generated by a worker exposure study designed in accordance with our guidelines.  We assume that a population of workers has been divided into several strata of relatively homogeneous exposure, that a number of workers have been randomly sampled from each stratum, and that one or more shifts have been randomly sampled for each worker.  For each of the sampled shifts, exposure concentrations representing arithmetic mean exposures over a period of one shift have been measured.  The goal of the analysis is to be able to estimate, for any given mean annual exposure concentration, what proportion of the population, within each stratum and globally, will exceed that exposure concentration. The results of the analysis might, for example, be presented in a table with seven columns, the first column listing a series of exposure concentrations of interest, columns 2 to 6 containing the estimated percentage of the population within each of the five strata that exceed those exposure concentrations, and the final column containing the estimated percentage of the combined study population that exceeds the given exposure concentrations.  (See Table A for an example.)  Three methods are described: Method A, a complete analysis which is applicable when exposure concentration measurements for more than one shift per person are available; Method B, an approximate method which produces conservative results and which is applicable when exposure concentrations for only one shift per person have been measured; and Method C, an even more conservative method which is applicable when only a few workers (e.g., 10) per stratum have been sampled rather than the recommended 30.  If more than one shift has been assessed for most individuals, but there are a few with missing values, then method A (the complete analysis) is still applicable.  All three methods are based on the same underlying statistical model of exposure, namely a mixed model that includes both fixed and random effects.    Guidelines for Antisapstain Exposure Assessment  A2 A.2  Notation  The following symbols are used in this appendix:  yijk : the actual concentration measurement for the i th shift of the jth person in the kth stratum  l.jk : the mean logarithmic exposure concentration for the j th person in the kth stratum  l..k : the mean logarithmic exposure concentration for the k th stratum  a.jk : the arithmetic mean exposure concentration for the j th person in the kth stratum  a..k : the arithmetic mean exposure concentration for the k th stratum  x : an arbitrary arithmetic mean annual concentration for an individual, used as a reference level  σSk 2: the shift-to-shift variance in logarithmic exposure concentration in the kth stratum  σPk 2: the person-to-person variance in logarithmic exposure concentration in the kth stratum  σk 2: σSk 2 + σPk 2  lx : ln ( x  ) – .5σP 2   (see explanation on page A5)  αijk ,βjk : independent values from a standard normal distribution  fk  : correction factor for job change  tk  : correction term for absence  Guidelines for Antisapstain Exposure Assessment  A3 A.3  The Statistical Model  We assume that exposure concentration follows a log normal distribution within each stratum.  However, the quantity which best predicts the biological effect of exposure is the arithmetic mean exposure concentration.  Hence, it is necessary to translate back and forth between logarithmic means and arithmetic means.  The data do not contain replicate measurements for the same shift, and therefore, the exposure concentration measurement error will be confounded with the shift-to-shift variation.  To deal with this, we assume that exposure concentration measurement error is small compared to shift-to-shift variation, and so can be ignored.  We make no assumption that shift-to-shift variance and person-to-person variance are constant across the different strata.  However, we do assume that shift-to-shift variance for different workers in the same stratum is constant.  The model can then be written:    i) ln( yijk )  =  l..k  + σPk  βjk + σSk  αijk   The stratum effect is thus fixed, whereas the person and shift effects are random. It follows that:    ii) a.jk  =  exp { l.jk  + .5 σSk 2 }  and   iii) a..k  =  exp { l..k  + .5 (σPk 2 + σSk 2) }    Guidelines for Antisapstain Exposure Assessment  A4 A.4  Method A  The parameters of the model, the 5 logarithmic means corresponding to the 5 strata and the ten variance components (5 person-to-person and 5 shift-to-shift), are estimated by the restricted maximum likelihood (REML) method.  Special computer software that performs REML estimation is required.  For example, the SAS procedure PROC VARCOMP or the BMDP procedure P3V can be used.  Here we will illustrate the use of PROC VARCOMP.  Suppose the data is arranged in a text file "EXPOSURE.TXT" with each line corresponding to one exposure concentration measurement, and five columns (separated by blanks) giving values for Y ( i.e., yijk , the exposure concentration measurements), STRATUM, PERSON, SHIFT, and VALID.  VALID is a special flag variable which is coded as 1 if the exposure measurement is ordinary, and 0 if exceptional.  (Exceptional values are those corresponding to absences or job changes and are recoded to missing for this part of the analysis.)  Then the following program can be used:  DATA EXPOSURE;   INFILE 'EXPOSURE.TXT';   INPUT Y STRATUM PERSON SHIFT VALID;   IF VALID=0 THEN Y=.;   LY = LOG(Y);   /* Natural logarithm */  PROC SORT;   BY STRATUM; /* Sort data by the STRATUM variable */  PROC MEANS N MEAN;   BY STRATUM;   VAR LY;  PROC VARCOMP METHOD=REML;   BY STRATUM;   CLASS PERSON;   MODEL LY =  PERSON; RUN;   This program produces estimates of the logarithmic mean exposures and variance components by  Guidelines for Antisapstain Exposure Assessment  A5 stratum.  What we have called σPk 2 will be labeled Var(PERSON) in the output.  What we have called σSk 2 will be labeled Var(Error).  The output will also include standard errors for those estimates (actually the covariance matrix for the estimates, from which standard errors are easily derived as the square root of the elements on the diagonal).  Once the parameters of the model are estimated, the arithmetic mean exposure concentration for each stratum is calculated from formula iii).  The arithmetic mean exposure concentration for the whole population can be estimated by taking the weighted average of the stratum specific arithmetic means (the weights being the actual population sizes of each stratum).  Given any individual mean exposure concentration x, the proportion of the population in stratum k which exceeds that exposure concentration is estimated as follows.  First calculate a logarithmic exposure concentration lx by solving:    iv) x   =  exp ( lx + .5 σSk 2 )  i.e.,   v) lx  =  ln ( x  ) – .5 σSk 2   Then, calculate a standardized version of lx according to the formula:    vi)  zk  =  ( lx  – l..k  ) / σPk  i.e.,   vii) zk  =  { ln( x ) – .5 σSk 2 – l..k  } / σPk    Guidelines for Antisapstain Exposure Assessment  A6 The proportion of workers in stratum k whose exposure exceeds x is found by looking up zk in tables of the normal distribution.  The proportion is equal to the area of the right tail beyond zk.  The proportion of the entire population exceeding the exposure concentration is the weighted average of the  proportions for each stratum, using the stratum population sizes as weights.  PROC VARCOMP may occasionally produce an estimate of zero for one of the variance components. This is particularly a problem if the estimate of σPk 2 is zero since formula vii) cannot be used.  A simple solution is to use the approximate formula xi) instead, on the plausible assumption that σPk 2 is not truly zero even if it is close to zero.  A.4.1  Adjustment for Absences and Job Changes  The assumption that shift-to-shift variation for a given person follows a log normal distribution is reasonable if that person remains in the same job and is not absent during any shifts.  The analysis described so far is only valid in the ideal circumstances that individuals never change jobs and have no absences.  To account for absences and job changes, adjustments are required.  An adjustment factor fk is calculated for each stratum to account for job changes, and an adjustment term tk, is calculated for each stratum to account for absences.  The adjustment for absences assumes that absence is sufficiently uncommon that the probability of a worker being absent for both of two randomly selected shifts is negligible.  Let Ak  be the set of workers in stratum k  who are absent at one of the two assessments (but not both), let Ck be the set of workers in stratum k who have changed jobs by the second assessment, and let nk be the total number of individuals in stratum k.  Then define:   viii) tk   = 1 nk   ∑  jεAk   | |y1jk  – y2jk  and   ix) fk   =  exp        1 nk   ∑  jεCk   ( )ln( y2jk ) – ln( y1jk )   Guidelines for Antisapstain Exposure Assessment  A7 (Note that in formula viii) either y1jk  or y2jk  will be zero.) For an annual mean exposure concentration, x, of interest, first calculate an adjusted mean exposure concentration x*:    x) x*  =  x fk    +  tk   Then carry out the same calculation of frequency of excessive exposure as before, using x* in place of x.  A value of fk  greater than 1 indicates added exposure due to job change.  In this case x* will be smaller than x, and the adjustment will result in a higher estimate of the frequency with which the exposure concentration is exceeded.  A value of tk  other than zero indicates reduced exposure due to absence.  In this case x* will be larger than x, and the adjustment will result in a lower estimate of the frequency with which the exposure concentration is exceeded.   A.5  Method B  In method A, both components of variation occurred separately in the calculation of zk, requiring a separate estimate of each (and, therefore, more than one measurement per worker).  However, it can be shown by simple algebra that formula vii) for zk can be rewritten:     .5 ( zk 2 – ( zk – σPk ) 2 ) = ln ( x  ) – .5 ( σSk 2 + σPk 2) – l..k  Since ( zk – σPk   ) 2 is always positive, it follows that    zk 2 ≥  2 { ln ( x  ) - .5 ( σSk 2 + σPk 2) - l..k  }   If we replace the inequality with equality, then we underestimate zk and hence overestimate the proportion exposed.  The advantage of this approximate method is that only the sum of the two  Guidelines for Antisapstain Exposure Assessment  A8 components of variation is required, rather than separate estimates of each.  Method B is, therefore, applicable where each worker's exposure has been measured on only one shift. The data can therefore be analyzed by elementary calculations on the natural logarithmic transform of the shift exposure concentration measurements.  The l..k  are the stratum logarithmic means, and the σk 2  =  σSk 2 + σPk 2 are estimated as the stratum logarithmic variances.  For example, the following SAS program could be used:  DATA EXPOSURE;   INFILE 'EXPOSURE.TXT';   INPUT Y STRATUM PERSON VALID;   IF VALID = 0 THEN Y=.;   LY = LOG(Y);   /* Natural logarithm */  PROC MEANS N MEAN VAR;   BY STRATUM;   VAR LY; RUN;  Then, for any mean annual exposure concentration of interest, x, calculate:    xi) zk  =   2 { ln ( x  ) – .5 σk 2 – l..k  }   The proportion with exposure concentration exceeding x is obtained by looking up zk in a tables of the normal distribution.  The proportion is equal to the area of the right tail beyond zk .  (If the quantity under the square root sign in equation xi) is less than zero, then the conservative estimate of the proportion exceeding that exposure level is 1.)  The proportion of the entire population exceeding the exposure concentration is the weighted average of the stratum proportions, using the stratum population sizes as weights.  The arithmetic mean exposure concentration for each stratum can be calculated as before using equation iii).  The arithmetic mean exposure for the strata combined is the weighted average of the stratum specific arithmetic means.   Guidelines for Antisapstain Exposure Assessment  A9 An adjustment for absences and job changes is not possible with this method unless some additional information is available on absence and job change. A.6  Method C  Method C is a variation of Method B suitable for use when the number of workers per stratum is too small (roughly, less than 30) to get a reliable estimate of the variance for that stratum.  In particular, this method applies to the limited exposure assessment described as the Pre-registration Option in Section 2.9 of the Guidelines.  Note that in the Pre-registration Option, not only is the sample small (10 per stratum), but the study is conducted over a relatively short period of time and so does not sample the full range of variation that would normally occur over a full year.  This method is identical to method B except that a predetermined conservative estimate of σk 2 is employed to account for the fact that the sample is not large enough and not representative enough to produce a reliable and fair estimate of σk 2.  The conservative estimate of σk 2 that we recommend is 3.21, which translates to a geometric standard deviation of 6.  This estimate was selected because data from urinary chlorophenate studies in sawmills (13,14) had total geometric standard deviations of between 3 and 4;  and data from 31 industrial work groups reported by Rappaport (47) showed between-person geometric standard deviations of greater than 6 in three groups.  Thus a geometric standard deviation of 6 is expected to be both high and therefore conservative, but not outside the realm of expectation.  To calculate the proportion of workers in each stratum exceeding various concentrations, the same program as for method B is used, except that we ignore the variance estimates coming from the data and replace them with 3.21.   A.7  Example  Data conforming to the model given in equation i) were generated using a random number generator.  A total of 300 observations were generated representing 2 measurements per person, 30 persons per stratum, and 5 strata.  The five strata logarithmic mean concentrations were assumed to be 2.80, 2.80, 2.30, 1.30, and -0.200 respectively.  σSk 2 was assumed to be 1.21 (corresponding to a geometric standard deviation of 3), and σPk 2 was assumed to be 0.164 (corresponding to a geometric standard deviation of 1.5).   Guidelines for Antisapstain Exposure Assessment  A10 Exhibit A shows the output generated by running the program given in Section A.4 (method A) on the simulated data.  It is seen that the stratum logarithmic means estimated from the data are:  3.11, 2.82, 2.52, 1.11, and -0.033, agreeing reasonably well with the true values.  Also it is seen that the variance component estimates for σSk 2 range from 0.721 to 1.28.  Those for σPk 2 range from 0.113 to 0.586. Table A show the estimated proportion of the worker population that exceeds various exposure concentrations based on the results in Exhibit A.  For convenience the logarithmic means have been converted to geometric means by exponentiating the logarithmic means, and the logarithmic variances have been converted to geometric standard deviations by taking the square root and then exponentiating.  Exhibit B shows the output generated by running the program in Section A.5 (method B) after deleting one half of the data (i.e., after deleting the second exposure concentration measurement for each person). The estimates of the stratum logarithmic means and variances are still reasonably close to the true values. Table B shows the estimated proportion of the worker population that exceeds various exposure concentrations based on the results in Exhibit B.  The estimated proportions exceeding each exposure level are uniformly higher than with method A.  Notice, however, that if one seeks the exposure level which ensures that less than 5% of the population will exceed that level, it is about 70 according to method A and about 90 according to method B.  In this instance, there is little practical difference between the two methods.  Exhibit C shows the output generated by restricting the data to the first 10 workers per stratum, and one shift per worker, and again running the program in section A.5.   Finally, Table C shows the result of estimating the proportion of workers exceeding various reference concentrations using method C and the output in Exhibit C.  It is seen that in order to ensure that less than 5% of the population is exposed one now has to go to an exposure level of about 300, considerably higher than before.  Guidelines for Antisapstain Exposure Assessment  A11 Exhibit A:  Method A Results  1. Stratum Logarithmic Means    ---------------------------------- STRATUM=1 -------------------------------- -                                   N          Mean                               -----------------                                60     3.1138618                               -----------------  ---------------------------------- STRATUM=2 -------------------------------- -                                   N          Mean                               -----------------                                60     2.8281871                               -----------------  ---------------------------------- STRATUM=3 -------------------------------- -                                   N          Mean                               -----------------                                60     2.5250107                               -----------------  ---------------------------------- STRATUM=4 -------------------------------- -                                   N          Mean                               -----------------                                60     1.1138319                               -----------------  ---------------------------------- STRATUM=5 -------------------------------- -                                   N          Mean                               -----------------                                60    -0.0336083                               -----------------  Guidelines for Antisapstain Exposure Assessment  A12 Exhibit A (con't):  Method A Results  2. Stratum Logarithmic Variance Components    --------------------------------- STRATUM=1 --------------------------------- -           Iteration        Objective      Var(PERSON)       Var(Error)               0          15.02008193       0.11312579       1.18364822              1          15.02008193       0.11312579       1.18364822  --------------------------------- STRATUM=2 --------------------------------- -           Iteration        Objective      Var(PERSON)       Var(Error)               0          21.80379752       0.48050108       1.05185591              1          21.80379752       0.48050108       1.05185591  --------------------------------- STRATUM=3 --------------------------------- -           Iteration        Objective      Var(PERSON)       Var(Error)               0           8.69696869       0.58646204       0.72081690              1           8.69696869       0.58646204       0.72081690  --------------------------------- STRATUM=4 --------------------------------- -           Iteration        Objective      Var(PERSON)       Var(Error)               0          32.82801578       0.58030346       1.26722980              1          32.82801578       0.58030346       1.26722980  --------------------------------- STRATUM=5 --------------------------------- -           Iteration        Objective      Var(PERSON)       Var(Error)               0          22.37554393       0.20298647       1.27562171              1          22.37554393       0.20298647       1.27562171   Guidelines for Antisapstain Exposure Assessment  A13 Table A:  Estimated proportion of workers in 5 exposure strata exceeding various concentrations   Method A (exact):   Stratum All Strata  Concentration  1  2 3 4 5 Combined    10.0 1.000 0.935 0.779 0.232 0.000 0.499  30.0 0.813 0.473 0.252 0.015 0.000 0.241  50.0 0.264 0.211 0.091 0.002 0.000 0.089  70.0 0.051 0.099 0.038 0.000 0.000 0.031  90.0 0.009 0.049 0.018 0.000 0.000 0.013  110.0 0.001 0.026 0.009 0.000 0.000 0.006  130.0 0.000 0.015 0.005 0.000 0.000 0.003  150.0 0.000 0.008 0.003 0.000 0.000 0.002  170.0 0.000 0.005 0.002 0.000 0.000 0.001  190.0 0.000 0.003 0.001 0.000 0.000 0.000    Based on method A output:   1) Stratum populations as percentage of total:   13.33 16.67 20.00 23.33 26.67  2) Geometric mean concentration by stratum:   22.42 16.95 12.55  3.03  1.03  3) Geometric shift-to-shift standard deviation by stratum:   2.96  2.79  2.34  3.09  3.10  4) Geometric person-to-person standard deviation by stratum:   1.40  2.00  2.15  2.14  1.57  Guidelines for Antisapstain Exposure Assessment  A14 Exhibit B:  Method B Results - Stratum Logarithmic Means and Variances      ---------------------------------- STRATUM=1 -------------------------------- -                            N          Mean      Variance                        -------------------------------                         30     3.2234967     1.0261790                        -------------------------------  ---------------------------------- STRATUM=2 -------------------------------- -                            N          Mean      Variance                        -------------------------------                         30     2.9523898     1.1616341                        -------------------------------  ---------------------------------- STRATUM=3 -------------------------------- -                            N          Mean      Variance                        -------------------------------                         30     2.7354386     1.3975117                        -------------------------------  ---------------------------------- STRATUM=4 -------------------------------- -                            N          Mean      Variance                        -------------------------------                         30     1.1867683     1.6757578                        -------------------------------  ---------------------------------- STRATUM=5 -------------------------------- -                            N          Mean      Variance                        -------------------------------                         30     0.1431212     1.1794050                        -------------------------------  Guidelines for Antisapstain Exposure Assessment  A15 Table B:  Estimated proportion of workers in 5 exposure strata exceeding various concentrations   Method B (conservative):   Stratum All Strata  Concentration  1  2 3 4 5 Combined    10.0 1.000 1.000 1.000 0.230 0.038 0.564  30.0 1.000 1.000 1.000 0.049 0.010 0.514  50.0 0.276 0.191 0.166 0.026 0.006 0.109  70.0 0.155 0.115 0.102 0.018 0.004 0.065  90.0 0.108 0.082 0.073 0.013 0.003 0.046  110.0 0.082 0.063 0.056 0.010 0.002 0.036  130.0 0.066 0.051 0.046 0.009 0.002 0.029  150.0 0.055 0.043 0.038 0.007 0.002 0.024  170.0 0.047 0.037 0.033 0.006 0.002 0.021  190.0 0.041 0.032 0.029 0.006 0.001 0.018    Based on method B output:   1) Stratum populations as percentage of total:   13.33 16.67 20.00 23.33 26.67  2) Geometric mean concentration by stratum:   25.03 19.11 15.49  3.29  1.15  3) Geometric total standard deviation by stratum:    2.76  2.94  3.26  3.66  2.96  Guidelines for Antisapstain Exposure Assessment  A16 Exhibit C:  Method C Results - Stratum Logarithmic Means and Variances     ---------------------------------- STRATUM=1 -------------------------------- -                            N          Mean      Variance                         ------------------------------                         10     3.7223508     0.7667791                         ------------------------------  ---------------------------------- STRATUM=2 -------------------------------- -                            N          Mean      Variance                         ------------------------------                         10     3.1377989     0.7442625                         ------------------------------  ---------------------------------- STRATUM=3 -------------------------------- -                            N          Mean      Variance                         ------------------------------                         10     2.5846468     1.0496606                         ------------------------------  ---------------------------------- STRATUM=4 -------------------------------- -                            N          Mean      Variance                         ------------------------------                         10     1.0933251     2.5739478                         ------------------------------  ---------------------------------- STRATUM=5 -------------------------------- -                            N          Mean      Variance                         ------------------------------                         10     0.2208569     1.2848269                         ------------------------------  Guidelines for Antisapstain Exposure Assessment  A17 Table C:  Estimated proportion of workers in 5 exposure strata exceeding various concentrations   Method C (conservative):   Stratum All Strata  Concentration  1  2 3 4 5 Combined    50.0 1.000 1.000 1.000 0.059 0.021 0.519  100.0 1.000 1.000 0.180 0.025 0.009 0.344  150.0 1.000 0.233 0.099 0.016 0.006 0.197  200.0 1.000 0.146 0.068 0.011 0.004 0.175  250.0 0.265 0.106 0.051 0.009 0.003 0.066  300.0 0.192 0.083 0.041 0.007 0.003 0.050  350.0 0.151 0.068 0.034 0.006 0.002 0.040  400.0 0.124 0.057 0.029 0.005 0.002 0.034  450.0 0.105 0.049 0.025 0.004 0.002 0.029  500.0 0.091 0.043 0.022 0.004 0.002 0.025    Based on method B output:   1) Stratum populations as percentage of total:   13.33 16.67 20.00 23.33 26.67  2) Geometric mean concentration by stratum:   41.26 23.10 13.20  2.97  1.25  3) Geometric total standard deviation by stratum:    6.00  6.00  6.00  6.00  6.00        


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items