Document 6b2Dq11vMLoXjBDJDeNXaKyxg

FROM BIOMEDICAL INFORMATION SERVICE ijtUME 28, No. 3 1984 (MON) 10. IS' 0 3 15:1 8/ST. 16:1 O/NO. 4862209231: F- 24 1984 i based ion of storical work- inufac- \ ( sented f tracer tples of ules on nbrane t Salable f i I ce 1983 .ochum. AN INTERNATIONAL JOURNAL PUBLISHED FOR THE BRITISH OCCUPATIONAL HYGIENE SOCIETY HON. EDITOR-IN-CHIEF MR W. H. WALTON, O.B.E. Institute of Occupational Medicine Edinburgh EDITORIAL ASSISTANT MRS S. M. COPPOCK HON. ASSISTANT EDITORS DR J. McK. ELLISON DR A. CRITCHLOW DR C. G. WARNER cover) )003--4878 290 (1984) 201 PERGAMON PRESS Oxford New York Toronto Sydney Paris HWBUI0008785 FROM BIOMEDICAL INFORMATION SERVICE r (MON) 1 0. 13' 03 1 6 : 1 8/ST. 1 6 : 1 O/NO. 48 6 2 2 0 9 2 3 0 P 25 tmples. Ann. jevel . TM/74/12. T mines as a ire to radon Am. occup. Hyg.. Vol. 28. No. 3, pp. 321-331. 1984. Printed in Great Britain. 0003 4S7S/84 S3.00 + 0.00 Pcrgamon Press Ltd. i 1984 British Occupational Hygiene Society. A COMPARISON OF LIGHT MICROSCOPY AND TRANSMISSION ELECTRON MICROSCOPY RESULTS IN THE EVALUATION OF THE OCCUPATIONAL EXPOSURE TO AIRBORNE CHRYSOTILE FIBRES A. Marconi,* E. Menichini* and L. Paoletti| ; "Laboratorio di Igiene degli Ambient! Confinati fLaboratorio di Ultrastrutture Istituto Superiore di Sanita, Viale Regina Elena 299, 00161 Roma, Italy Abstract-- A comparison has been made between light microscope (LM) and transmission electron microscope (TEM) counts of airborne chrysotile fibres, collected on 30 membrane filters from different working processes. For TEM counts, fibres of length l> 5 were classified according to diameter d, namely 0.3 ftm<d<3 fan (microscopic fibres) and d<,0.3 fan (submicroscopic fibres); short fibres, l<5 ftm, were taken into account separately, A very good correlation has been found between LM and TEM counts, both for microscopic and total (microscopic + submicroscopic, /> 5 ftm) fibres, for densities 40-500 fibres mm"!, The median of the ratios between TEM and LM counts has been found to be 1.1 for microscopic fibres and 1.2 for total fibres. From TEM counts, submicroscopic fibres have been found to account for about 10% of total (>5 fim fibres. The mean ratio between short fibres and those with 1 > 5 fan was in the range 0.8-3.6, depending on the working process. These results suggest the possibility of adopting factors, varying with the kind of working process, to be applied to LM counts in order to estimate the exposure to all the asbestos fibres considered important for a comprehensive evaluation of the risk. i INTRODUCTION At PRESENT, the membrane filter method (MFM), based on optical phase microscopy, is commonly used for measuring airborne asbestos dust levels in occupational environments. It has been used in most epidemiological studies relating asbestos dust exposure to lung disease. The light microscope (LM) is considered a suitable instrument for routine control of airborne asbestos fibres, but, because of its low detection limit (approximately 0.2-0.4 pm), it cannot detect all fibres; consequently, the results obtained by this instrument are intended as ah indication of exposure to fibres, rather than a concentration of all harmful fibres. On the other hand, animal experiments have indicated the importance of fibres less than 0.5 pm in diameter (in particular, less than 0.25 pm) in the carcinogenic process (Stanton et al, 1981; Pott, 1978). In addition, data on fibre recovery from human lung tissue have shown that most fibres are less than 0.5 pm in diameter (Bignon ei al, 1978) and that the optically visible fibres are only a small fraction of the fibres measured by electron microscopy (Poolev and Clark, 1980). A different carcinogenic potential according to fibre size has been suggested to explain the incidence of mesothelioma in South African mines (Timbrell et al., 1971). The above findings indicate the inadequacy of LM in evaluating the true exposure 321 HWBUI0008786 FROM BIOMEDICAL INFORMATION SERVICE (MON) 1 0. 1 3' 03 1.6'18/ST. 1 6 : 1 O/NO. 4862209230 P ^6 322 A. Marconi, E. Menichini and L. Paoletti to all harmful airborne fibres: in order to make the optical data more representative of the true exposure, it is necessary to transform them, by appropriate conversion factors into the true number of airborne fibres having pathogenic dimensions. To obtain such factors, the best available method nowadays appears to be electron microscopy, which is capable of resolving the finest asbestos fibres: traditionally it has been used for si/ins submicron fibres, but its routine use is considered too expensive in terms of time and costs (Walton, 1982). Up to now, there have been few published studies on the direct comparison of the results obtained by light and electron microscopy (Beckett and Jarvis, 1979, for amosite; Hwang and Gibbs, 1981, for crocidolite; Chatfield, 1979, for chrysotile). However, from the results of some electron microscope studies on airborne fibre dimensions, it is possible to calculate the percentage of fibres not detectable by LM and hence, indirectly, to estimate the relationship between the results obtained by the two instruments (Gibbs and Hwang, 1975,1980; Hwang and Gibbs, 1981; Rendall and Skikne, 1980; Winer and Cossette, 1979). A comprehensive analysis of these studies indicates the difficulty of establishing a generally valid relationship, since differences in size distribution of the airborne fibres can occur, owing both the mineralogical variety and to the stages of the production process (Gibbs and Hwang, 1975; Hwang, 1983). These studies have obvious implications in the establishment of occupational hygiene standards for airborne asbestos fibre concentrations, compliance which is normally controlled by optical methods. Detailed and systematic studies involving different types of asbestos and different production processes are therefore required in order to correlate results from the two microscopical techniques. This paper reports a comparison between LM and transmission electron microscope (TEM) counts on airborne dust samples collected during three different working processes where chrysotile was used. The results are then analyzed and discussed in comparison with the other results previously published on the same subject. MATERIALS AND METHODS Samples were collected in a brake and clutch (BC) factory (8 samples) and at a railway workshop dealing with maintenance and repairs (22 samples); in the latter case, fibre exposure resulted from asbestos tapes and paperboards (ATP) handling, and from asbestos-cement (AC) board sanding. Fibre type was chrysotile, for all processes. Air samples were collected by the MFM (AIA, 1979) on 25 mm and 37 mm filters. In order to detect a possible influence of fibre density on filters (observed fibres per mm2) on the relationship between LM and TEM, samples were selected so as to cover a wide density range (10-1500 fibres mm-2 by LM). For each filter, a portion was examined by LM and another one by TEM, measurements being made on areas of the same order of magnitude. Counting criteria were those reported in AIA (1979) and it was verified that they were interpreted in the same way by the two operators who performed separately the LM and TEM analyses. In order to eliminate coarse errors due to a possible uneven fibre distribution on the samples prepared for microscopic examination, for each sample we tested the goodness of fit of the Poisson distribution to that observed in LM and TEM counts. In ideal conditions fibre distribution is random and' therefore describable by the Poisson law HWBUI0008787 FROM BiOMEDI CAL INFORMATION SERVICE (MON) 1 0. 1 3' 03 16:1 9/ST. 16:1 O/NO. 4862209230 P 27 A comparison of light microscopy and transmission electron microscopy 323 It"til ive of vhich sizing e and }f the for >tile), fibre 1 and : two - and udies :es in riety 983). onal oh is Iving od in Iron ame at a itter and sscs. s. In un2) vide ined rder fied ned (Rajhans and Bragg, 1975; Leineweber, 1978; Millette et ai, 1978). The Kolmogorov-Smirnov test was used to evaluate the goodness of fit. In addition, on the samples examined by LM, the relationship between the mean and the variance of the fibre number per microscopic field was studied by linear regression analysis (Battisti et ai, 1981). Light microscopy: sample preparation and analysis A filter portion, corresponding to one wedge-shaped quarter for 37 mm filters and to one-half for 25 mm filters, was cleared by acetone vapours and triacetin (AIA, 1979). Fibre counting was performed by a Leitz Ortolux phase-contrast microscope, with a magnification of 500 x, and a practical detection limit ca 0.3 pm, checked by the test slides produced by the Asbestos Institute for Occupational and Environmental Safety and Health (AIA, 1979) and by the Health and Safety Executive/National Physical Laboratory, Mark I and Mark II versions (Le Guen, in AIA, 1980, p. 83). Fibres were counted using a Walton-Beckett graticule (Walton and Beckett, 1977), following the rules and criteria now generally accepted and reported in AIA (1979); for each sample, the scanned area was between 0.16 and 0.79 mm2, according to the fibre density on filter. Transmission electron microscopy: sample preparation and analysis Membrane filter preparation for TEM analysis was performed according to the `collapsed membrane filter' method (NIOSH, 1977): a filter section (equivalent to that used for LM) was removed, placed on a microscope slide, sample side up, and fastened to the slide by a narrow transparent tape. The slide was exposed to acetone vapours to destroy the microporous structure of the filter (normally 2-3 min were sufficient to produce a smooth surface) and then coated with carbon (20-30 nm) by conventional techniques. Sections were cut from the coated filters and placed, sample side down, on 200-mesh electron microscope grids. The filter matrix was gently dissolved by acetone, leaving particles adhering to carbon film ready for TEM analysis. Copper grids with opening of 85 x 85 pm were used; in the analysis of the filter the full grid opening was used as field of view for fibre counting. A Siemens Elmiskop 102 transmission electron microscope was used at a magnification of x 104. Fibre sizes were measured on the microscope screen with the help of known reference marks on it. The counting rules and criteria were the same as for LM (AIA, 1979); in particular, ^ fibre intersecting one side of the field of view was considered a half-fibre. All the fibres belonging to one of the following size classes were considered for analysis: (1) l>5 pm, 0.3 pm <d<3 pm, I/d>3 (microscopic fibres); (2)/> 5 pm, d^0.3 pm, l/d>3 (submicroscopic fibres); (3) 1^5 pm, d <3 pm, l/d>3 (short fibres). In the text, the whole of microscopic plus submicroscopic fibres will be reported as `total' fibres (l > 5 pm). The recording of fibres with a length ^ 5 pm can be useful in order to provide full information on the size distribution of fibres, particularly in view of future epidemiological studies. Minimum diameter and minimum length of fibres detected by TEM were estimated to be, respectively 10" 2 and 3 x 10 ~ 2 pm. I 1 I f I HWBU10008788 INFORMATION SERVICE (MON) 10. 13' 03 16: 21/ST. 16 : 1 n -fT~ P 28 324 A. Marconi, E. Menichini and L. Paoletti RESULTS Statistical test of the random distribution offibres on filters For both TEM and LM counts performed on all the samples the null hypothesis (H0 = fibres are randomly distributed on filter) is not rejected at an a = 0.20 significance level. Therefore, the fibres can be considered as randomly distributed on filters. Table i reports an analysis of the relationship between the mean and the variance of the number of fibres detected in each LM graticule area; the results are reported as regression lines of the variance on the mean and as correlation coefficients (r). Table 1. Relationship between mean (x) and variance of the number of fibres PER LM GRATICULE AREA x (range) 0.00-1.00 1.01-5.00 0.00-5.00 0.00-12.00 No. of filters 16 10 26 30 Regression line r(P) y=1.00x-0.02 y =0.72x4-0.75 y=0.95x4-0.07 y=1.50x --0.57 0.98 (P<0.001) 0.85 (0.01 <P<0.001) 0.96 (P< 0.001) 0.96 (P< 0.001) In this analysis, samples have been classified according to fibre density, reported as mean number of fibres per graticule area; a very good agreement with the Poisson distribution has been obtained for the samples with a density up to 5 fibres per graticule area (about 650 fibres mm~2), with a maximum agreement for the samples with density up to 1 fibre/graticule area. As already reported (Battisti et al, 1981), these results are in agreement with the particular suitability of the Poisson distribution to describe rare events. Comparison between LM and TEM counts All the LM and TEM counts performed on the 30 samples are reported in Table 2, in increasing order of fibre density determined by LM. Comparing LM and TEM counts of microscopic fibres, it is possible to observe that, in four out of the five samples with the lower LM density (less than 20 fibres mm-2), no fibre has been detected in the area examined by TEM. Moreover, the two samples with a very high LM density (more than 1000 fibres mm-2) showed a particularly high ratio between LM and TEM counts. As indeed samples with such densities are not generally accepted in the good practice of this kind of analysis, because of the poor reliability of the results, we considered separately the 7 above-mentioned samples in the overall analysis of the results. Table 3 shows the results of the comparison between LM and TEM counts; TEM counts are reported both for microscopic fibres and for total fibres (!> 5 pm). Samples are classified according to working processes and also to fibre density ranges, where appropriate, in order to find out possible relationships between results and fibre density. The mean value of the ratios between the LM and TEM counts on each sample has been calculated for each working process and each density range. To evaluate the 1 HWBUI0008789 BIOMEDICAL INFORMATION SERVICE (MON) 1 0. 1 3' 03 1 6 : 21/ST. 16:1 0/NO. 4862209230 p 29 I A comparison of light microscopy and transmission electron microscopy 325 Table 2 Comparative summary of LM and TEM counts (fibres mm 2) Filter Process LM counts Microscopic fibres TEM counts Total fibres Short fibres 1 ATP 2 AC 3 AC 4 AC 5 AC 6 AC 7 ATP 8 ATP 9 ATP 10 ATP 11 AC 12 ATP 13 ATP 14 AC 15 AC 16 ATP 17 AC 18 BC 19 ATP 20 AC 21 BC 22 BC 23 BC 24 BC 25 BC 26 BC 27 ATP 28 BC 29 AC 30 AC 9 10 12 13 16 42 45 48 50 58 69 70 94 113 121 133 167 179 225 229 291 354 406 411 439 577 837 1000 1218 1446 ND ND 4 ND ND 54 7 33 40 83 78 85 94 184 181 133 119 218 552 589 346 419 415 564 768 439 872 2137 609 346 ND 1 4 ND ND 66 7 33 40 119 82 87 145 189 211 152 149 325 552 619 412 419 467 676 858 439 969 2137 740 526 83 9 3 ND ND 226 78 43 137 291 217 206 295 548 294 414 1535 402 421 774 301 266 353 670 914 277 1239 129 845 955 Microscopic fibres: 0.3 pm <d< 3 pm; total fibres: microscopic and submicroscopic fibres (/> 5 pm); short fibres: 1^5 pm; ATP: asbestos tapes and paperboards; AC: asbestos-cement boards; BC: brake and clutch; ND: no fibre detected. degree of association between the two different methods, the correlation coefficients between TEM and LM counts have been calculated. Table 4 shows the percentages, obtained by TEM counts, of submicroscopic fibres with respect to total fibres (/> 5 pm) and of fibres up to 5 /tm with respect to those more than 5 pm in length. Samples 1-5 have been excluded because no fibres, or very few, were detected in the scanned area. The percentage values turned out to be widely distributed. DISCUSSION TEM counts of microscopic fibres were expected to be roughly the same as LM counts. The results for the samples with densities in the range 40-1000 fibres mm-2 (Table 3) show a very good correlation between the two instruments (rs0.9) and a median ratio between TEM and LM counts about or more than.l, with a maximum of 1-4 for AC samples. TEM counts are almost always higher ; such a difference between j 'I HWBUI0008790 FROM BIOMEDICAL INFORMATION SERVICE o (MON) 10. 13' 03 16:21 /ST. 16:1 0/NO. 4862209230 P 3 U 326 A. Marconi, E. Menichini and L. Paoletti ^ vo r- OOOOoOo <2' g o s o V p tv, V 88 3 ^a >r( q^ <ONs vei ^ ^s-** ' s|| 88 (3 S O 'E. 0 2 o 1 S P O a. o o E E --j1 X<s ; nnno I W'f *>* aM ^ N <N rvi --4 v ^iv vod020d o v V V8 v V Si5s5d2s.s.5i2i M h ^ 03 o woV ^ ' *- --i a_ 8O 8O 8 "v faVt, aV, ika-V<i, o8<o5 <mfVNt- HoaVI* *n e c4 * '*Ti ^i Vhl> ^i e*Vs**i*j ^i ei A <^ <H fOl2twSSW2W * B 2 ri JJ5H At H 41 2 x> <c 13 w >i tf * `e. 2 o B U < u < f^e^f --^dN Ea. in A s > o rn *n V/ 8 rc-4* ^r4 oo --* E "S a. <n A ^5 Xi o s a sHo orn m--^ i odo odos-<so -! u vi n so >o dridd O E SD Os 00 <0 ON o Z p a. & <s4s OHu ^ <<aS H A C , ATP, BC, M : refer to Tables 2 and 3. HWBUI0008791 FROM BIOMEDICAL INFORMATION SERVICE (MON)10. 13' 03 16:22/ST. 16:10/NO. 4862209 A comparison of light microscopy and transmission electron microscopy 327 TEM and LM counts, in any case within the reproducibility of the MFM, can be attributed to the following causes: (1) operator's subjectivity in interpreting counting criteria (Walton, 1982, p. 205); (2) uncertainty in determining LM detection limit (a non-negligible amount of fibres with a diameter near 0.3 fxm has been found by TEM) (Rendall and Skikne, 1980; Walton, 1982, p. 234); (3) presence of a non-negligible amount of fibres with a non-uniform diameter, and in particular with d<0.3 /im in the central portion and with d> 0.3 fim but l < 5 nm in the external portions: as the central portion is not detectable by LM and the external portions do not conform to the definition of fibre, such fibres are not counted by LM while they are by TEM. On the other hand, as the number of fibres crossing the sides of the TEM opening is negligible, we can exclude the possibility that the result is significantly affected by the fibres which can touch particulate matter out of the opening, in non-visible zones. The lack of fibres detected by TEM in 4 out of the 5 samples with LM density less than 20 fibres mm-2 can be explained by a background on the samples prepared for the LM analysis due, e.g., to an ambient contamination or to the clearing procedure. This confirms (AIA, 1979) the importance of regarding as acceptable in the MFM the samples with a density of at least 20 fibres mm-2, also of minimizing the `blank' error. The maximum of the agreement between the results obtained with the two instruments has been found for the ATP samples; such samples do not generally present non-fibrous particulate matter, which is known as disturbing in counting and as leading to ambiguity in interpreting counting criteria. On the other hand, the agreement is minimum for the AC samples, which show an opposite situation. When TEM total fibres (submicroscopic + microscopic fibres, l > 5 jj.m) are considered, the median ratios between TEM and LM counts increase up to 25% as compared with the ratios found for microscopic fibres, previously discussed; the correlation coefficients are still very good (r^0.9), and they are basically the same as those found for microscopic fibres. In these conditions, and for the type of samples considered in this study, when considering the mean value for each process, TEM (including submicroscopic fibres in the counts) detects up to 60% more fibres than LM; the median increase for all the samples is 20%. As regards the kind of working process, results for total fibres (/> 5 /mi) are similar to those previously reported for microscopic fibres: with increasing amounts of non-fibrous particulate matter, the difference between the results obtained with the two instruments increases. Finally, separate TEM counts of submicroscopic and microscopic fibres indicate that submicroscopic fibres are on average about 10% of total fibres (1>5 /*m), although with a wide distribution of values. In Table 5 the results of this study are compared with those of other similar studies previously reported, concerning chrysotile fibres. Two studies (Chatfield, 1979, data reported also in Winer and Cossette, 1979; Hwang and Wang, 1983) have involved a comparison by applying the two techniques on the same filter: the three sets ofdata (the two above reported and ours) are considerably different; Chatfield (1979) reports, however, that TEM data can be over-estimated because of ultrasound used in preparing samples. Besides, Lynch et al. (1970) have performed a similar comparison over a great number of samples: data are not reported but they conclude that the TEM counts of the longest fibres do not exceed the LM counts. In the-other studies, fibre size distributions have been studied only by TEM: from such distributions it is possible to INFORMATI ON SERVICE 328 (MON) 10. 1 3' 03 1 6 : 22/ST. 1 6 : 1 Q/NO. 4862209230 P 32 A. Marconi, E. Menichini and L. Paoletti > S<0 S g vs I < z <r*n ON D T3 8 G& O 2 S8 <& a I 13 X & Xoz>- as 5 eS m d s+o1 o+! o+o1 r*i o m o r^ <rnn| On vnoS S .2 oo o Os n do n-Ho r+n14m-) oS d *j Z no os oo at at at zzz S3 e<o xt o <u T5 0.^3 S3 U CC I S g 2 ! go SS ea "O g T? O S S 5 ca s O' a-a M W JS < ^ lM A &ca & 2> e M illin g (bagging) T e x tile (carding) 2 2 43.7 42.5 13.1 26.4 O D L : 0.5 /im Measurements by Scanning Electron Microscope ( i HWBUI0008793 BIOMEDICAL INFORMATION SERVICE (MON) 10. 1 3' 03 1 6: 23/ST. 16:1 O/NO. 4862209230 P 33 (, i A comparison of light microscopy and transmission electron microscopy 329 T able 5. (continued) a E OrO--Ns a T3 sg I 4ll |V g v; CM SCOM +3 fN G 8 X> A G < < Su 1o <Ossc 0 'Z < =t 7.g 0 . *r0os ni-- w 4- ss o0 i<s=* o .oS s 0 Si 3 X o Sj SO 00 xI0zs< c3ss0?,, XT0S&g<.do0 Q s w H o w OS ^ wO cn 4 e"i 00 cm *0 ON -- v rO --< ^r0- - c- Os c4 =6 O E* 8 i0 1 -a J4S> i o 'E, .0 S-o e^-H Sl|c o 22 Z ON vr"> m-f} $ 06 fi & ZZ On -i ND N On ZX0< 03 wCM 12 "S 3CaoO aG S 60 &0 .s .s s 0 c 2 5s> > 1 ! HWBU10008794 FROM BIOMEDICAL INFORMATION SERVICE (MON) 10. 13' 03 1 6: 23/ST. 6:1O/NO. 4862209230 P 34_ 330 A. Marconi, E. Menichini and L. Paoletti calculate the proportions of submicroscopic fibres and of short fibres (1^5 /rm); in this case too, a remarkable disagreement among the various results is found. Possible causes of the discrepancies reported in Table 5 are: (1) different working processes, and also a different `way of working' for the same process, on which the fibre size distribution can depend (Gibbs and Hwang, 1975; Rendall and Skikne, 1980; Hwang, 1983); (2) the different diameter value adopted to differentiate submicro scopic from microscopic fibres; (3) the different mineralogical origin of the fibre. In particular, when considering only fibres with /> 5 ptm, we found a lower percentage of submicroscopic fibres than any other author. This discrepancy too can be explained by the causes reported above. We think that a loss of fibres during the preparation for TEM is unlikely; in fact, by comparing with LM results, we verified that no loss of microscopic fibres longer than 5 /an occurred (as reported in Table 5) and there seems to be no reason to suppose a selective loss of thin long fibres (ds 0.3 ftm, /> 5 jum), as could be possible for thin but very short fibres. Finally, it is clear that owing to both the discrepancies in reported results and the low number of studies undertaken, and especially of samples examined, it is not yet possible to derive generally valid conclusions about the percentage of submicroscopic fibres. As regards the correlation between the two instruments when only the microscopic fibres are considered at TEM, Hwang and Wang (1983) report a correlation coefficient r=0.87 (25 samples), which is in very good agreement with ours (see Table 3: r=0.90; 23 samples with LM density of 40-1000 fibres mm-2). CONCLUSIONS The results reported in this paper show a very good agreement between the two techniques considered and, therefore, suggest the possibility of deriving, for each working process, conversion factors which would allow a rough evaluation of the total number of airborne chrysotile fibres from LM results in occupational environments. In fact, LM might be a suitable instrument for routine control of asbestos dust, but it is unable to detect all the fibres which are considered to be harmful. The differences among the few results available up to now for this type of comparison indicate that the problem ofcorrelating the two measurements is far from being solved. Consequently, it is not correct to determine the exposure to all the fibres considered as harmful, by using conversion factors derived from other surveys. It seems, therefore, clear that further systematic studies on a wider number of samples from different working processes are required to determine such factors. Acknowledgements--We thank Mr M. Maccione for his effective technical collaboration in LM analysis and Mr S. Notargiacomo for his help in TEM counts. Air sampling in the brake and clutch factory has been performed by the Environmental Survey Group of the Laboratory di Igiene e Profilassi--Reparto chimico, Ascoli Piceno. REFERENCES AIA (1979) Reference method for the determination of airborne asbestos fibre concentrations at workplace by light microscopy (membrane filter method). Asbestos international Association, London. AIA (1980) Third international colloquium on dust measuring technique and strategy. Cannes, 10--12 June 1980. Asbestos International Association, London; Association Frangaise de PAmiante, Paris. i [ ! HWBU10008795 FROM B.I OMEDI CAL INFORMATION SERVICE (MON) 10. 13' 03 16: 23/ST. 1 6 1 O/NO. 4862209230 P 35 md size 80; :roIn i Of by for ! of ms , as the nd lid pic on In is :es he , it ng er re nd ;en ;0, A comparison of light microscopy and transmission electron microscopy 331 pATTisn, S., Comba, P., Munafo, E. and Serio, A. (1981) Distribuzione di fibre di asbesto sui filtri nei prelievi secondo il metodo AIA: adattamento della curva di Poisson. Fifth National Congress on Industrial Hygiene, Rome, 15-16 December 1981. University Cattolica del Sacro Cuore, Rome. Beckett, S. T. and Jarvis, J. L. (1979) A study of the size distributions of airborne amosite fibres in the manufacture of asbestos insulating boards. Ann. occup. Hyg. 21, 273-284. Bignon, J., Sebastien, P. and Gaodichet, A. (1978) Measurement of asbestos retention in the human respiratory system related to health effects. In Proc. Workshop on Asbestos: Definitions and Measurements Methods, Gaithersburg, 18-20 July 1977. National Bureau of Standards Special Publication 506. Chatfield, E. J. (1979) Measurement of asbestos fibres in the workplace and in the general environment. In Short Course in Mineralogical Techniques of Asbestos Determination (Edited by Ledoux, R. L.), pp. 111-163. Mineralogical Association of Canada, Toronto. Gibbs, G. W. and Hwang, C. Y. (1975) Physical parameters of airborne asbestos fibres in various work environments--preliminary findings. Am. ind. Hyg. Ass. J. 36, 459-466. Gibbs, G. W. and Hwang, C. Y. (1980) Dimension of airborne asbestos fibres. In Biological Effects of Mineral Fibres (Edited by Wagner, J. C.), Vol. 1, pp. 69-77. IARC Scientific Publication No. 30.1ARC, Lyon, 1980. Hwang, C. Y. (1983) Size and shape of airborne asbestos fibres in mines and mills. Br. J. ind. Med. 40, 273-279. Hwang, C. Y. and Gibbs, G. W. (1981) The dimensions of airborne asbestos fibres: 1. Crocidolite from Kuruman area. Cape Province, South Africa. Ann. occup. Hyg. 24, 23-41. Hwang, C. J. and Wang, Z. M. (1983) Comparison of methods of assessing asbestos fiber concentrations Arch, environ. Health 38, 5-10. Leineweber, J. P. (1978) Statistics and the significance of asbestos fibre analyses. In Proc. Workshop on Asbestos: Definitions and Measurements methods, Gaithersburg, 18-20 July 1977, pp. 281-294. National Bureau of Standards Special Publication 506. Lynch, J. R., Ayer, H. E. and Johnson, D. L. (1970) The interrelationship of selected asbestos exposure indices. Am. ind. Hyg. Ass. J. 31, 598-604. Millette, J. R., Clark, P. J. and Pansing, M. F. (1978) Sizing of particulates for environmental health studies. Scanning Electron Microscopy (1978) (Edited by Johari, O.), pp. 253-258. IIT Research Inst., Chicago, U.S.A. NIOSH (1977) Review and evaluation of analytical methods for environmental studies of fibrous particulate exposures. DHEW (NIOSH) Publication No. 77-204. U.S. Dep. HEW, NIOSH, Cincinnati, U.S.A. Pooley, F. D. and Clark, N. (1980) A comparison of fibre dimensions in chrysotile, crocidolite and amosite particles from samples of airborne dust and from post-mortem lung tissue specimens. In Biological Effects ofMineral Fibres (Edited by Wagner, J. C), Vol. 2, pp. 79-86. IARC Scientific Publication No. 30. IARC, Lyon, 1980. Po rr, F. (1978) Some aspects on the dosimetry of the carcinogenic potency of asbestos and other fibrous dusts. Staub-Reinhalt Luft 38, 486-490. Rajhans, G. S. and Bragg, G. M. (1975) A statistical analysis of asbestos fiber counting in the laboratory and industrial environment. Am. ind. Hyg. Ass. J. 36, 909-915. Rendall, R. E. G. and Skikne, M. I. (1980) Submicroscopic fibres in industrial atmospheres. In Biological Effects of Mineral Fibres (Edited by Wagner, J. C.), Vol. 2, pp. 837-843. IARC Scientific Publication No. 30. IARC, Lyon, France, 1980. Stanton, M. F., Layard, M., Tegeris, A., Miller, E., May* M. Morgan, E. and Smith, A. (1981) Relation of particle dimension to carcinogenicity in amphibole asbestoses and other fibrous minerals. J. natn. Cancer Inst. 67, 965-975. Timbrell, V., Griffiths, D. M. and Pooley, F. D. (1971) Possible biological importance of fibre diameters of South African amphiboles. Nature, Load. 232, 55-56. Walton, W. H. (1982) The nature, hazards and assessment of occupational exposure to airborne asbestos dust; a review. Ann. occup. Hyg. 25, 117--247. Walton, W. H. and Beckett, S. T. (1977) A microscope eyepiece graticule for the evaluation of fibrous dusts. Ann. occup. Hyg. 20, 19--23. Winer, A. A. and Cossette, M. (1979) The effect of aspect ratio on fibre counts: a preliminary study. Ann. N.Y. Acad. Sci. 330,661-672. .ce i HWBUI0008796