Document 6Ba7xyYo3eXZ68LjK4o4OKy9o

Calibration Standards for Counting Asbestos ZCOW- 7 05 I % * VI.Lawrence Ortiz J.Harry Ettinger Charles i- Fairchild -#30- Health Division Los Alamos Scientific Laboratory HeUniversity of California, Los Alamos, w Mexico* Asbestos dust has long been recognized as an industrial hazard resulting in asbestos is, and on the BASIS Oe MORE recent informa tion, LUNG CANCER. INCREASING AWARENESS 0e THIS SERIOUS OCCUPA TIONAL RISK HAS RESULTED IN THE INSTITUTION OF AN EMERGENCY TEMPORARY STANDARD BY THE SECRETARY OF LABOR, CONCERNING HUMAN EXPOSURE TO i asbestos. This emergency standard, published in the Federal Register ' AND EFFECTIVE AS OF JULY, 1972 -- DICTATES THAT THE MAXIMUM PER- : MISSIBLE OCCUPATIONAL EXPOSURE TO AIRBORNE ASBESTOS FIBERS SHALL ! NOT EXCEED FIVE (5) FIBERS LONGER THAN 5 PM PER CUBIC CENTIMETER OF AIR. IN ADDITION, THE FEDERAL REGISTER FURTHER STIPULATES A REDUC TION IN THIS MAXIMUM PERMISSIBLE LEVEL TO 2 FIBERS/CC, BY JULY, 1976. TLV' TWA'These s abe 8-hour s as determined by a microscopic fiber COUNTING PROCEDURE. THE STANDARD ALSO*SETS A SHORT TERM MAXIMUM EXPOSURE LEVEL OF 10 FIBERS/CC. TO ASSURE UNIFORM FIBER COUNTING PROCEDURES BY THE VARIOUS HEALTH PROTECTION AGENCIES MONITORING EXPOSURE TO ASBESTOS, A LABGRAi 1 TORY PROFICIENCY COUNTING PROGRAM, CALLED THE PROFICIENCY ANALYTICAL Testing or PAT program, was initiated by the National Institute for NICSH.Occupational Safety and Health or This program will permit * Work supported by the National Institute for Occupational Safety and Health, Cincinnati, Ohio, Project R-072, and Under Contract W-7405-ENG-36 with the U.S. Atomic Energy MAST, commission. J"Al ncHNKAt DtSTfaOUTICN OF THIS DOCUMENT US WN LIMIT INFORMATION SERVICE fl - 0 5 7 6 GAF 15674 QUANTITATIVE EVALUATION OF ASBESTOS MONITORING PROCEDURES BY VARIOtt DESIGNATED STATE AGENCIES/ AS WELL AS HELP DEFINE THE VARIABLES IN T EHERENT IN FIBER COUNTING. h BASIS OF THIS STANDARDIZATION PROCES: IS A MICROSCOPIC FIBER COUNTING METHOD ORIGINALLY DEVELOPED BY THE U.S. Public Health Service. -Briefly/ this technique involves air SAMPLING THROUGH A MEMBRANE FILTER/ CLEARING THE MEMBRANE/ AND COUNTING FIBERS LONGER THAN 5 PM UNDER PHASE CONTRAST OPTICS AT MAGNIFICATIONS OF 40(M30 X. The Los Alamos Scientific Laboratory or LASL -- at the request OF NIOSH -- DEVELOPED A TECHNIQUE FOR THE PREPARATION OF MULTIPLE MEMBRANE FILTER SAMPLES CONTAINING 'IDENTICAL AND -PREDICTABLE' CONCENTRATIONS OF ASBESTOS FIBERS/ WITH CONSISTENT PARTICULATE BACKGROUNDS. THESE SAMPLES WERE TO BE REPRESENTATIVE OF SAMPLES OBTAINED FROM MANY WORK SITUATIONS. THF NEEDS OF KIOSH, COUPLED WITH THE INHERENT HEALTH HAZARD OF AIRBORNE ASBESTOS DUST DICTATED THAT A SIMPLE/ EASILY CONTROLLABLE AND SAFE TECHNIQUE BE USED. THE APPROACH CHOSEN WAS THE FILTRATION OF MEASURED VOLUMES OF DILUTE LIQUID DISPERSIONS OF MILLED CHRYSOT1L ASBESTOS/ AND AL2O3 THROUGH MEMBRANE FILTERS. iHE RESULTANT DE POSITS ON THE MEMBRANE FILTER SURFACE CAN THEN BE MONITORED BY THE USPHS FIBER COUNTING TECHNIQUE. CHRYSOTILE PRODUCED BY THE INTER NATIONAL Union Concerning Cancer (UICC) was selected as the basic MATERIAL BECAUSE/ (1)/ CHRYSOTILE IS THE MOST WIDELY USED FORM OF ASBESTOS IN INDUSTRY/ AND (2), UICC MATERIAL IS READILY AVAILABLE/ WELL CHARACTERIZED/ US'D EXTENSIVELY IN MEDICAL RESEARCH, AND HAS TBEEN PROPOSED AS A WORLDWIDE STANDARD FOR RESEARCH PURPOSES. he UICC MATERIAL REQUIRES NO MULING TO OBTAIN THE DESIRED RANGE OF 2 GAF 15675 ASBESTOS FIBERS; AS DO OTHER SOURCE MATERIALS. A BRIEF DESCRIPTION OF THE DISPERSION-FILTRATION TECHNIQUE follows: Accurately weighed quantities of hilled chrysotile are DILUTED TO FIXED VOLUMES WITH REAGENT GRADE TOLUENE. THE ASBESTOS IS THEN DISPERSED THROUGHOUT THE LIQUID BY SONICATION (60 HZ) AND INTERMITTANT; VIGOROUS HAND SHAKING; UNTIL UNIFORM TYNDALL LIGHT SCATTER CONDITIONS EXIST. A UNIFORM DISPERSION HAS MANY SMALL FIBERS; BUT VERY FEW LARGE FIBROUS CLUMPS. SEPARATE DISPERSIONS OF AL2O3 PARTICLES ARE PREPARED IN SIMILAR FASHION. The LOADED FILTERS ARE THEN PREPARED BY SLOW; CONTROLLED VAC UUM FILTRATION OF MEASURED VOLUMES OF THE PREPARED DISPERSIONS; THROUGH INDIVIDUAL MEMBRANE FILTERS. THE FILTRATION VOLUMES ARE ADJUSTED TO OBTAIN THE DESIRED FIBER AND PARTICULATE DEPOSIT CONCEN TRATIONS ON THE MEMBRANE SURFACE. MULTIPLE FILTERS CONTAINING 4 LEVELS OF ASBESTOS CONCENTRATIONS RANGING FROM 500 TO 1700 FIBERS/MM^; A RANGE WHICH IS TYPICAL OF INDUSTRIAL FIELD SAMPLING WERE PREPARED AND SENT TO N!0SH. SAMPLE WEDGES FROM EACH FILTER WERE RETAINED AT Los Alamos. Duality control was accomplished by multiple counters; each COUNTING THE ^AME-WEDGES TAKEN FROM REPRESENTATIVE PORTIONS OF THE filters. Rules for counting fibers were consistent with jhose SPECIFIED BY NI0SH. A SAMPLE OF TYPICAL FIBER COUNT DATA OETAINED ON FILTERS PRE PARED BY THE DISPERSION-FILTRATION TECHNIQUE IS SHOWN IN THE FIRST SLIJBJL- This slide illustrates typical count data obtained on SEPARATE FILTER SAMPLES CONTAINING FOUR (4) DIFFERENT ASBESTOS CON CENTRATIONS. The SIGNIFICANT FEATURE ILLUSTRATED BY THESE RESULTS 3 GAF 15676 ARE THE SPECIFIC LEVELS ACHIEVABLE BY THE DISPERSIOH/FILTRATION TECHNIQUE. An ADDITIONAL EXAMPLE (SLIDE 2) OF COUNT DATA SUMMARIZES RESULTS OBTAINED ON TEN "IDENTICAL" FILTERS. The COEFFICIENTS OF VARIATION, THAT IS, THE STANDARD DEVIATIONS DIVIDED BY THE MEAN, RANGE FROM TO + 2fiZ. Specifically, the variation between these ten samples, counted LASLby 5 counters is + 17Z. In this idealized situation, each of THE 5 COUNTERS USED THE SAME MICROSCOPE AND ALL COUNTED THE SAME FILTER WEDGES. REPETITION OF THIS EXPERIMENT AT A LATER DATE PRO DUCED A VARIATION BETWEEN 4 COUNTERS OF + 20Z. EXTENSIVE FIBER COUNT DATA FROM THIS QUALITY CONTROL PROGRAM DEMONSTRATES THAT UNDER IDEAL CONDITIONS (i.E., SEVERAL EXPERIENCED COUNTERS COUNTING THE SAME WEDGES AND USING THE SAME OPTICS) VARIATIONS BETWEEN COUNTERS AS LARGE AS 5QX ARE OBSERVED, WITH TYPICAL VARIATION ON THE ORDER of + 301. Moreover, limited unpublished data contributed by counting FACILITIES PARTICIPATING IN THE NIOSH PAT PROGRAM INDICATES THAT LARGER VARIATIONS MAY BE EXPECTED WHEN ADDITIONAL VARIABLES ARE INTRODUCED. THESE VARIABLES INCLUDE COUNTERS OF VARIED EXPERIENCE! DIFFERENT SAMPLE WED6ES; DIFFERENT OPTICS; VARIED LEVELS OF SKILL IN LABORATORY AND MICROSCOPY TECHNIQUES; THICKNESS OF BOTH SLICE AND COVER SLIP; RELATIVE SIZE OF COUNTING FIELD; RELATIVE PARTI CULATE AND/OR FIBER CONCENTRATION; AND QUALITY OF OPTICAL MOUNT. All these are significant factors which affect fiber count results. Of all variables, the two most important are the quality of micro scope AND ATTITUDE OF COUNTER. A MAJOR REASON FOR LARGE VARIATIONS IN COUNT LIES IN THE FACT THAT FIBERS COLLECTED ON MEMBRANE FILTERS DO NOT ALWAYS APPEAR AS 4 GAF 15677 DISTINCT ENTITIES. THE NEXT SLIDE. SHOWS A TYPICAL SAMPLE OF UICC CHRYSOTILE WITH AN AL2O3 BACKGROUND. A LARGE PERCENTAGE OF FIBER DIAMETERS/ ARE NEAR THE LIHiT OF OPTICAL RESOLUTION. THIS PHOTO GRAPH PROVIDED A MAGNIFICATION OF APPROXIMATELY 50C X. FURTHER DIFFICULTY IS CAUSED 2Y THE FACT THAT ALL FIBERS CO NOT LIE IN THE SAME FOCAL PLANE/ NECESSITATING A CAREFUL 3-DIHENSIONAL SEARCH OF EACH COUNTING FIELD. A SINGLE EXPERIMENT WAS CONDUCTED IN AN ATTEMPT TQ QUANTITATE THIS BUILT-IN VARIABLE OF THE MICROSCOPIC FIBER-COUNTING TECHNIQUE. Four experienced counters each counted two separate wedges from REPRESENTATIVE FILTERS PREPARED BY THE DISPERSION/FILTRATION TECH NIQUE. In COUNTING THESE WEDGES/ EACH COUNTER COUNTED THE IDENTICAL OPTICAL FIELD AND RECORDED HIS COUNTS WITHOUT INFORMING THE OTHER COUNTERS OF HIS RESULTS. THE NEXT SLIDE SUMMARIZES THESE DATA ( OBTAINED UNDER EXTREMELY IDEALIZED CONDITIONS. THAT IS/ SEVERAL EXPERIENCED COUNTERS/ USING THE SAME OPTICS/ EACH COUNTED IDENTICAL VIEWING FIELDS. VARIATIONS OF + 132 AND + 222 ARE OBSERVED/ WHICH IS CONSISTENT WITH THE VARIATIONS PREVIOUSLY NOTED BETWEEN'"IDENTICAL SAMPLES* PREPARED FOR THE NIOSH PAT PROGRAM. THIS INDICATES THE CON SISTENCY OF THE PREPARATION TECHNIQUE AND SUGGESTS A MINIMUM ERROR WHICH SHOULD BE CONSIDERED WHEN COMPARING COUNT DATA. The IMPORTANCE OF THE QUALITY OF MICROSCOPIC OPTICS IS DIPECTLY RELATED TO THE INTRINSIC NATURE OF THE MINERAL ASBESTOS (ESPECIALLY CHRYSOTILE). IT IS WELL KNOWN THAT SAMPLES OF AIRBORNE ASBESTOS MAY CONTAIN MANY FIBERS OF A SIZE V.HICH LIE WELL BELOW THE LIMIT OF RESOLUTION OF EVEN THE BEST OPTICAL MICROSCOPES (ESPECIALLY AT *100 to 500 X). Samples of asbestos containing individual fibers 5 GAF -'5678 vv\ WITH DIAMETERS AS SMALL AS 250 - 300 ANGSTROMS HAVE BEEN MEASURED USING THE ELECTRON MICROSCOPE. THE NEXT SLIDE SUMMARIZING (EM)COMPARATIVE FIBER COUNTS OBTAINED USING ELECTRON MICROSCOPY versus Optical Microscopy (OK) illustrates a significant feature. The final entry in this table is of specific interest. These fiber (EH/OM) 22 23count ratios of or indicate a possible magnitude of error associated with counts obtained by Optical Microscopy when TOTAL FIBER CONCENTRATIONS ARE ASSESSED. THESE COMPARATIVE COUNTS WERE OBTAINED FROM A SINGLE FILTER SAMPLE PREPARED BY THE DISPERSION/ FILTRATION TECHNIQUE. ELECTRON MICROSCOPE COUNTS WERE OBTAINED BY TRANSFERRING A PORTION OF THE TEST FILTER ONTO Efl GRIDS/ PHOTOGRAPH ING RANDOM FIELDS AND COUNTING THE TOTAL AREAS PRESCRIBED. OPTICAL MICROSCOPE COUNTS MERE OBTAINED FROM A DIFFERENT SECTION OF THE SAME TEST FILTER BY THE CONVENTIONAL PHASE CONTRAST MICROSCOPIC FIBER AsCOUNTING TECHNIQUE. NEAR AS PRACTICAL/ THE SAME COUNTING CRITERIA APPLIED TO EACH TECHNIQUE. AN IMPORTANT ASPECT/ WHICH SHOULD NOT BE IGNORED/ IS THAT THE RELATIVE HEALTH HAZARD ASSOCIATED WITH SMALL FIBERS IS AT PRESENT UN KNOWN. IN FACT/ THESE SMALL FIBERS ARE MORE APT TO BE 'RESPIRABLE*. AND MAY CONSTITUTE A GREATER HAZARD. - woici. - j>MM* M MMMN M MM MtMaM Imm am Ite V*m4 U*m ammi Imv Cammmmaa, am m* * mr 1 ----- -------... --mat mmmm. iMniii iii. at tmm tipWyin, -- --- r| ItoMMr aa mgammamt im M* wnmr, ram. Umarnam m iiHUiib at mr \almamtm. mwim RmBwI fmm aa rajtaaaam MM m m 6 GAF 15679 In conclusion: Evaluation of the dispep.sion/filtration technique indicates THAT THE METHOD KILL PRODUCE ASBESTOS LOADED MEMBRANE FILTERS IN A CONTROLLABLE AND PREDICTABLE FASHION. SAMPLES PREPARED BY THIS TECHNIQUE CAN BE USED TO EVALUATE THE PROFICIENCY OF PERSONNEL RESPONSIBLE FOR FIBER COUNTING. Minimum errors inherent in the "standard" fiber-counting tech nique HAVE BEEN CUANTITATED AND INDICATE THAT MICROSCOPIC FIBER COUNTING MAY UNDERESTIMATE THE POTENTIAL RISK OF HUMAN EXPOSURES TO ASBESTOS DUST. EVEN UNDER IDEAL CONDITIONS MICROSCOPIC FIBER COUNTING MAY BE TERMED AS INADEQUATE, SINCE THE VARIABLES IN THE TECHNIQUE TEND TO REDUCE THE NUMBER OF FIBERS COUNTED AND THUS, AGAIN, UNDERESTIMATE THE ACTUAL EXPOSURE. These limitations should be considered in connection with the PROPOSED REDUCTION FROM 5 FIBERS/CC TO 2 FIBERS/CC IN 1976. CON CURRENTLY, IT IS APPROPRIATE TO CONSIDER THE DEVELOPMENT OF ALTERNATE METHODS TO PROVIDE MORE ACCURATE MONITORING OF ASBESTOS EXPOSURE. Moreover, the course of development of accurate methods depends upon THE PARALLEL INVESTIGATION OF THE MECHANISM OF FIBROUS PNEUMOCONIOSES. Until improvements in monitoring methods are realized, involved agencies AND INDIVIDUALS SHOULD BE AWARE OF THE UNCERTAINTIES IN THE PRESENT MEHOD. GAF 15680 7 ASBESTOS COUNT DATA SERIES 149 - 152 Sample and Fibers/mm^ Filter Area Counter 1 2 3 4 15-A 15-B 15-C 706 961 1255 863 1255 1216 1000 1627 2216 470 627 804 Average: a: Coef.Var.: 760 227 0.30 1118 426 0.38 1373 598 0.44 15-D 2294 1882 3039 1353 2142 711 0.33 /\ SLIDE 1 GAF 15681 SAMPLE 1 2 3 4 5 6 7 8 9 10 modl;>hK> Lu'UNi UA1A - StlULS 12j INDIVIDUAL FIBER COUNT DATA - TEN "IDENTICAL" ACounter and Fibers Counted/mm^ Filter A 1207 1586 1448 1310 1793 1586 1586 1793 1241 1414 B 1414 1690 1448 1828 1724 1759 1724 1655 1655 1690 C 1172 966 896 1310 1310 1552 966 1310 1276 1103 SLIDE 2 Average: : 1496 209 1659 131 COEF. OF VAR: 0.14 0,08 LASLMean of five (5) counters: 1496 Standard Deviation (c): 203 Average coef. of var, 0,19 -7 1186 204 0.17 GAF 15682 12.j/UW-OIUO loUH I UAl/l - 5LK1LS JIVIDUAL FIBER COUNT DATA - TEN "IDENTICAL* FILTERS Counter and Fibers Counted/mm^ Filter Area B 7 1014 6 1690 8 1448 0 1828 3 1724 5 1759 3 1724 3 1655 l 1655 I 1690 1659 131 0.08 1496 203 0.19 C 1172 966 1310 1310 1552 966 1310 1276 1103 D 1621 1931 --------(262^) 2069 2655 1310 1103 1103 1517 1000 1186 204 0.17 1693 610 0.36 E 1828 - 1 1552 - i-- 1690 1517 1724 - 1069 - r 1241 - n 1345 1276 - =i 1207 - i 1445 254 0.18 JO\. losNU/Blamo* nllda IMr*Mfy i mt--***f i (mwm /\ GAF 15683 SUDS 3 // GAF 15684 Asbestos Count Data Four Counters Each Counting the Same Twenty Five Fields COUNTERS SAMPLE AND FIBERS/k.H FILTER AREA 1 2 3 4 Avg. A 1034 862 1000 1414-- 0 -aT * 1078 236 COEF. OF VAR. 221 B 1966 2276 " ` 1862 2759 . 3. m +182 y JosVXXalamos scientific laboratory I h Uni*r*i>7 *1 California // \ SLIIZ 4 A* - GAF 15685 CWPARATIVE FIBER COURTS Electron Microscopy vs Optical Microscopy COUNTING MODE SAMPLE AND FI3ERsWfIBER ARE/ OPTICAL microscopy LASLAverage of Four (4) Counters; Fibers >5m* Total Fibers: A 1078 1079 B 2216 2218 electron microscopy Fibers >5r: Total Fibers: *Em/om Em/om Fibers >5u*: Total Fibers 3695 25261 3.4 23.4 *Em/om Fiber counts obtained by electron microscopy Fiber counts obtained by optical microscopy JL losWolo met /V 7162 4968* 3.2 22.4 losMValamo* t\ SLI18 5 13 GAF 15686