Document b5VOOEyZe4X5k1ovo8xBE7qbD

AM. IND. HYG. ASSOC. J. 56;:J66-873 (1995) Relationships Between Phase Contrast Microscopy and Transmission Electron Microscopy Results of Samples from Occupational Exposure to Airborne Chrysotile Asbestos 401 Dave K. Verma Nancy E. Clark Occupational Health Laboratory, McMaster University, 1200 Main Street West, Hamilton, Ontario I.8N 3Z5, Canada Transmission electron microscope (TEM) andphase contrast microscope (PCM) fiber counts from the same filter were compared. Industrial hygiene samples from operations within a chrysotile mine, crusher, mill and tailings site, a brake manufacturing industry, and a taping products man ufacturer were used. A total of 10 318 individual fibers were analyzed from 65 filter samples. The aim of the study was to derive the relationships between TEM counts and PCM counts that could be used in the extrapolation ofrisk estimate from occupational exposure to low-level nonoccupational exposure to asbestos. The results show that ratios of EMPCM equivalent to PCM counts vary between 1.4 to 3.2. The proportion of long thin fibers (L > H pm, D < 1.5 pm) in creased as the asbestos operation moved from primary sec tor (mining) to end use sector (manufacturing). A sbeslos has long been known lo be a health hazard to /-% people occupationally exposed. Miners, millers, insu lators, and workers using asbestos in industries (e.g., asbestos cement ftransite pipe] industry, brake manufacturing, etc.) have suffered numerous exposure-related diseases (e.g., asbestosis, lung cancers, and mesothelioma)." '4| The health effects ;ire well documented, and exposure levels for occupational settings have been determined since the late 1960s by the membrane filter method/4* While this method is a good index of exposure, in occupational settings, it does not pro vide the necessary information to be applicable to nonoccupa tional or even paraoccupational settings. To assess the true as bestos fiber exposure in such situations, electron microscopy (EM) (preferably transmission electron microscopy [TEM]) must be used. The phase contrast microscope/membranc filter methods can neither distinguish between asbestos and nonasbeslos fibers, nor between the different types of asbestos. Also, due to different regulations respecting asbestos, not all fibers are counted. Established methods define a fiber as greater than 5 pm in length with an aspect ratio of 3 to 1 or greater/4 5) Some methods also exclude any fibers with a diameter greater than 3 pm/f'~') These fiber definitions include many nonasbestos libers as well as excluding many asbestos fibers present in the air. As an index of tire concentration of asbestos fibers in the air, the phase contrast microscope (PCM) method is adequate when the predominant fiber type is asbestos; however, in nonoccupa tional and paraoccupational settings where many types of fibers are present, electron microscopy techniques that positively iden tify the types of fibers present must be used/14* The fiber counts obtained by PCM differ greatly from EM. It has been suggested that counts obtained by EM could be 50 to 70 times higher than those obtained using PCM on the same air sample/9"101 This difference essentially relates to the issue of fiber dimensions; that is, fibers of sizes visible by EM and those visible by PCM. It is known that PCM cannot detect fibers smaller than 0.25 pm (about 0.3 pm in practice) while EM, with its superior resolution, can detect unit fibrils of diameters as small as 0.01 pm. This is graphically depicted in Figure 1. The fibers that are thin, generally with diameters less than 1.5 pm (variously from 0.25 pm to 1.5 pm), and lengths longer than 8 pm, are thought to be most carcinogenic/A vast majority of such long thin fibers cannot be detected by PCM and can only be seen by EM (see Figure 1). The relative number of fibers in the different size ranges likely depends on the type of asbestos and operations involved. Furthermore, PCM methods may in clude nonasbestos fibrous material when it conforms to the def inition of a fiber. Accurate identification and quantification of asbestos in nonoccupational and paraoccupational situations can only be reliably obtained by using EM methods. It is generally accepted that the health risks associated with fiber levels typical of nonoccupational and paraoccupational ex posure are very low/2'3,15* At present, data does not exist that relates low asbestos liber levels obtained by EM to health cffects/2,3,15* however, a grading of relative hazard potential can be developed if the following data is accumulated: (1) an inven tory of fiber levels in outside air by EM, and (2) in selected occupational exposure situations, air samples analyzed by both EM and PCM. A correlation between EM and PCM liber counts could be established for specific occupational settings and fiber types. The AM. IND. HYG. ASSOC. J. (56) ! September 1985 866 Copyright 1995, American Industrial Hygiene Association counts from mining, milling, brake manufacturing, and a dump 1 n ing and loading operation, and (2) to compare this study's find ings to those previously reported. -1012 Log10 Length (pm) FIGURE 1. Ability of electron microscopy (EM) and phase-contrast microscopy (PCM) to detect asbestos based on fiber lenplh and diameter health effects based primarily on PCM counts could then be re lated to EM fiber counts. It is postulated that the relationship between asbestos concentration measured by PCM and by EM is: Cpcm " f(CtM. type of asbestos, operation, process, ?) In this equation the question mark represents unknown fac tors such as the source of the asbestos--for example, Canadian or Rhodesian chrysotile. Using a linear dose response relation ship assumption, the magnitude of health risk of low-level ex posure could be estimated by extrapolation as conceptually shown in Figure 2, where the Y axis represents health risk in terms of morbidity or mortality. Chesson et al.<lw realized the difficulties and uncertainty in volved in determining the quantitative relationship between lev els of exposure and the risk of disease at the levels associated with nonoccupational asbestos exposure. They presented a mathematical model for investigating asbestos risk estimates based on measurement by one method (PCM) to an equivalent measurement by another method (EM). The model suggests ap proaches for obtaining conversion factors that will allow TEM measurements to be used in a PCM-based risk equation. Of the three ingredients in the model the most crucial is the distribution of asbestos fibers/structures and the ability of the measurement methods to detect the size of each structure. The relationships between fiber counts obtained by PCM to those obtained by EM in occupational environments is crucial to the determination of risk estimates at low levels. This study attempts to provide needed data relating PCM fiber counts to those obtained by EM for use in risk estimates. More specifically, the objectives of this study were (1) to obtain data relating PCM counts to the TEM MATERIALS AND METHODS A total of 65 samples, consisting of 52 personal and 13 area samples, taken for routine industrial hygiene monitoring of air borne asbestos concentrations, were analyzed by phase contrast microscopy and by transmission electron microscopy. Chrysotile was the only type of asbestos in all of the operations sampled. The air samples were collected on 37-mm diameter, 0.8 pmpore size, cellulose ester filters at a sampling flow rate of 2 L/ min. Air monitoring of occupational exposure to asbestos in Can ada routinely used 37-mm diameter filters as per NIOSH P&CAM 239 (predecessor of NIOSH 7400 method) at the onset of this project. This filter size was used throughout the project to avoid introducing another variable, and also because there was a larger area available for eight good-size wedges for replicaLc analyses. Most of the samples were long-term samples. The sam pling time was varied to accommodate the different operations and dust levels. The details of samples arc given in Table I. For analysis each filter was divided into eight equal wedges. A one-eighth wedge was cut from the filter and prepared for PCM analysis by the standard acetone/triacctin method07' and analyzed according to the Ontario Ministry of Labour regulation respecting asbestos made under the Occupational Health and Safety Act, revised July 1987.t/' This method is essentially the NIOSH 7400 method01 (counting rules A) with an upper diam eter cutoff of 3 pm. The wedge cut for PCM was assigned num ber 8. Of the remaining seven wedges, four were randomly se lected and prepared for TEM analysis. The selection of filter Fiber Concentration (CpcM) In Fibers per CC CpCM = * (CEM, tyP*. operation, process,?) FIGURE 2. Conceptual relationship between risk and as bestos concentrations 867 AM. IND. HYQ. ASSOC. J. (56) t September 1995 TABLE I. Origin of industrial Hygiene Samples and Number of Samples and Fibers Analyzed Industry Operation Type* and Number of Samples (filters) Range of Air Volumes (L) Mining Crusher Milting Tailings Taping products manufacturer Brake manufacturer Driller Shovel operator Pit electrician Pit laborer Primary Secondary Crusher and dryer Baghouse Floors 2 to 6 First floor (packaging) Helper and operator Dumping and loading Various tasks 3-P 5-P 4-P 3-P 5-A 3-P 2-A 5-P 5-A 5-P 1-A 3-P 8-P 13-P 80-440 280-510 170-428 250-260 118-125 170-310 216-218 120 120 120-238 290-456 37-52 10-446 Total A A = area samples, P - personal samples 65 (13-A + 52-P) Total Number of EM Fibers 852 573 200 58 843 780 632 438 1103 1351 638 1236 1614 10318 wedges and the direct transfer technique used have been de scribed previously.08* Since ail samples received were from oc cupational settings where chrysotile was the only type of asbes tos used and the only type of fiber expected, identification of asbestos fibers during TEM analysis was by morphology and tubular structure alone. All fibers were sized with respect to length and diameter, and the results entered into a program writ ten for a VAX 8530 minicomputer (Digital Equipment of Can ada, Ontario, Canada) and size distributions generated. A total of 10 318 EM fibers were sized. RESULTS Table I gives the industries from which the samples were taken, the operation being performed during sampling, the number of samples, the air volume sampled, and the total number of EM fibers analyzed. In Table II the distribution of fiber concentrations of various size ranges for individual samples of the mining operations are given. The data in Table II will permit calculation of different ratios of EM to PCM for each sample. For example, all fibers (all lengths, all diameters) or all asbestos fibers (all lengths, all diameters) or fiber concentrations of defined size ranges can be ratioed to the PCM concentrations. The EM all-fiber counts in clude all fibers as defined by an aspect ratio of 3 to 1 irrespective of length and diameter, and includes nonasbestos fibers. The EM confirmed asbestos fibers are presented as all lengths, all diam eters and lengths greater than 5 pm, and lengths smaller than 5 pm. The asbestos fibers greater than 5 pm in length have been subdivided into all diameters, diameters less than 3 pm, and di ameters between 0.3 pm and 3 pm. The EM fibers greater than 5 pm in length, with diameters between 0.3 pm and 3 pm (lim ited by PCM resolution and the counting rules), should theoret ically be visible by PCM and equivalent to PCM counts. For convenience these EM fibers are termed EM-PCM equivalent (EMpcM,;). Likewise, the fibers greater than 5 pm in length with diameter less than 3 pm that were detected by EM are termed since this group contained those with diameters less than 0.3 pm in addition to EMpcMii fibers. The individual raw data has been ratioed (EM/PCM) and grouped by location, operation, and industry in Table III. In Figure 3 the long thin fibers as percent of the total EM fibers are shown for various operations under the broad industrial classification of primary and end-use sectors. In Figure 4 the data showing ratio of EM to PCM counts has been presented. DISCUSSION There are numerous indices from which to choose a criterion to determine the health hazards due to asbestos exposure. Long thin fibers, sometimes inappropriately referred to as Stanton fibers, have been considered most carcinogenic. Lippman(f,t has pro posed asbestos exposure indices relating to asbestosis, mesothe lioma, and lung cancer. Unfortunately, the indices proposed re late mainly to amphibole asbestos and are expressed as total surface area of fibers within a specific size range for asbestosis. These fibers are greater than 2 pm in length and greater than 0.15 pm in diameter. For mesothelioma and lung cancer, Lippman's exposure indices are expressed in number of fibers with minimum lengths of 5 pm and 10 pm and diameters less than 0.1 pm and greater than 0.15 pm, respectively. These indices are based on animal as well as human studies. When analyzing airborne asbestos by TEM, many different results can be obtained, such as total fibers present, PCM equiv alent fibers, long thin fibers, and ratios of TEM to PCM results. Depending on which result is used to assess the asbestos health hazard, one can arrive at different conclusions. When comparing AM. IND. HYG. ASSOC. J. (56) / September 1995 868 TABLE II. Distribution of Fiber Concentrations of Individual Samples from Mining Operations Operation All Fibers All Lengths All Diameters All Lengths All Diameters Fibers Per Millimeter Squared EM Fibers Asbestos Length > 5 (.im All Diameters D < 3 p/n 0.3 \im < D < 3 \im Mining Tractor operator Shovel operator Shovel operator Shovel operator Shovel operator Mine electrician Pit electrician Pit electrician Pit electrician Secondary driller Secondary driller Primary driller Pit laborer Pit laborer Pit laborer Tailings helper Tailings helper Tailings operator 1596 557 401 400 600 798 103 187 71 8900 1856 406 186 169 49 457 3714 4060 1496 531 386 393 592 732 91 162 65 8860 1802 379 172 141 49 450 3714 4060 278 149 21 28 105 200 19 32 0 2228 271 142 14 49 7 14 1386 1402 272 149 21 28 98 188 13 26 0 2201 257 122 7 35 0 14 1344 1336 167 91 14 21 42 172 6 19 0 1237 176 95 7 21 0 7 748 494 L < 5 pun All Diameters 1218 382 365 366 488 532 71 129 65 6631 1531 237 158 92 42 436 2328 2658 PCM 35 10 13 10 9 45 6 17 9 52 47 25 5 9 2 11 141 331 TABLE HI. Ratio of EM to PCM for Various Operations and Locations Ratio of EM to PCM Sector Operation and Location All EM Fibers Standard Mean Deviation All EM Asbestos Fibers Standard Mean Deviation EM Asbestos Fibers L > 5 pm L > 5 \im D < 3 pm 0.3 pm < D < 3 pm Standard Mean Deviation Mean Standard Deviation Primary mining driller 75.6 68.2 74.6 68.4 17.6 17.5 10.4 pit electrician 13.5 4.2 12.0 3.8 2.0 1.5 1.5 shovel operator 47.8 12.4 46,1 12.4 7.6 5.0 4.3 pit laborer 26.7 7.7 24.8 7.7 1.8 1.6 1.2 crusher primary 31.7 4.0 29.7 6.1 2.5 0.6 1.8 secondary crusher 32.9 7.5 32.9 7.5 7.2 1.5 4.0 crusher and dryer 36.7 1.5 36.7 1.4 6.8 1.3 4.4 milling baghouse 145.6 130.6 144.4 128.7 3.1 2.5 1.2 floors 2 to 6 39.6 14.5 39.5 14.5 3.7 1.5 2.7 first floor (packaging) 47.7 12.1 46.7 12.5 5.2 3.6 2.8 tailings helper and operator 26.7 12.0 26.5 11.7 4.9 3.4 2.5 End Use taping products manufacturer dumping and loading 18.5 6.7 18.4 6.6 2.9 0.9 1.4 brake manufacturer various tasks 19.5 12.6 19.4 12.7 5.4 4.3 3.2 9.4 1.4 2.8 1.0 0.7 0.9 0.6 1.8 0.9 1.7 2.0 0.7 2.5 869 AM. IND. HYG. ASSOC. J. (56) / September 1995 total TEM fibers, the milling operation has the highest results when ratioed to PCM counts (see Figure 4); however, when long thin fibers are used as the criterion to assess the health hazard, the tailings operation has the greatest number, followed closely by brake manufacturing (sec Figure 3). As can he seen in Figure 4, the ratio of EM to PCM counts is highly variable, ranging between 19 and 76. The high ratio of total EM to PCM of 76 in milling is due to the high result found in the baghouse, when; the dust collector bags more effectively capture the larger than the smaller fibers. This predominance of small fibers greatly increases the ratio of total EM to PCM. The ratio between EMkimi; tmd PCM is fairly consistent, ranging be tween 1.4 and 3.2 based on data grouped by operation (sec Table IV). The small range indicates that the techniques used for count ing fibers by PCM and counting and sizing fibers on EM were consistent. The results in Tables II and 111 show that the airborne fibers in the operations studied were mainly asbestos. Of the total EM fibers, 93 to 100% were confirmed as asbestos. This is not un expected considering the nature of the operations. The EMrcmk fibers as a percent of the total range from 4 to 18%. When EMpcMB is compared to PCM counts, the ratio for samples grouped hy occupatiou/location, as shown in Table III, ranges between 1.2 to 10.4; mostly below 4.4. Even when comparing similar operations, there seem to be differences in the results. The taping products manufacturer should be similar to the packaging operation at the mill, since one is placing the asbestos into packages and the other is dump ing the contents into a hopper. The origin of the asbestos from the taping products operation (sampled) is unknown and prob ably came from a different source than the mill in this study, or it is possible that there were some nonasbestos fibers identified as asbestos by morphological features. This may account for the differences observed. The results within an industry vary as they do between in dustries. Only a few studies have compared PCM to TEM and scanning electron microscopy (SEM)<,J'"I 2<I"271 fiber counts. Only four of these studies involved applying two techniques on the same filler. The study of Winer and Cossette,w and also reported by Chatfield,(l()) involved I I filters. The Hwang and Wang(24) study was carried out on 25 filters, while the Marconi et al.<25) study involved 23 filters, and (he Breysse et al.l27) study involved 5 filters. The current study is somewhat more comprehensive, involving 65 filters. Table IV summarizes for comparison earlier studies along with the results of this study. It becomes apparent when reviewing the literature that while many studies have been done, a consistent procedure has not been adopted either for analysis of the samples or for presentation of the results. In many of the studies some of the data needed to calculate the ratio between EM|,cm[. fibers and PCM fibers is not available cither because light microscope counts were not performed or the data was not presented in the publication. With all of the variation inherent in the methods, it would seem that a consistent way of presenting data may be more im portant then the specific method used. The data of comparative studies should be presented in the manner shown in Table II, which would permit derivation of various desired ratios. To ob tain such data each fiber has to be sized accurately with respect to its diameter and length using an EM, which admittedly is very time consuming. In earlier studies PCM counts have been compared to total EM fibers of all diameters but length greater than 5 pm, which produced a very high ratio of about 66 (EM to PCM). The criteria used here is not truly a PCM equivalent, since it will include diameters greater than 3 pm as well as those smaller than 0.3 pm. In other studies the PCM counts have been com pared to the EM*TM,* fibers (i.e., fibers of diameter between 0.3 and 3 pm and lengths greater than 5 pm) and to the plus (i.c., fibers of diameter less than 3 pm and length greater than 5 pm).124,25 27' AM. INO HYG. ASSOC. J. (56) / September 1995 870 TABLE tV. EMpcme Ratioed to PCM and as Percentage of Total EM Fibers Process Number of Filters EMpcme- PCM Mean Sfd. Dev. EMpcmea Percent of TotalB Mean Sfd. Dev. Drying chrysotile 2 * 10 Bagging chrysotile Carding chrysotile Total = 6 Various 2* 2* 4 2.1 11 66.3 63.9 4.1 3.4 Mining NR * 0.8 Bagging Making brake linings Milling Mining (underground) Mining NR 66 fibers 139 fibers 249 fibers NR * * * 1.5 21,2 14.4 10.0 0.4 Bagging Various NR * 25 4.8 1.2 5.8 Asbestos cement 6 1.5 0.6 25.7 12.5 Insulation Brake and clutch manufacturing Total = 23 Generated standard samples Water tank insulation removal 9 8 4 5 1.1 0.6 29.4 14.7 1.3 0.4 53.9 19.3 2.2 0.5 11.5 0.8 D 27.2 10.0 Mining Crusher Milling Tailings Dumping and loading Brake manufacturing 15 3.2 3.3 9.5 6.8 10 2.8 1.2 9.0 3.9 16 1.9 1.4 4.3 2.5 3 2.5 1.8 11.3 7.4 8 1.4 0.6 8.6 4.1 13 3.0 2.4 18.2 10.3 Total = 65_____________________________________________ A EMpcme fibers are asbestos fibers greater than 5 pm in length B Total fibers include non-asbestos fibers c NOTE: Results have been derived from data in the referenced publications D Data not available Source and Notesc (Reference Number) Gibbs and Hwang, 1975(20) 0.5 pm diameter Winer and Cossette, 1979(9) No diameter specified Gibbs and Hwang, 1986iJ1> No diameter specified Rendall and Skikne, 1980132' 0.4 pm diameter Hwang and Gibbs in Walton, 1982|4> 0.5 pm diameter Hwang and Wang, 1983'3'11 0.3 pm diameter Marconi et al., 1984(2S1 0.3 pm diameter Pang et at., 1984|26> Breysse et al., 1989(2?) European Reference Method (ERM) Fibers Verma and Clark (this study) As c;m be seen from Table HI, the ratio of EM/PCM is signif icantly different for various fractions. For example, for Driller it is 75.6 (o = 68) for total EM fibers; 17.6 (0 = 17.5) for EMpcme* and 10.4 {a = 9.4) for EMhjmK. It is thus important that criteria selected for comparison be clearly slated and understood. This is the main reason for the widely different ratios being reported in the literature. From a health hazard point of view it may be more suit able to use the EMpcmk* to PCM ratio. The total EM/PCM ratio is not really relevant. Previous studies generally have data related to EMpcme, and that is the reason in Table IV the EMpcme ratio of earlier studies has been listed, wherever it was jxjssible to determine from published information, along with this data. In Figure 4 grouped data from this study is graphically shown as EM to PCM ratio for total EM fibers, EM-mi'.+ , and EMpcme. Provided that the lengths and diameters of the fibers are mea sured accurately, and the criterion used for expressing the data (EMpcme or EMpcme*) is clearly established, it may be possible to pool results from the studies for extrapolation of the doseresponse curve for assessment of health effects related Lo lowlevel asbestos exposure. A single conversion ratio of 10 has been recommended be tween EMp^E and PCM counts for asbestos (presumably for all asbestos types and operations).(1> The results of this study as well 871 AM. INO. MYG. ASSOC J. (56) I Saptambe, 1995 as a few previous studies, listed in Table IV do not support this relationship. Since the relationship depends on several factors as discussed earlier, it would be prudent to use the relationship from industry and operations that are similar to the one in question. The majority of ratios between EMk.-mh and PCM seems to range between I to 3. Most of the previous studies were conducted generally on a small number of samples and only four of these involved analysis by two techniques on the same filter. This study, conducted on a reasonably large number of samples (65 fillers and 10 318 fibers), shows the ratios to be within this range (see Table IV). The authors realize that this study was limited to chrysotile asbestos only. Further studies, similar to this, should be conducted involv ing different asbestos fiber types (amosite and crocidolite) on occupationally exposed samples. This may be somewhat difficult now in North America, where occupational exposure to asbestos generally is being phased out. It is important, however, that such studies be conducted, so that conversion factors can be derived for use in the risk esdmate of nonoccupational asbestos exposure. If one excludes tailing samples because of the small sample size, the data in Figure 3 shows a trend of the proportion of long thin fibers increasing as the asbestos processing moves from pri mary sector (mining) to the end-use sector (manufacturing). This is important in terms of health effects such as occurrence of lung cancer, which has generally been much higher in the end-use sector (e.g., among insulators and manufacturing workers) than in the primary sector (.such as mining). CONCLUSIONS The following conclusions were reached: (1) To quantify the health effects of low-level asbestos ex posure, relationships between the PCM and TEM counts in occupational settings must be established. The data should be presented in these PCM/TEM com parison studies in a consistent manner so that results can be pooled from different studies. It is also impor tant to include the number of samples and/or fibers an alyzed to determine whether or not statistical signifi cance can be placed on the results. (2) This study indicates that the ratio of to PCM is likely to he in the range of 1 to 3. (3) There appears to be a trend of the proportion of long thin fibers increasing as asbestos processing moves from primary sector (mining) to the end-use sector (manufacturing). (4) Further work of a similar nature on occupationally ex posed samples involving asbestos fibers of different types is needed. ACKNOWLEDGMENTS The authors are grateful to the Ontario Ministry of the Environ ment, Air Resources Branch for their financial support. They are grateful to their colleagues for providing them with the field samples for the study. They thank Jim Julian for his help in the study design and statistical analyses. Thanks are also due to Mar garet King and Jean Bodnar for their help, REFERENCES t. Ontario Ministry of the Attorney General: Report of the Royal Commission on Matters of Health and Safety Arising from the Use ofAsbestos in Ontario, by J.S. Dupre, J.F. Mustard, and R.J Uffen. fISBN 0-7743-8508-1] Toronto, Ontario, Canada: Queen's Printer for Ontario, 1984. pp. 1 -920, 2. National Research Council, Committee on Non-Occupational Health Risks of Asbcstiform Fibers: Asbestiform Fibers--Non-Occupational Health Risks. Washington, D.C.: National Academy Press, 1984. 3. U.S. Environmental Protection Agency: Airborne Asbestos Health Assessment Update, [EPA/600/8-84/003F June 1986] Environ mental Criteria & Assessment Office, 1986. 4. Walton. W.H: The nature, hazards and assessment of occupational exposure to airborne asbestos dust: a review. Anri. Occup. Hyg. 25:117-247(1982). 5. National Institute for Occupational Safety and Health: Fibers Method 7400, Revised 1989. In NIOSH Manual ofAnalytical Meth ods, P.M. Eller, ed. 3rd ed. IDHHS/NIOSH Pub. No. 84-100] Washington, D.C.: Government Printing Office, 1984. pp. 7400-1 7400-14. 6. Asbestos International Association: Reference Method for (he De termination of Airborne Asbestos Fibre Concentrations by Light Microscopy (Membrane Filter Method). London: Asbestos Inter national Association, 1979. Recommended Technical Method 1 (RTMt). 7. Ontario Ministry of Labour: Regulation Respecting Asbestos (Created under the Occupational Health and Safety Act, Revised Statutes of Ontario, 1980, Chapter 321, filed 20 August 1982, Re vised, 1987). 8. Breysse, P.N.: Electron microscopic analysis of airhome asbestos fibers. Crit. Rev. Anal. Chem. 22:201-227 (1991), 9. Winer, A.A. and M. Cossette: The effect of aspect ratio on fiber counts: a preliminary study. Annal. N.Y. Acad. Sci. 330:661-672 (1979). 10. Cbatfield, E.J: Measurement of Asbestos Fibres in the Workplace and in the General Environment. In Short Course Handbook, R.L. Ledous, ed, Quebec: Mineralogical Association of Canada, 1979. pp 111-157. 1 i. Stanton, M.E. and M, Layard: The Carcinogenicity of Fibrous Mineral. In Workshop on Asbestos: Definitions and Measurement Methods, G.C. Graven, P.D. LaFleur, and K.F.J. Heinrich, eds. [Na tional Bureau of Standards Special Pub. #506] Washington, D.C.: U.S. Government Printing Office, 1978. 12. Stanton, M.F. and C. Wrench: Mechanism of mesothelioma in duction with asbestos and fibrous glass. J. Natl. Cancer Inst. 45:797-821 (1972). 13. Stanton, M.F., A. Tegeris, E. Miller, M. May, E. Morgan, and A. Smith: Relation of particle dimension to carcinogenicity in amphibole asbestos and other fibrous minerals. J. Nail. Cancer Inst. 67:965-975 (1981). 14. Wylie, A.G., K.F. Bailey, J.W. Kelse, and R.J. Lee: The importance of width in asbestos fiber carcinogenicity and its implication for public policy. Am. hid. Hyg. Assoc. 54:239-252 (1993). 15. Health Effects Institute: Asbestos in Public and Commercial Build ing. A Literature Review and Synthesis of Current Knowledge. Cambridge, MA: Health Effects Institute, 1991. AM IND. HYG. ASSOC. J. (561 > SsptombBr 1995 872 16. Chesson, J., R.D. Rench, B.D. Schultz, and K.L, Milne: Inteipretation of airborne asbestos measurements. Risk Anal. /0:437 -446 (! 990). 17. National Health and Medical Research Council: Membrane Filter Method for Estimating Airborne Asbestos Dust. Canberra, Austra lia, 1976. pp. 18-20. 18. Verma, D.K., N.E. Clark, and J.A. Julian: Asbestos fiber charac terization using an analytical transmission electron microscope and a microfilm reader. Am. Ind. Hyg. Assoc. J. 52:113-119 (1991). 19. Lippman, M.: Asbestos exposure indices. Environ. Res. 46:86-106 (1988). 20. Gibbs, G.W. and C.Y. Hwang: Physical parameters of airborne asbestos fibers in various work environments--preliminary find ings. Am. Ind. Hyg. Assoc. J. 36:459-466 (1975), 21. Gibbs, G.W, and C.Y, Hwang: Dimensions of Airborne Asbestos Fibres. In Biological Effects of Mineral Fibres, J.C. Wagner, cd. [IARC Sci. Pub. No 301 Lyon, France: Internationa] Agency for Research on Cancer, 1980. pp. 69-78. 22. Rcndall, R.E.G. and M.I, Skikne: Submicroscopic Fibres in Indus trial Atmospheres. In Biological Effects of Mineral Fibres, J.C. Wagner, ed. flARC Sci. Pub. No. 30] Lyon, France: International Agency for Research on Cancer, 1980. pp. 837-843. ------ 7---------------------------------------------------------------- 23. Hwang, C.Y. and G.W. Gibbs: The dimensions of airborne asbestos fibres: 1. crocidolite from Kuruman Area Cape Province, South Af rica. Ann. Occup. Hyg, 24:23-41 (1981). [Also personal commu nication] Cited in "Nature, hazards and assessment of occupational exposure to asbestos,'1 by W.H. Walton. Ann. Occup. Hyg. 25(2):\ 17-247 (1982). 24. Hwang, C.Y. and Z.M. Wang: Comparison of methods of as sessing asbestos fiber concentrations. Arch. Environ. Hlth. 38:510(1983). 25. Marconi, A., E. Menichini, and L. Paolclli: A comparison of light microscopy and transmission electron microscopy results in the evaluation of the occupational exposure to airborne chrysotilc fi bres, Ann. Occup. Hyg. 2S:32f-331 (1984). 26. Pang, T.W.S., W.L. Dicker, and M.A. Nazar: An Evaluation of the Precision and Accuracy of the Direct Transfer Method for the Anal ysis of Asbestos Fibers with Comparison to the NIOSH Method. Am. Ind. Hyg. Assoc. J. 45:329-335 (1984). 27. Breyssc, P.N,, J.W. Cherrie, J. Addison, and J. Dodgson: Evalu ation of airborne asbestos concentrations using TEM and SEM dur ing residential water tank removal. Ann. Occup. Hyg. 32:243-256 (1989). 873 AM. IND. HYG. ASSOC. J. (56) / Saptember 1995