Document xdvQy5v82Ja1oba79kbezvLrQ

i evibeen o dis- .s for S T n n r,'.f)96 J Morwork* 'terenee peiown, 19T0, pp allty of id Med evskiys rkers in j iami i its reationai *sbu.rg jversity *. mine* ?nce on rnnee. i of Dm- :os in a London. rvey of irds. Br -oncen-5:1271- ;n as- IARC jDestos. id et al Dust-Fiber Relationships in the Quebec Chrysotile Industry Graham William Gibbs, MSc, PhD, Maurice LaChance, Eng, Montreal Epidemiological studies o< Quebec chrysotile miners and millers have related various Indices of health to dust expo sure, as measured with the midget Implnger. The interpretation ot these rela tionships In terms ot fiber necessitates the conversion of midget Implnger (par ticle fiber) counts to membrane filter (fiber) equivalents. An Investigation In which 87 side-by-side midget Impingermembrane filter samples were taken at five mines and mills showed that the cor relation was poor and no single conver sion factor was justified. Until more de tailed Information on the relationships between midget Implnger and membrane filter counts can be obtained. It Is recom mended that safety standards, at least In this Industry, should continue to be based on dust counts, for which there la consid erable epidemiological support rather than on fiber counts, tor which there Is no . direct evidence. In Canada and the United States, for many years routine measure counts derived from the membrane filter technique.*10 In North America, ments of airborne concentrations of there are no asbestos standards based asbestos have been made by the on gravimetric or other methods. midget impinger method. In conse Unfortunately, safety standards quence, the few epidemiological stud have little validity unless based on ies that have related health of as epidemiological evidence and, as has bestos workers to dust exposure have, been stated, this evidence has of ne of necessity, been forced to use con cessity been related to midget im centrations measured by this tech pinger counts. There might be no se nique. This has been true of surveys rious difficulty in converting dust in Quebec of mortality, roentgen- counts to fiber counts if some reason ographic appearance, respiratory able constant relationship were found function and symptoms in chrysotile to exist between the two types of miners and millers,'1 and also of sur measurement. Ayer et al" measured veys of production workers in the the ratio between the membrane fil United States.*-7 ter count and the midget impinger Because of its relatively low effi counts for textile plants, and found ciency for the collection of fibers, the that ratios differed for different oper midget impinger is not ideal for the ations in the same plants and for sim determination of airborne asbestos ilar operations in different plants. fiber concentrations.' This limitation This report describes a preliminary and the belief that fiber is the essen investigation of this problem in the tial cause of disease have led to the Quebec mining and milling industry. recommendation that standards for exposure should be based on fiber Methods Submitted for publication June 8, 1973; ac cepted July 31. From the Department of Epidemiology and Health, McGill University, Montreal (Dr. Gibbs) and the Quebec Asbestos Mining Association, Thetford Mines, Quebec (Mr. LaChance). Reprint requests to Department of Epidemiol ogy and Health, McGill University, 3775 Univer sity St. Montreal H3A 2B4 (Dr. Gibbs). Surveys in August 1971 and 1972 were conducted at up to nine selected sites in each of five mines and mills of the Quebec chrysotile mining and milling industry. A total of 87 pairs of membrane filter sam ples and midget impinger samples were collected side-by-side at each location with sampling rates of 2 liters per minute and Arch Environ Health/Vol 28. Feb 1974 Dust-Fiber Relationships/Gibbs A LaChance 69 2.8 liters per minute, respectively. The fil ter and impinger inlets were both at ap proximately 1.5 meters above the floor. Sampling times ranged from 5 to 30 min utes. and were the same for both methods. The membrane filters were 37-mm plain white (pore site, 0.8(i), and the samples were taken with the entire filter exposed. These samples were fixed, mounted with use of the British method,'1 and counted at 500 x magnification with a Zeiss standard binocular WL microscope with phase con trast optics. One observer counted all sam ples; 20 random fields, of area 0.0055 sq mm. were observed, and all fibers greater than 5(i in length were counted. A fiber wa3 defined as any particle with a length-tobreadth ratio greater than 3:1. The midget impinger method, essentially that ap proved by the Asbestos Textile Institute/1 used isopropyl alcohol, and particles were counted at 100 x magnification with iightfield illumination. These samples were counted by two observers. One observer < M.L.i. who had made all previous routine midget impmger surveys of the Quebec mining and milling industry, counted 51 of the total number of samples, and the other observer (trained by M.L.i counted the remainder. Counts made on the same samples by the two observers were within 15% Results A plot of all 87 membrane filter and midget impinger counts is shown in the Figure. As membrane filter sam ples were collected over the same sampling period as the midget im pinger samples, the fiber concentra tions on the filter were not always op timal (2 to 6 fibers per field). This was A plot of membrane filter and midget impmger counts; mpcf represents millions of par ticles per cubic foot. owing to the low ratio of fiber to other particulates in certain processes and the time limit imposed by evapo ration of isopropyl alcohol from the impinger. The linear correlation coef ficient for 56 samples with more than 1 fiber per field was 0.32. and, after a logarithmic transformation of both variables, the correlation was in creased to 0.45. For 31 samples with less than 1 fiber per field, the linear correlation was very dose to zero (-0.03), and the correlation of loga rithmically transformed data was 0.25. The highest correlation obtained for these data (0.45) demonstrated an association between the results ob tained by the two methods. However, these correlations suggest that for ail mines the regression lines are unsat isfactory for the prediction of fiber counts from impinger counts, as the improvement in prediction for the best correlation (0.45) is only 13% better than a conversion obtained at random. A 50% improvement of pre diction would require a correlation co efficient of 0.87. If membrane filter and midget im pinger counts were considered by work area (Table), it was clear that the ratios of the two in some mines were of a different order from those in others where the same process was Ratios of Membrane Filter Counts to Midget Impinger Counts Mine A B C D E All mines Overall Mean 45 (28) 11 4 (18) 21.9 (18) 5.9 (11) 1.7 (12) IMdergroqgl 17 (D 1.7 (1) Open Pit Drill ShoveJ ao 54 (2) 12) 91 24 (2) (2) 52 79 12) (2) 06 05 in (1) 53 46 (7) (7) Dryer and Crusher Dryer 3.2 (2) 29 8 (2) 15 9 (2) 20 13) 04 (2) 95 (11) Crusher 25 (6) 37 (4) 12 5 >4) 3.3 (1) 1.1 (2) 53 (17) Rock Screening 49 (5) 29 8 (2) 32 6 (2) 10.4 (2) 03 (1) 14.2 (121 Mill Fiber Screening 35 (5) 85 (2) 47 4 (2) 43 (2) 53 (3) 11 0 (14) Bagging 8.1 (3) 61 (2) 31 1 (2) 38 (3) 4.5 (D 10 4 (11) Storage 3.3 (2) 2.0 (2) 22 8 (2) 08 mJ 8.1 (7) J * Membrane filter counts were measured in fibers per milliliter, midget impinger counts in millions of particles per cubic foot. Numbers m parentheses are numbers of membrane filter-midget impmger pairs empi with sam>. a; were lion, stra: tor call 'a suit Aver Th ease min:: disea poss: infer' the : each conn: sites, effect. count. based amirustand. count. W:;hc mat:r. GO --I 70 Arch Environ Health Vol 28. Feb 1974 Dust-Fiber Relationships/Gibbs A LaChanc* Arrr- =- 35 oar- ver. r all .sat'iber the the m d at pren co- t im1 by that !ines hose was I employed. In addition, ratios varied with stage of production within the same mine and mill Conclusion* Although only 87 pairs of samples were collected in this pilot investiga tion. these were sufficient to demon strate that no single conversion fac tor could be applied to all mines or to all work areas within a mine. This re sult coincides with the findings of Ayer et ai" in textile plants. Thus, the conversion of dust-dis ease relationships for the Quebec mining and milling industry to fiberdisease relationships does not seem possible at the present time. Further information on the relationships of the two methods at each site within each plant is required. With adequate counts of pairs of samples at selected sites, and with assumptions as to the effects that dust control and changes in ore have had on impinger-filter counts, dose-response relationships based on fiber exposures could be ex amined. and it is possible that safety standards based rationally on fiber counts could then be established. Without this, there is no way of esti mating with any accuracy the total fiber exposure of persons in the Que bec asbestos mining and milling in dustry, or indeed of deciding whether the dose-response relationships would be better or worse than those ob tained from particle counts. Until sat isfactory conversion factors are de rived, it would seem best to base standards, at least for mines and mills, on particle counts, for which there now exists a reasonable body of evidence. A similar argument probably applies equally to all epidemio logical studies of asbestos workers in North America in which exposure has been considered. This investigation was supported by a grant from the Institute of Occupational and Environ mental Health of the Quebec Asbestos Mining .Association. H. Hui. BSc. and A. Corsillo earned out the field surveys and performed the midget irapmger and memorane filter counts References I McDonald JC. et al Mortality in the chryso tile asbestos mines and mills of Quebec Arch Environ Health 22:677-686, 1971. 2. Rossiter CE. et al. Radiographic changes in chrysotile asbestos mine and mill workers of Quebec Arch Environ Health 24:388-400, 1972. 3. Beddake MR. et aL Lung function in diryso tile asbestos mine and mill workers of Quebec Arch Environ Health 24:401-409, 1972 4 McDonald JC. et ai: Respiratory symptoms :n enrysotile asbestos mine and mill workers of Quebec .4rcA Environ Health 24-358-363. 1972 5. Gibbs GW, LaChance M: Dust exposure in the chrysotile asbestos mines and mills of Que bec ArcA Environ Health 24:189-197, 1972. 6. Enterime P, Weill H: Asbestotis in asbestos cement workers. Read before the IARC Confer ence on the Biological Effects of Asbestos, Lyon, France, 1972 7. Enterline P. De Coufie P, Henderson V: Mortality in relation to occupational exposure in the asbestos industry. J Occup Med 14:897-903, 1972. S. RendaJl REG. Van Sittert GCH: "Hie suit ability of convention..! dust instruments for sampling asbestos dust clouds, in Shapiro HA led): Pneumoconiosis: Proceedings of the Inter national Conference, Johannesburg, 1969. Cape Town. South Afnca. Oxford University Press, 1970, pp 74-80. 9. Standard for Asbestos Dust Concentration for i'se mth Asbestos Regulations 1969, technical note 13. London. Dept of Employment and Productivity. Her Majesty's Factory Inspectorate. 1970. 10 Threshold Limit Values for Substance tn Workroom Air Adopted by ACGIH for l97t. Cin cinnati, American Conference of Governmental Industrial Hygienists, 1972, p 37 11 Ayer HE. Lynch JR, Fanney JH A com parison of impinger and membrane filter tech niques for evaluating air samples in asbestos plants. Ann HY Acad Set 132:274-275, 1965. 12. Hygiene standards for chrysotile asbestos dust. Committee of Hygiene Standards of the British Occupational Society Ann Occup Hyg 11:47-69, 1968. 13 Asbestos Textile Institute Method for Determimng Asbestos Dust Concentration Ap proved by the Asbestos Textile Institute, Air Hygiene and Manufacturing Committee. Pompton Lakes, NJ, Asbestos Textile Institute. 1963. f \ It 1 / * [ r I bers `n Chance *rch Environ Health/Vol 28, Feb 1974 i Dust-Fiber Relaftohships/Gibbs & LaChance 71