Document gwZbZpza3jVzE7prYZZyMO8J

Friday November 4, 1983 Part VII Department of Labor Occupational Safety and Health Administration Occupational Exposure to Asbestos; Emergency Temporary Standard LAM 025239 ABS-008414 51086 Federal Register / Vol. 48, No. 215 / Friday. November 4,1983 / Rules and Regulations DEPARTMENT OF LABOR Occupational Safety and Health Administration 29 CFR Part 1910 [Docket No. H-0330] Occupational Exposure to Asbestos agency: Occupation Safety and Health Administration [OSHA], Labor. action: Emergency temporary standard. SUMMARY: OSHA is issuing an emergency temporary standard (ETS) to reduce the permissible exposure limit (PEL) for asbestos from 2 fibers (longer than 5 micrometers) per cubic centimeter (2 f/cc) as an eight-hour time- weighted average, to 0.5 f/cc. During the period of the emergency standard, employers may use all practicable control methods, such as engineering controls, work practices and personal protective equipment to meet the new limit of 0.5 fibers per cubic centimeter (0.5 f/cc). Training programs are also required to be instituted within 30 days. The basis for this ETS is OSHA's determination that continued employee exposure to asbestos under current conditions that exceed 0.5 f/cc presents a grave danger of developing asbestos- induced cancer and asbestosis to exposed employees, and that an emergency standard is necessary to protect them. The ETS serves also as a proposal to revise the current asbestos standard pursuant to section 6(b) and 6(c) of the Act. This notice also requests comments on the appropriateness of including the - provisions of the ETS as permanent revisions to the asbestos standard. In addition, OSHA will soon publish a separate notice of proposed rulemaking that further explains the issues under consideration for the permanent standard and which raises some additional issues regarding the application of certain provisions of the asbestos standard to the construction industry. That notice will invite public comments and will schedule a rulemaking hearing pursuant to Section 6(b) of the Act concerning the proposed permanent revision to the asbestos standard. ___ dates: The effective date fer this ETS is November 4,1983. Comments and evidence concerning the proposed revisions to the permanent standard made by the ETS must be received on or before January 3.1984. As noted. OSHA will publish a notice of proposed rulemaking shortly that will set due dates for submissions to the asbestos docket for the issues raised therein. ADOMSSCS: Written comments should asbestosis) among currently exposed be submitted to the Docket Officer,. workers which are attributable to Docket No. H-033C. Room S-6212. U.S. cobtlnued exposures under present Department of Labor, Third Street and ' .w&king conditions. OSHA estimated Constitution Avenue. N.W., Washington, f' Ihe number of cancer deaths avoided D.C. 20210. thrbugh lowering of the exposure to 0.5 FOR FURTHER INFORMATION CONTACT: James F. Foster, Director. Office of f/cc, based on 6 months, 1-year, 20-year and 45-year periods of exposures at Information and Consumer Affairs, cuisent levels.* The predicted cancer OSHA. U.S. Department of Labor. Room deaths avoided that result from these N-3637, 200 Constitution Avenue, NW, periods of exposure are respectively 210. Washington. DC 20210. Telephone (202) 428, 5725 and 7815. A more detailed 523-8151. discussion of these estimates is found in SUPPLEMENTARY INFORMATION: Table of Contents the sections. Quantitative Risk Analysis (Section V), Crave Danger (Section III A), and Need for an ETS (Section ill B). I. Introduction II. Legal Authority and Background III. OSHA Rationale for the ETS A. Crave Danger B. Need for an ETS Supporting Documentation: Risks of this magnitude at permitted and actual exposure levels defined a situation which demanded immediate regulatory action. Evaluation of the relevant scientific data, policy considerations and exposure patterns of IV. Occupational Health Data workers has convinced OSHA that V. Quantitative Risk Analysis VI. Technical and Economic Feasibility A. Technical Feasibility B. Economic Feasibility VII. Environmental Impact. Regulatoiy Flexibility, and Regulatory Analysis VIU. Summary and Explanation of the ETS continued exposure at the current PEL and under actual workplace conditions constitutes a "grave danger" to employees, and that an ETS is needed to protect them. The ETS requires that employees may IX. Public Participation. Authority and no! be exposed to concentrations of Signature, and Part 1910 (Amended) asbestos exceeding 05 f/cc on an 8-hour 1. Introduction This is an emergency temporary standard (ETS) issued pursuant to section 6(c) of the Occupational Safety and Health Act of 1970 (the Act] (84 Stat. 1596; 29 U.S.C. 655), the Construction Safety Act (Pub. L 91-54; 40 U.S.C. 333), the Longshoremen's and Harbor Workers Compensation Act (33 U.S.C. 941), the Secretary of Labor's Order No. 9-83 (48 FR 35738). and 29 CFR Part 1911. This emergency standard consists of revisions to the existing standard covering occupational exposure to asbestos, 29 CFR 1910.1001, in order to protect employees from a grave danger from continued exposure to asbestos at current exposure levels. OSHA has determined that workers exposed to asbestos under exposure conditions existing under the current standard face a grave danger of time-weighed average basis, and permits the employer to choose among engineering controls, work practices and respirators to reduce exposures to the new PEL. However, the requirement in { 1910.1001(c) to utilize feasible engineering controls and work practices to reduce exposure levels to 2 f/cc remains in effect under this temporary standard. The ETS also requires employers to institute a training program within thirty days of the effective date of this emergency standard. The training progam will include instruction on respirator fitting and use. handling of asbestos, medical information, the relationship between smoking and lung cancer, and a review of the asbestos standard. The ETS applies to all industries covered by the Act including "general industry", construction and maritime industries. developing incurable cancer and f * OSHA Is aware, at cows*. that Sactton 0(c) of asbestosis. Further, OSHA has found tha Act Ualti tba effective Una of an ETS to 6 that an immediate reduction in exposure levels to 0.5 f/cc or below is necessary to protect such employees from this months sod OSHA concludes that a gravo danger exists and an ETS is necessary avsn If OSHA focuses exclusively on this six month period However, the Agency believes it is appropriate to grave danger. Therefore OSHA is calculate benefits deriving bom an ETS using issuing this ETS to compel reduced exposures to respond to this emergency situation. OSHA's determination that a grave Ufettme risks from 20 snd SS years of sxposurs to the PEL of OS f/cc ostabUshod by tha ETS. Although tha ETS expires within S months Section 8(c) requires that rulemaking on a parmansnt standard * also be completed within months, so that than danger currently exists is predicated upon quantitative risk estimates in this record which point to a large number of excess deaths from cancer (and will be no gap in protaction for exposed employees. In 06HA s experience and Jndpnant. complying arith this statutory directive and completing rulemaking for a permanent standard within 8 months of an ETS has and can be dons. ABS-008415 LAM 025240 Federal Register / Vol. 48, No. 215 / Friday, November 4. 1983 / Rules and Regulations 51087 As provided in 29 CFR 1953.22, the 24 States with their own OSHA-approved occupational safety and health plans are expected to adopt a comparable standard within 30 days of this publication date. These States are: Alaska, Arizona.-Califomia, Connecticut (for State and local government employees only)., Hawaii, Indiana, Iowa, Kentucky, Maryland, Michigan, Minnesota, Nevada. New Mexico. North Carolina. Oregon. Puerto Rico. South Carolina. Tennessee. Utah. Vermont, Virginia, Virgin Islands. Washington, and Wyoming. Also, in response to the emergency conditions faced by exposed workers, OSHA is stepping up its enforcement activities in targeting workplaces where asbestos is handled. The temporary provisions that are being added to the existing asbestos standard prescribes the major components of an occupational safety and health standard In general, most of the current rdt)uir*shents remain unaffected by the emergency temporary standard. However, compliance with some requirements will be triggered by the new exposure limit of 0.5 f/cc instead of the former PEL of 2 f/cc. For example, requirements such as change rooms remain unaffected by the ETS and the trigger level for change rooms remains the former PEL of 2 f/cc (11910.1001(d)(4)). However, the ETS requires that where concentrations may exceed the new PEL, the employer must post signs indicating such locations (8 moioaukUQ)).. OSHA based its decisions concerning appropriate provisions for the ETS on its determinations of the kind and degree of protective action needed to protect against a grave danger and the feasibility of instituting these provisions during the period of the ETS. Under section 6(b) of the Act. OSHA will shortly schedule an informal rulemaking hearing on the proposed permanent changes to the standard. OSHA will also propose other revisions to the asbestos standard that will be explained in the separate notice of proposed rulemaking. Application of certain provisions of the asbestos standard to the construction industry will be raised at' that time. OSHA anticipates that this notice will be published shortly. OSHA Is now accepting written comments on those issues raised in the ETS which are relevant to revising the permanent rule. The Secretary most promulgate a section 6(b) standard no later than six months after publication of the emergency standard. .II. Legal Authority and Background. A. Legal Authority Section 6(c) of the Act provide;: for Immediately effective emergency temporary standards in certain circumstances. The Secretary has the authority to issue such a standard, without rulemaking, "if he determines (A) that employees are exposed to grave danger horn exposure to substances or agents determined to be toxic or physically harmful or from new hazards, and (B) that such emergency standard is necessary to protect employees from such danger." 29 U.S.C. 655(c)(1). Thus, the danger must be "from" worker exposure and it must be "grave." not merely significant and the regulation. must be "necessary" to address "such danger," not merely reasonably necessary or appropriate to provide safe employment. In addition, as in section 6(b) permanent standards, OSHA may impose requirements in the emergency standard only to the extent that they are "feasible" within the meaning of section 6(b)(5) of the Act. The Fifth Circuit has emphasized the importance of the severity of health effects in OSHA's consideration of whether an ETS is appropriate: .........the danger of incurable, permanent. or fatal consequences to workers, as opposed to easily curable and fleeting effects on their health, becomes important in the consideration of the necessity for emergency measures to meet a grave danger." Florida Peach Growers Association. Inc. v. United States Department ofLabor, 4368 F.2d. 132 (CAS). OSHA also has interpreted relevant judicial-decisions to require that its evaluation of the gravity of the danger and the necessity for emergency action must be made in the context of actual workplace conditions. B. Background 1. Events Leading to the ETS. In june, 1983, OSHA received a petition for an ETS from the International Association of Machinists and Aerospace Workers (Exs. 84-244). Subsequently letters supporting this petition were received from 16 other unions (Ex. 84-284 to 84294,84-387 to 84-391). The petition requested an ETS to reduce the PEL to 0.1 f/ce. to modify some existing work practice requirements, and to require other protective provisions regardless of exposure level. The main reasons set forth in the petition and supporting letters are that continued exposure under current workplace conditions constitutes a grave danger and that delaying remedial action until a permanent standard is promulgated would cost large numbers of currently- exposed workers their lives. OSHA shares the genuine concern expressed by the petition and letters from unions representing thousands of employees who are directly faced with the threat of asbestos-related disease from continued exposure. And OSHA concludes that failure to issue this ETS may cost a number of exposed workers their lives. In August 1963. OSHA completed a comprehensive risk assessment based on numerous human studies which estimated the number of excess deaths from the three major asbestos-related cancers, i.e.. lung cancer, mesothelioma and gastrointestinal cancer, at the current permissible exposure level of 2 f/cc and at various reduced exposure levels. OSHA placed this document, "Quantitative Risk Assessment for Asbestos-Related Cancers", in the Asbestos docket in August 1983 (Ex. 84- 349). OSHA subsequently revised this document (See Ex. 84-392). 2. History of the Asbestos Standard. OSHA has regulated asbestos since 1971. A12 f/cc limit for asbestos was included in the initial promulgation of OSHA standards pursuant to section 6(a) of the Act. on May 29,1971 (36 FR 10466). In response to a petition by the Industrial Union Department of the AFL-CIO, OSHA issued an ETS on asbestos on December 7.1971, which established a PEL of 5 f/cc on a timeweighted average basis and peak exposures of 10 f/cc. The current standard, promulgated in June 1972, first established an 8-hour time-weighted average PEL of 5 f/cc and a ceiling limit of 10 f/cc. The limits were intended primarily to protect employees against asbestosis. and it was hoped that they would provide some incidental degree of protection against cancer. Effective July 1978. the TWA limit was reduced by the standard to 2 f/cc. This limit has remained in effect since that time. The standard also includes provisions covering methods of compliance, monitoring, medical surveillance and housekeeping. Court review of this standard upheld all major provisions, but remanded two Issues for OSHA's reconsideration. IUD v. Hodgson. 499 F. 2d 467 (CADC 1974). These issues were whether the July 1976 date for the 2 f/cc standard should be accelerated for some industries and the adequacy of the 3 year retention period for exposure monitoring records. Subsequently, OSHA Increased this retention period to 20 years (41 FR 11504} and the acceleration issue became moot After reviewing the then available scientific data, in October 1975 OSHA'. published a notice of proposed " ........ ABS-008416 LAM 025241 SIMS Federal Register / Vol. 48. No. 215 / Friday. November 4. 1083 / Rules end Regulations rulemaking to revise the asbestos standard because OSHA believed that "sufficient medical and scientific evidence had been accumulated to warrant the designation of asbestos as a human carcinogen" and that advances in monitoring and protective technology made reexamination of the standard "desirable." This proposal would have reduced the time-weighted average to 0.5 f/cc and imposed a ceiling limit of 5 f/cc foe 15 minutes (40 FR 47652). The basis for the 1975 proposal's reduced permissible level of 0.5 f/cc was OSHA's then current policy for carcinogens that no safe threshold level was demonstrable and. therefore, the Act required OSHA to set the PEL as low as technologically and economically feasible. This policy was rejected by the Supreme Court in the benzene decision (IIID vs. API. 446 U.S. 601 (I960)). OSHA limited the proposed revisions in the 1975 notice to all industries except construction. No hearing was scheduled on the 1975 proposal. Also, although OSHA announced its intention to develop a separate proposed revision applicable to the construction industry, no snch proposal has been published to date. OSHA is basing its present decision to issue an ETS. and to propose revisions to the permanent standard covering all employees, on information and analyses which postdate the 1975 proposal. Therefore, this ETS marks a new regulatory initiative, related to, but not part of the 1975 proceeding. On May 24, 1983. OSHA consulted the Construction Advisory Committee for Occupational Safety and Health (CACOSH) concerning applicability of a new asbestos standard for the construction industry.-CAC'OSH endorsed OSHA's position that changes in the PEL made for general industry should also apply to the construction industry. OSHA is also inclnding the comments and data received in response to the 1975 notice in the record of this proceeding, with the understanding that some commenters may have changed their positions based on intervening scientific developments and policy shifts. ___ In the decision to issue an ETS, the Agency has reviewed this regulatory history of asbestos. However, the Agency bases the decision to issue an ETS on the actual emergency conditions which now confront exposed workers, on epidemiologic studies that provide data to make numerical estimates of risks and on interpretations of these studies using quantitative risk analysis. 3. The Rulemaking to Revise the Permanent Standard. The 6(b) rulemaking initiated by this proposal is intended to be limited to the revisions mad(| by the ETS and tbe additional proposed amendments raised in the notice of proposed rulemaking to be published soon. The major subjects of the 6(b) rulemaking will be: reducing the permissible exposure limit revising the definitions of asbestos and asbestos fibers, reassessing (he methods of compliance to achieve anch limits, revising the provisions regarding respirator selection, revising die sampling and analytical method to improve reliability, and adding a training requirement OSHA will also raise issues regarding the application of the permanent standard to the construction industry. As previously stated. OSHA will soon publish a separate notice of proposed rulemaking to further explain these issues. IIL OSHA's Rationale for the ETS OSHA bas determined that prevailing conditions involving worker exposure to airborne asbestos dust justify the promulgation of an emergency temporary standard. OSHA estimates that approximately 375.000 workers are exposed to asbestos at various levels (Table 1). ranging from a high value of 20 f/cc to below 0.5 f/cc. OSHA has estimated that under currant exposure conditions asbestos-exposed workers face an extraordinarily high risk of contracting asbestos-related cancer - whether the risk is computed over a working lifetime of exposure or for exposure periods as short as 6 months. The average excess cancer risks for all workers exposed above 0.5 f/cc vsmg available exposure data and relying on the risk assessment are estimated as approximately 196 excess cancer deaths per 1000 workers for 45 years of exposure, 139 deaths par 1000 workers for 20 years. 10 par 1000 workers for 1 year, and 6 per 1000 workers for 6 months of exposure. OSHA believes that risks of these magnitudes, taking into account all relevant considerations such as total numbers of workers at risk and quality of supporting data, constitute an emergency situation which requires immediate response by die agency. The Act states that when certain statutory criteria are met OSHA is authorized to respond to an emergency situation by issuing an ETS The two pronged statutory teat for an ETS is that (1) employees must be exposed to a grave danger from exposure to substances or agents determined to be toxic or physicially harmfuL and that (2) an emergency standard is necessary to protect employees from such dangers. After evaluating all the evidence available to the agency concerning the severity and magnitude of the risk of asbestos-related disease to the current asbestos-exposed working population, comparing these risks to other occupational rides, applying relevant policy considerations, and reviewing all relevant judicial uecisioue for guidance, the agency has determined that both prongs of the statutory teat are mat and that an ETS should be promulgated. For purposes of darity. the discussion is divided into two parts. "Gave Danger" and "The Need for an ETS." OSHA believes, however, that the factors which indicate that a substance constitutes a grave danger are related to and overlap those which determine that an ETS is necessary. A. Grave Danger OSHA has determined (hat the risk to workers from exposures to asbestos at conditions that exist in the workplace pose a grave danger of death from cancer and of severe disability from the lung disease, asbestosis. In making a "grave danger" determination, the severity of the disease produced by exposure to the reflated substance and the magnitude of foe predicted risks of disease must be considered. In addition, the Supreme Court has suggested that a determination of "grave danger" indicates a situation where the risk is more than "significant" (fUD v. API. supra n. 45J. OSHA has applied that analytic approach endorsed by the Supreme Court for "significant risk" determinations in evaluating the gravity of the danger faced by asbestos-exposed workers. The Supreme Court gave some general guidance at to file process to be followed. It recognised fitet while the Agency muet support its finding that a certain level of risk exists with substantial evidence H also recognized that its determination that a particular level of risk is "significant" will be based largely ori policy considerations (IUD v. API. 448 U-S. 663,656. a. 62). OSHA bedeves, therefore, that its determinations regarding the magnitude of the risk faced by employees should, to the extent possible, rely on quantitative expressions of that risk, utilizing the best available date. The Court stated that the significant risk determination required by the OSH Act is "not e mathematical straitjacket" and "OSHA is not required to support its finding that a significant risk exists with anything approaching scientific certainty. ***A reviewing court (is) to give OSHA some leeway where its findings must be made on the frontiers of scientific knowledge (and that) ** * the Agency ie fine to ose conservative assumptions in interpreting the data ABS-008417 LAM 025242 Federal Register / Vol. 46. No. 215 / Friday. November 4. 1983 / Rules and Regulations 51089 with respect to carcinogens, risking error on the side of overprotection rather than underprotection" [488 US at 655. 658]. In the case of asbestos, the data available are of unusual breadth and high quality. However, because risk assessment itself involves many uncertainties. OSHA made certain assumptions in its analysis and evaluation of these data. In assessing the risk for asbestos-expose4 workers, OSHA has attempted to use realistic assumptions, although the court stated that the Agency was free to use "conservative assumptions" in interpreting data. OSHA, in many cases, has indicated where different assumptions may produce different results. In addition OSHA cautions that because the risk figures finally derived are the products of a process which, as the Supreme Court acknowledged, is "on the frontiers of science," they should be viewed as approximations of the degree of risk faced by asbestos-exposed workers and not as precise fixed predictions of the number of workers who will actually develop disease. OSHA has evaluated the kinds of dangers presented by asbestos exposure, the quantification of those dangers under present asbestos exposure conditions, the quality of the data on which risk estimates are based, a comparison of asbestos risks to other occupational risks, and relevant policy and legal considerations in concluding that workers are exposed to a grave danger from asbestos. 1. Nature of the Diseases. As stated above, the nature of the disease associated With exposure to a toxic substance is one of the most important elements OSHA evaluates in determining whether a grave danger exists. This factor was discussed in Florida Peach Growers Association, Ina v United States Department of Labor, supra. The court in overruling OSHA's organophosphate pesticide ETS, observed: We refect any suggestion that deaths must occur before health and safety standards may be adopted. Nevertheless, the danger of incurable, permanent orfatal consequences to workers, as opposed to easily curable and fleeting effects on their health, becomes important in the consideration of the necessity for emergency measures to meet a grave danger. 488 F 2nd at 132 (emphasis added) OSHA Is aware of no instances in which exposure to a toxic substance has more dearly demonstrated detrimental health effects on humans than has asbestos expoem Tha diseases caused by asbestos exposures are in large part life-threatening ot disabling. Among these diseases are lung cancer, cancer of the mesothelial lining of the pleura and peritoneum, and asbestosis. In addition, workers exposed to asbestos are at increased risk of gastrointestinal cancer, as shown by epidemiologic studies. Although colo-rectal cancer may be curable if detected in an early stage, other gastrointestinal cancers are usually fatal. OSHA also believes that asbestos might Induce cancers at other sites, which are also often fatal. Of these, lung cancer constitutes the greatest health risk for American asbestos workers and has accounted for more than half of excess mortality in some occupational cohorts. About 90% of lung cancer patients die within 5 years of diagnosis. Mesothelioma is an incurable cancer which is usually fatal within a year after diagnosis. It is epidemiologically linked to asbestos exposure, and occurs very rarely, if at all, in persons never exposed to asbestos. Asbestosis, a type of pulmonary fibrosis, is usually nonreversible, its advanced stages are disabling, and can be fatal. OSHA concludes that all these diseases are very serious, and that the excess mortality from such severe diseases must be considered an important factor for making a grave danger_____ _____ _ determination. 2. Degree ofRisk ofDeveloping Dangerous Disease. OSHA based its calculations of extent of risk faced by workers under current exposure conditions primarily on the results of a quantitative analysis which derived numerical estimates of cancer risk at various cumulative exposures corresponding to levels at which workers are exposed (Ex. 84-392). Although 2 f/cc is the current PEL for asbestos exposure, actual exposure conditions vary widely, mostly by irfdustry segment As explained later in this document and as set forth in Table 1, average ambient exposure levels in various industries include high exposure levels such as 20 f/cc in drywall removal, renovation and demolition: 5 f/cc in shipbuilding and repair mid range exposure levels such as 2 f/cc in secondary fabricating of cement sheet packing and gaskets and paper products and rebuilding and refadng brakes; IS f/cc for dry processing of textiles: and lower exposure levels such as 0.5 f/cc and 0.2 f/cc in the manufacture of floor tile. Because OSHA is required to consider the sctnal danger faced by workers In assessing whether exposure to a substance presents a "grave danger". OSHA looked at the risk of developing disease not only at the 2 f/cc permissible level but at all exposure leyels which workers currently face. Most of the results of these calculations for cancer are presented in Table 11 in the risk assessment section of this document. The table sets forth predicted excess lifetime cancer risks for exposures of one year. 20 years and 45 years. Risks for exposures of 6 months are closely approximated by one-half the risks for exposures of one year. Although average exposures in demolition and renovation are estimated at 20 f/cc, the table presents risks only for selected exposure levels up to 10 f/cc. These calculations show that the risks of asbestos-related disease are alarmingly high at current occupational exposure levels. For example, an estimated total cancer risk of 265 excess deaths per 1000 workers exists for workers exposed for a 45-year lifetime at 10 f/cc, a level which currently exists on some construction sites. At 5 f/cc. the exposure levels which are considered average in shipbuilding and repair, the risk of developing asbestos-related cancer for a 45-year exposure period is 149 excess deaths per 1000 workers. At the current permissible level of 2 f/cc which also represents actual exposure levels in such industries as secondary fabricating of cement sheet, packing gaskets and paper products and rebuilding and refacing brakes, risk Is estimated as 64 excess cancer deaths per 1000 workers for a 45-year exposure period. These risks remain very high when the period of exposure for which calculations are done is shortened to 20 years, which OSHA believes is another appropriate point for examination. The period of 20 years is the approximate midpoint between 1 year and 45 years of exposure: also many workers receive 20 years of exposure. Counterpart risk calculations using a 20-year exposure period are: for workers exposed to 10 f/ cc, 140 excess cancer deaths per 1000 workers; for exposures to 5 f/cc, 105 excess cancer deaths per 1000 workers and for exposures to 2 f/cc. 44 excess cancer deaths per 1000 workers. OSHA also estimated risks of developing cancer for a one year period of exposure at various levels to which employees are exposed. The counterpart risks for exposures to 10 f/cc for one year are: 15 excess cancer deaths per 1000 workers; to 6 f/cc 7 excess cancer deaths per 1000 workers and to 2 f/cc, 3 excess cancer deaths per 1000 workers. Even at current woikplace exposure levels which are less than the current PEL extraordinarily high risks of disease exist At 0.5 f/cc. 17 excess cancer deaths per 1000 workers are ABS-008418 LAM 025243 51090 Fedacal Register / Vol. 48i No. 215 / Friday. November 4. 1863 / Rules and Ratgiilntiooa predicted for a 45-year bletirae expoaura, aad 11 excess cancer death* pec 1000 workers for a 20-year exposure period. OSHA notes that the above calculations are for cancer risk only. In addition, asbe&ioa-exposed workers face a high risk of developing asbestosis, a disabling and often fatal disease. Predictions concerning the estimated magnitude of the asbestosis risk have been performed by OSHA and are discussed in the risk assessment section of this document Accordingly. OSHA estimates that at 2 f/cc, 50 workers per 1000 exposed to asbestos for 45 years will develop disabling asbestosis. At 0.5 f/cc for 45 years, it is estimated that 12 workers per 1000 will develop disabling asbestosis. Asbestosis risks can also be estimated for exposures for duration* shorter than 45 years. For example, the risk of disabling asbestosis from exposure to 1.0 f/cc for 22-5 years is 12 cases per 1000 workers. OSHA'8 estimates of the magnitude of the asbestosis risk are based on sound data from good epidemiological studies. OSHA believes, however, that the confidence which can be placed in predictions of asbestosia risk is not as great as for the predictions of cancer risk. This is because the cancer risk estimates are based on a larger and more varied data base and are derived from dose-extrapolation models that are better established. Because OSHA has determined that the risks for cancer ulone indicate a grave danger, the additional risks of developing asbestosia are not necessary to justify this ETS. However. OSHA has considered that the additional and independent risk of developing asbestosis increases the danger faced by exposed workers and underscores the gravity of the health threats to employees posed by asbestos. 3. Quality of Data oa Which Risk Estimates are Based. The underlying data upon which the quantitative risk assessments for asbestos are based are high quality epidemiologic studies, conducted in occupational environments. OSHA emphasizes that the data bases for asbestos are of unusual quality and size. Unlike most potential occupational carcinogens, asbestos has been studied often and thoroughly.for evaluation of its effects on occupational populations. In deriving these quantitative estimates for cancer risk. OSHA utilized eleven studies for the calculation of the lung cancer risk, four of which were also used to calculate (he mesothelioma risk. Investigations involved "cohort" studies where the frequencies of various types of cancars in workers exposed to asbestos were compered to those in "control''-group* not exposed to asbestos or to those of general populations such as U-S. males. Studies of such design are able to provide direct estimates of excess risk. The studies ased bp OSHA in deriving dose-response relationship* toe its risk assessment covered s variety of work situsboas and indostrial process**. This variety improves the paddirt value of the risk assessment because it lessens or eliminates the possibility that the results were uuiqve to any one occupational situation or were in fact aberrational. The occupational settings studied were: workers exposed at a chrysolite textile plant from 1930-1975 (Demeat et oL Ex*. 84-03 and 84-03771 Canadian workers at an asbestos cement facility (Finketstein Ex. 84-240); Italian chrysotife miners and mUten who worked Ailing 1930-1985 (Rtabtno et al. Ex. 84-88* workers hi an asbestos cement pipe plant (Weill et at. Ex. 84206); workers in an asbestos production plant and asbestos cement pipe factory (Henderson and EhteriineEfc. 84-48); British workers manufacturing asbestos textile products (Mo Ex. 84-199): asbestos miners and milters in Quebec, Canada (Liddell et at. Ex. 84-69); and in the Thetford Mines, Canada (Nicholson el al. Ex. 8+^72)Land workers manufacturing asbestos friction materials (Berry and Newhouse Ex. 84- 21). "Well-conducted epidemiologic studies that show a positive association between an agent and a disease are accepted as the most convincing evidence about human risk** (Risk Assessment in the Federal Government: Managing the Process, National Research Council, 1983, p. 21. Ex. 84322). No extrapolation from animal dale to human data is necessary in order to show carcinogenicity of asbestos. For most substances. OSHA must infer human health effects, such as carcinogenicity, from animal data. The results of this risk assessment performed by OSHA agree well with other recent risk assessments performed by other governmental and outside scientists (see Acheson and Gardner) (Ex. 84-218 and 84-243); EPA (Ex. 84180); Kang and Chu (Ex. 84-001]; Slilikoff et al. (64-002); aad CHAP (Ex. 84-258). 4. Comparative Analysis. Insight into the magnitude of the risk associated with asbestos exposure can be gained by reviewing ether occupational risks. OSHA believe* it is inatrwetiv* to compare asbestos risks with other workplace hazards spread on as presenting as umsuaily high degree of hazard, where the data ate considered both avattabie and wiiebie. The risk of access moilably estimated as a result of exposure to asbestos at the condition* in the workpiece today appear* to be substantially higher than other risk* experieneed by workers from occupational injury hazards. The National Safety-Cbwiei) (NSC) has reported the annuel death rates m 1981 from work accidents in a variety of indostriee (Ex. 84-339). Using the NSC data OSHA has reviewed the annual mortality from work accidents per 1000 workers in several industries in light of the excess caneer mortality from a single year of exposure to asbestos per 1000 workers. For example, in (he high risk occupations of agriculture and mining-quarrying, the annual mortality rates from work accidents were 0.54 and 0.55 per 1080 workers respectively in 1991 (Ex. 84-339). In contrast the death . rate from work accidents for sH industries combined was 0.12 per 1000 workers in 1901. OSHA has estimated that the lifetime risk for one year of exposure to 2 f/cc of asbestos is about 3 excess cancer deaths per 1000 exposed workers during the remainder of the workers' lifetimes (Ex. 4-349). Thus, asbestos workers' risk of . excess cancer mortality from a single year of exposure to 2 f/cc is roughly 5 times higher than the risk of accidental occupational fatalities from one year of employment in agriculture and mining- quarrying. As shown in Table 1. OSHA estimates that many workers are exposed to asbestos in the vicmity of 2 f/cc. In addition, OSHA calculated the average excess cancer risk to workers exposed at conditions that exiat in the workplace today (for those above 0.5 f/cc and using the scenario described in Table 3). OSHA estimates that 10 excess cancer deaths will occur per 1000 workers for 1 year exposure; thus the average risk to workers (exposed above 0.5 f/cc) in the workplace today is approximately 20 times the annual fatal accident rates in apiculture and mining quarrying. These comparisons are striking. They show that the estimated risk of dying of cancer from asbestos exposure at levels existing at the workplace today far exceeds the accidental death rate in the rfckieri of industries. Although the estimated mortality rates for cancer due to asbestos exposure are not completely comparable to the total actual accidentalfatalifiea.- the review is clearly useful in showing that t magnitude of the asbestos risk Is pave. Oa* exaa^ke of psedfetad eances risk as a result of occupational exposure la the following cancer rfek estimated from 4BS- LAM 025244 Federal Register / VoL 48. No. 21S / Friday. November 4, 1963 / Rules and Regulations 51081 occupational exposure to ionizing radiation. The --Seriated excaaa.roncw fa tality rate frera-47 yeara of ewp--ra to the maximam permissible occupation*] exposure to iooicing radiation (5 rents) is 17 to 29 per 1600 workers (Committee on the Biological Effects of ionising Radiation (BBR) ill predictions, see 46 FR 1902). However, most radiation standards (unlike OSHA standards) require that exposure limits be reduced to the lowest level reasonably achievable below the exposure timit (the ALARA principle). Approximately 96 percent of radiation workers have exposures less then one-tenth the maximum permitted limit. The excess cancer deaths ad one-tenth the permitted level are 1-7 tsllpsr WOO workers exposed 47 years. Asbestos exposures of 45 yesra to 2i/cc are predicted by OSHA to result ha M excess cancer deaths per 1088workers beginning work at age 25 (Bx. 94-89Z). OSHA's calculation for tfie average excess cancer risk to worker exposed at conditions that exist hr the workplace today (tor those above 9.5 f/cc) for a 45year expotare, is 198 excess cancer deaths per MWworkers. This figure was calculated byTaking the number of cancer deaths estimated from exposure to existing conditions for 45 years for those workers exposed to greater then 0.5 f/cc of asbestos and dividing by the number of workers expaaed to asbestos greater than 0.5 f/cc (multiplied by 1000). Therefore, the excess cancer risk at 2 f/cc for asbestos workers is estimated as more than twice as high as the maximum permitted radiation cancer risk and about 25 times higher than the estimated cancer risk of 95 percent of the workers exposed to radiation. At existing conditions, asbestos workers' excess cancer risks are estimated to be 85 times higher than the cancer risk faced by 95 percent of the workers exposed to radiation. The risk of asbestosis further increases the significance of the risk from asbestos exposure. At 0.5 f/cc, OSHA estimates that 17 excess cancer deaths will occur in 1000 workers exposed 45 years. This risk is approximately 7 times higher than the cancer risk faced by 95 percent of the workers exposed to radiation. OSHA finds that these comparative risks strongly support OSHA's finding that workers exposed to air concentrations above 0.5 f/cc are far above the point of significant risk and are at grave danger of dying from cancer. 5. Conclusion. OSHA's finding of "grave danger" is based on evidentiary and policy considerations. OSHA's determination that the magnitude of the estimated risk to exposed workers is alarmingly high constitutes the major component of the ''grave danger" finding. The overall extraordinary degree of risk, the extent that very high risk is found in many asbestos nsing industries, and the unusually high quality of the data utilized to make -these assessments present a very strong evidentiary basis for a "grave danger" finding. ]ust as iaiportantly. the unique gravity of asbestos-cansed diseases, in x particular cancer, such as mesothelioma which is linked almost exclusively to asbestos exposure, strongly supports OSHA's finding of grave danger. Also OSHA's comparison of the risk of asbestos-related disease to other induetrial risks underscores the extraordinarily high risk estimated for asbestos exposure. OSHA has also noted the concerns of workers about current workplace conditions end the numerous petitions for an ETS from unions representing many expaaed workers. Finally OSHA has relied on its experience ia evaluating and regulating workplace hazards in recognizing the extraordinary degree of risk currently faced by asbestos workers and in determining that such risk constitutes a grave danger to those workers. B. Needfor an ETS OSHA has determined that this ETS Is necessary to protect employees from grave danger, the second prong of the Act's test of OSHA's exercise of its ETS authority (Section 8(c) of the Act). As explained in detail, the effect of this ETS is to save many lives which would otherwise be lost to asbestos-related disease if current working conditions were not changed. OSHA believes that employees can be adequately protected against this grave danger only by issuing an ETS. This is because no other Agency action and no other foreseeable event would result in sufficiently reduced asbestos exposures that would alleviate the grave danger. Further, the provisions of the ETS are tailored to effect the necessary exposure reductions expeditiously. 1. Lives Saved by Issuing an ETS. OSHA has estimated the number of deaths avoided as a result of an ETS which would reduce the PEL to 0.5 f/cc (see Tables 2 and 3). For cancer only, based on continuing exposures under currently existing conditions for 8 months, the potential number of lives saved is estimated as approximately 210. Based on continuing exposures at currently existing conditions for 1 year, the potential number of lives saved is estimated at approximately 420. Also, OSHA has estimated that the promulgation of an ETS setting u 0.5 f/cc PEL may avoid 572S cancer deaths assuming 20 yeara exposure to asbestos of the current workforce at current conditions and 7815 cancer deu'hsassuming 45 years exposure. OSHA is aware, of course, that Section 6(c) of the Act limits the effective time of an ETS to 6 months, and OSHA concludes that a grave danger exists and an ETS is necessary even if OSHA focuses exclusivity on this six month period. However, the Agency believes it is appropriate to calculate benefit* deriving from an ETS using lifetime risks from 20 and 45 years of exposure to the PEL of 0.5 f/cc established by the ETS. Although the ETS expires within 0 months, Section 6(c) requires that rulemaking on a permanent standard also be completed within 6 months, so that there will be no' gap in protection for exposed employees. In OSHA's experience and judgment complying with this statutory directive and completing rulemaking for a permanent standard within 0 months of an ETS has and can be done. OSHA also believes, based on its experience, that it is very likely that the PEL established alter 0(b) rulemaking will be no higher than 0.5 f/cc. the ETS limit. Therefore. OSHA believes that the ETS will result in a reduced lifetime worker exposures of0-5 f/cc or lower for 20 or 45 years, and that the benefits derived from these exposure reductions for these time periods are appropriately attributed to OSHA's promulgation of this emergency standard. S. Employee Exposures. To derive these estimates of numbers of lives saved. OSHA depended on its knowledge of the following factors: (1) The employee exposure levels from the ambient asbestos air concentrations in the workplace; (2) the number of workers exposed at the various asbestos levels; (3) the duration of the exposure; and (4) the probability of the disease (or the risk) associated with the cumulative exposure. Employee exposure levels are conventionally measured in terms of the number of asbestos fibers that are 5 microns or more in length in one cubic centimeter of air. f/cc. In these terms, an ambient concentration may seem to be a small amount of asbestos. However, in physical terms. 2.0 f/cc equals 2.000.000 fibers per cubic meter (f/m3). Humans inhale about one cubic meter of air per hour, depending on degree of physical activity. Thus, at this concentration, a worker would inhale roughly 10.000.000 fibers. 5 microns or more in length, over an eight hour workday. ABS-008420 LAM 025245 51092 Federal Register / Vol. 48, No. 215 / Friday. November 4, 1963 / Rules and Regulations Note.--Depending on (he industrial process, up lo 784.000.000 additional asbestos fibers less than 5 microns in length may also be inhaled, assuming that 98* of airborne asbestos fibers are less than 5 microns. OSHA continues to use the term. f/cc. for convenience but cautions that the numerical estimates of air concentrations given in these terms are only one way of viewing asbestos concentrations and should not be evaluated without interpreting the meaning of the units. OSHA used existing information to estimate worker asbestos exposure in each affected industry. Data sources included government contractor reports (F.x. 84-002 and Ex. 84-009), various studies reported in the literature. NIOSH I lealth Hazard Evaluations, and OSHA compliance data (Ex. 84-355). OSHA reviewed the information, and decided to use the 1980 Research Triangle Institute (RTI) report (Ex. 84-009) as the primary basis for exposure estimates because the RTI estimates appear to OSHA to be the most comprehensive. In addition. RTI used a large number of different data sources to make their exposure estimates, including interviews, existing documentation, and industrial hygiene surveys of worksites. Using the RTI report as the primary reference, OSHA compared RTI data with other exposure information available for each industry. For example, specific reports were found for asbestos cement manufacturing (Ex. 84248). textile manufacturing (Ex. 84-267). removal of sprayed asbestos material from buildings (Ex. 84-282). and brake repair (Exs. 84-283. 84-298). The Asbestos Information Association provided data concerning exposures during field fabrication and installation of asbestos cement pipe and sheet in controlled conditions (Ex. 84-295). The Environmental Sciences Laboratory (Ex. 84-002) also made best estimates of worker exposure in various industries for 1975. In addition. OSHA used its own field inspection experience to estimate exposures. OSHA reviewed selected case files and obtained information that listed OSHA asbestos measurements during compliance.inspections from 8/1/ 79 to 5/31/83 (Exs. 84-354, 84-355). OSHA adjusted the RTI estimate as appropriate, based upon a qualitative judgment as to which data best represent existing exposure conditions. OSHA's exposure estimates are based, therefore, upon a substantial data base and upon considerable experience in enforcing the existing asbestos standard. OSHA used the employee exposure levels in each industry to calculate the number of deaths avoided by reducing the exposure from current levels to the emergency PEL. Alternative exposure distributions used for sensitivity analysis will be provided in the Preliminary Regulatory Impact Analysis. OSHA believes that the exposure estimates are relatively good, given the state of the art of worker exposure estimation techniques and data available today. OSHA is not aware of any other available current exposure estimates. Industrial hygiene reports of operations in the same industry, but at different work sites, invariably report different asbestos exposures among the workers. OSHA reviewed these reports and made decisions regarding typical industry practices. For some operations, such as brake relining, several reports were available with complete descriptions of the working environment and OSHA was able to use these to make direct exposure estimates. Other reports, such as "Removal of Pre-formed Asbestos Insulation" (Ex. 84-296), described careful processes for asbestos handling that did not appear , representative of the methods used throughout industry, since the reported air concentrations were not consistent with other reports showing higher exposures (Exs. 84-306,84-262). Table 1 shows OSHA's estimates of employee exposure to asbestos by industry segment. aajjMa coot 4sie-4i ABS-008421 LAM 025246 Federal Register / Vol. 48. No. 215 / Friday, November 4. 1983 / Rules and Regulations 51093 r-. ui o O-- i: a- >* u a 4. vx v i-- o O r^. in -- f'O rs .s o II c +* .o* ua 3e C9 c o -3 <- -- Q. 9 C *e Og Jtdo *--<0 o u u a. Q. 9 O 9*C~ c - -e-* x X 1 a0ica1 w* >* sD O O -n 9 vO OCD 9 tamni s9--o <4cT\m> ,O'*> <CC--\DJ 01 41 X --& a01 > CL. wO 01 -E C C S- o */ * a. a-a-- 01 (| JbO*Otofl --4^--a 'Aaa3- 00&C1 _./*>iS o1 a- O O ^1 iOAj 9-e 3 -- . a10A1 0a1A qO 5 0Q>1> < c -w> cm o o> r. <sj oo OSHA E s tim a te o f Employee Exposure to Asbestos ABS 008422 lam 025247 51094 Federal Register / Vol. 48, No. 215 / Friday. November 4, 1983 / Rules and Regulations USl.A Estim ate of Employee Exposure to Asbeslus fV xaf3 <*. o Z O3 O ('J w n CoV o o OO'' O' CO 41 --(Vi CrnO < gvOsfi J> t/> vO tS <Tt CO o --(Vi G*O o 4o1 Oc ' C 41 c <V " 11 <uv <9 O' -- c t/l a - 41 */ <o v. -o -o o C j u *- 3 wOl V4>i ^41 = -a u e 41 o> 0>- --<Q wn OX 0 41 V* ,'" O C U -- U */ TCJ >*- C'O OO WU C -- C JS 4# < Q lk0/u1 UoXeJ 0x1 cOr oH<- <Ao-- 41 W V Lu 'O br-k <O*_ 4w> .X.*.><O--0-jn' jLw*ii V --- 4* O3 tn v> QO S' i- * X u i o < e w tjT 2 ^ O - ** "v> a-- (VI - S. 41 "O 41 < 3 L. 5 x- ^ c -- S- *< OJ -- ` co. W au c4 ioJ CO S* 1 --* ^- 2o 4, w u_jrOe a. o o4 a -*C- 4u *a>o s '->Va -S9* 2* 41 -- r l wt ill t/k i/I 4> 41 *C4J* 1/>1* 'O--C* II ab. tA IsOS2 t*lxC,"2GOu*^^, OCl - 40 aj^z O V* <u *; 3 *= CL i- * 4> "^5 X w--' 41 Ho Q>v^ = --2S. s*-; ^"^es41seUW;*fwOO?l a-wf4Ql1l,--tM;l ** wc> S"i >- op> 41- h O | '-C2o*-*oo -y" J9-* i^ - t! a ai^ *s I is* ai s =<( c Sw 5-- 41 >*- u. ou ^ lit ' f" w w .u"^"S3S t/l w O 2->* f IV G O \r 41 = vi a *v 11 cW O*--' JoW )o -* CU"vJi +eJ l'' a 22 3i-g *i c *a. eo u e mx 25 "C5 *J -1* -Q ** e <o *a x 4 5 a 4> O ^ S Ol - ** 1 CO i* .x- 4i O'O e > A ** *0. m ^ ^O S 4ki1 w ao - ji ^^ ". *,- U ^o n 6 41 0B " L w_ x- O v5f4l (VI 4J*1- C3 2i * 4* - - v* L 41 4U-1O1^-3O --V- **zl$ c x 41 * 2 * O' J2 8 oal -C vv ' V41 Ca w 2 9 ABS-008423 LAM 025248 Federal Register / Vol. 48, No. 215 / Friday, November 4, 1983 / Rules and Regulations 51095 Data from OSHA's Management Information System (MIS) includes a list of all asbestos samples reported by OSHA field staff during |une 1,1979 to May 31.1983 (Ex. 84-355). During this period. 949 eight-hour lime-weighted average (TWA) samples were collected in many different industries. Of the 949 samples. 731(77%) were below 0.5 f/cc 156 (16.4%) fell between 0.5 f/cc and 2.0 f/cc and 62 (6.5%) were above 2 f/cc. OSHA also reviewed the MIS summary report where the measurement data were grouped within Standard Industrial Code (SIC) classifications (Ex. 84-354). All these data contributed to the development of OSHA's best exposure estimates. In conclusion, OSHA has used its experience in enforcing the existing asbestos standard to determine the best estimates of worker exposure in the industries. On the whole, in comparison to other toxic substances present in the workplace, asbestos exposure information is well reported. Given that OSHA experience is considerable, OSHA believes its estimates are reasonable and appropriate and provide a satisfactory basis for Judgments regarding the extent of risk existing in workplaces. OSHA acknowledges, however, that the exposure information concerning asbestos removal and renovation in the construction industry is less certain than the other estimates. Despite uncertainties in qualifying exposure in the demolition/removal category, OSHA feels that this best estimate is based upon evidence that is both reasonable and the best available. Furthermore. OSHA believes that any changes made to the estimate as a result of the uncertainities would not be of a sufficient magnitude to warrant a change in its basic findings. 2. Calculation ofLives Saved and Disease Avoided. The benefits provided by this ETS consist of cancer deaths and disabling asbestosis avoided. The estimates of deaths avoided by lowering of the exposure limit by the emergency temporary standard are based on the mortality rates developed in the risk assessment discussion. To generate the expected number of excess deaths attributable to asbestos in the U.S. workforce, the expected rate of mortality at each exposure level was multiplied by the population exposed at that level. The expected mortality rate from each exposure level was derived from OSHA's quantitative risk analysis (Ex. 84-392). The expected number of deaths at the reduced PEL was then subtracted from the expected number of deaths at the current worker exposure levels to determine the estimated number of deaths avoided by reducing the permissible exposure limit. The benefits are expressed in terms of estimated deaths rather than disability because the types of cancer associated with asbestos exposure have a very poor survival rate. An example of such a calculation follows. To calculate the lung cancer deaths avoided during one year of exposure to a worker population in the construction industry who are engaged in the installation of asbestos cement sheet: Population=1765 (Table 1) Estimated Current Exposure Level=2.0 f/cc (Table 1) Estimated lung cancer risk for one year at 2 f/cc = 144/100,000 (Table 11) 1. Calculate expected deaths for installers of A/C sheet having one year of exposure at 2 f/cc exposure level: Population Xrisk~=expecteddeaths (1765) (144/100,000) = 2.54 or approximately three deaths among installers. 2. Calculate expected deaths for installers of A/C sheet at PEL of 0.5/cc for one year of exposure: Estimated risk at 0.5 f/cc=36.1 per 100,000 (Table 17) Population x risk=expected deaths (1765) (36.1/100.000) =0.84 or approximately one death among installers 3. Calculate lung cancer deaths averted or lives saved for A/C sheet installers by reducing exposure to 0.5: subtract expected deaths at 0.5 f/cc from those expected at 2/f/cc (2.54) -(0.64) = 1.9 or approximately two lives saved. % The estimated benefits derived from exposure reductions to several PEL's which OSHA is considering for a final standard are presented in Table 2. These benefits represent the estimated number of cancer deaths avoided from those expected due to 20 years exposure at the estimated current exposure levels assuming worker exposure begun at age 25. The benefits were calculated using a base of 20 years exposure because, as stated earlier, 20 years is an estimate uf typical lifetime exposures for some workers. Exposures of 6 months, one year and 45 years duration are also used to show the grave danger and the need for the standard, besed upon a 0.5 f/cc PEL (see Table 3). ' For Table 2. OSHA bases its respirator assumptions in the construction industry and shipbuilding industry on its enforcement experience, on its familiarity with the industries, trade unions, and industrial hygienists, and on available studies on this subject. OSHA believes that these estimates may overstate .the degree of respirator use. This would lead to underestimation of benefits of the ETS for the following reason: The fewer workers in compliance with the current requirements for respirator use for ambient exposures exceeding 2 f/cc, the greater the potential benefits from an ETS mandating that workers be trained regarding the significant risks from asbestos exposure and the importance of using respirators. Also, workers currently using respirators will benefit by further reductions from 2 f/cc to 0.5 f/cc. However, if respirators mandated by the ETS are not used properly or consistently, then the number of cancer deaths avoided is an overestimate. To test the results OSHA also calculated the benefits assuming that all shipbuilding/repair and construction/ renovation/demolition operations comply with the current standard. All of the shipbuilding/repair workers would thus be protected by respirators which would reduce exposures to 0.5 f/cc. The construction workers would have an exposure of 2 f/cc. The benefits of reducing the PEL to 0.5 f/cc would, therefore be calculated as the difference between expected deaths at 2 f/cc and the expected deaths at 0.5 f/cc for all workers in these segments. Table 4 shows these benefit estimates. aiLUMO COOC 4S10-M-W ABS-008424 lam 025249 51096 Federal Register / VoL 4*. No. 215 / Friday, November 4, 1983 / Rules and RegnlaWowe ESTIMATES uF CANCER DEATHS AVOIDED ;? c: - * o i/t lO uo au> o n CM o ao M tn CM <o 00 <T> cn CM <M m i o fc c S "i -I o i; - >*n 3Al e<3 ti/ tM O Qi > J= * U4/ w>* 41 o <Vo CO i/ O O Ko A CM r-* O N o * S2 N *3.3 u si? 2 S * V. * 4>1 v* loife* CL 2 c o&3 g e S?c O |S2 O oa S"2 SSS 5<- -S I**! |V - ** O 4**t-- *4* ^41 4 ^ C */ M ^ OV ^3v* wLi <4** h a o c 9cc>o o * o a41 aX, s'* f <?st i b* a<>4s o. 5 823 cks . 3* D* *u-l! ss "I t* 5f*me+ -- Sij: is* S?" T eoS# 44>f *3 or <4 4 ** | * t o>t iwe Ci PW4 4u 4 4( S 83 SW So 5 wS *0ti 4Oa e8 s II: *l/* J** 4 |l'4 lift s:s**- I?? S3 A 41 u ABS-008425 lam 025250 Federal Register / VoL 48, No. 215 / Friday, November 4, 1983 / Rules and Regulations 51097 Although the reaulte in Table 4 show fewer benefits OSHA believes the number of deaths avoided under this scenario. 80 for six months. 182 for 1 year, 2416 for 20 years, and 3513 for 45 years indicate a grave danger and necessitate this ETS action. Table 4 Cancer Oeaths Avoided with a 0.5 f/cc ETS Assuming Compliance with the 2 f/cc Standard Industry Sector 6 months 1 year Primary manufacturing 2 8 117 Secondary manufacturing 17 31 471 Automotive Aftermarket 4 8 134 Shipbuilding/Repair* 00 0 Construction** 57 115 1694 TOTAL 80 162 2416 20 years 45 years 174 684 194 0 2461 3513 *0SHA assumes that the entire industry sector coagjlies with f/oc standard through use Of controls or respirators capab reducing exposure to 0.5 f/cc. **0SHA assumes- that the renovation and demolition Industry sectors are protected to 2 f/cc by use of half masic air purifying respirators. 3. No Other Agency Action Is Adequate To Protect Employees Against This Grave Danger. OSHA believes that Congress intended that the ETS authority may be exercised when OSHA determines that a grave danger is presented and when the provisions of the ETS are crafted to be an immediately effective means of protecting employees against such danger. OSHA. however, has evalauted the potential capability of other possible actions to reduce the risk from asbestos exposure. OSHA has determined that no other agency action short of issuing this ETS which compels an immediate four fold exposure reduction will sufficiently protect employees against the grave danger of developing asbestos related disease. In making this determination OSHA considered the usefulness of stepping up enforcement of the current asbestos standard, and initiating a 8(b) rulemaking proceeding to permanently reduce exposure levels. These actions will be discussed in turn. OSHA notes that these actions are not mutually exclusive, and in fact, OSHA is both stepping up its enforcement activities and embarking on a required 6(b) rulemaking as coordinated activities in addition to issuing an ETS. (a) OSHA first concludes that merely increasing its enforcement of the current . asbestos standard would not sufficiently reduce the grave risk of asbestos-related disease to exposed workers. Even if such increased enforcement resulted in immediately uniform compliance with the currant 2 f/cc standard In aO industries, risks to asbestos-exposed - workers would remain unacceptably high. As stated above. OSHA has estimated that at a 2 f/cc level employees exposed over a working lifetime of 45 years are predicted to have an excess risk of dying from cancer of 64 in 1000 and of contracting disabling asbestosis of 50 in 1000; employees exposed over 20 years are predicted to have excess cancer risk of 44 in 1000 and a disabling asbestosis risk of approximately 22 in 1000. Even at exposures lasting one year, an estimated 3 employees out of 1000 are predicted to die of asbestos related cancer, and there is also additional risk of developing asbestosis. Risks at the 2 f/cc level have been acknowledged as unacceptable by other governments which have reduced their permissible levels below 2 f/cc. To estimate the number of lives at stake if only the current standard were immediately enforced instead of compelling a 0.5 f/cc level through this ETS, OSHA made other calculations using certain assumptions about actual compliance levels. For this analysis, OSHA assumed that all construction, renovation and demolition operations comply with the ' current standard so that workers in this industry would have exposures of 2 f/cc. The difference in the number of deaths avoided by compliance with the 2 f/cc standard compared with those avoided from compliance with the 0.5 f/cc ETS is still very high and shows that asbestos exposure even at 2 f/cc represents a grave danger to such exposed employees. Table 4 sets forth the results of these calculations. Thus, for a 20 year period of exposure, OSHA estimated that merely ensuring compliance with the current 2 f/cc limit may cost 2418 employees their lives. For one year of exposure, OSHA estimated that 182 employees may die if exposures are not reduced to 0.5 f/cc, assuming full compliance with the current 2 f/cc limit. OSHA believes that these benefits (l.e. lives saved] represent the lower bound of those that would result from the ETS for two reasons. Half mask respirators which may be used at air concentrations of 20 f/cc to comply with the current standard may not effectively reduce employee exposure below 2 f/cc Further under the ETS, half mask respirators may only be used in workplaces where concentrations do not exceed 5 f/cc. Employees in workplaces between 5 and 20 f/cc must use more protective respirators under the ETS; hence, it is likely that their actual ABS-008426 LAW 025251 51096 Federal Rogiste* / Vot 48, No. 215 / Friday. November 4, 1983 / Rates and Regulations exposure* may be reduced to below 0-5 f/cc and the benefits of the ETS wiH be a gaf in protection between the parts null be needed for many yearn. For expiration of the ETS and the imposition example, automobile manufacturers may correspondingly increased. Further, an of the permanent standard for a switch to nen-asbeetoe brakes and KI S by nature of the action itself, and substance already determined to present clutch facing in new retdetee in the next the accompanying enforcement program, a grave danger. will undoubtedly boost the incentive* to OSHA also believes that it is helpful comply with all protective provisions of to evaulate the extent of risk resulting the asbestos standard. from lifetime exposure periods of 20 and (b) OSHA rejected relying on merely 45 years in the absence of this ETS. As beginning Section 6(b) rulemaking shown in Tables 10 and 11. these risks proceedings to revise the standard to are extraordinarily high. reduce the PEL as an inadequate 4. Other Factors Indicating a Need for response to the grave danger faced by an ETS. Although worker exposure to asbestos-exposed workers. Beginning asbestos has been declining over the rulemaking proceedings results in no years. OSHA believes that exposure immediate workplace changes. conditions will continue to present grave Employees would still continue to be danger in the near future unless an exposed to those conditions which define a grave danger for at least the pendency of the rulemaking. In OSHA's experience, completing 6(b) rulemakings emergency standard is promulgated. OSHA bases this finding on its evaluation of exposure information, asbestos use statistics, consideration of not initated by an ETS concerning the nature of industrial exposure today, hazardous substances can take many and the degree of compliance with the years. For example, the coke even current standard. emission standard took approximately Although OSHA anticipates a decline Vh years, the lead standard, more than H years and the cotton dust standard, more than 4 years. These periods do not in use of asbestos in products, this decline will not materially affect asbestos use in the near future. The include any of the additional delays in extensive tort litigation regarding the effective dates of OSHA standards that were due to judicially imposed slays, which have resulted in delays asbestos and the ewetenees of health S'effects associated with asbestos exposure provides strong inducements lasting several years. Under the most for producers and users of products to .favorable circumstances, however. switch to substitutes for asbestos. In OSHA believes that it is possible that a fact, asbestos consumption has declined section 6(b) rulemaking limited to the \ over the years as shown; issues raised herein might be completed several years. Old vehicles, containing asbestos parts, will remain in use for many years thereafter. Paik et al. found over half of all multistory buildings surveyed (60 oat of 427} contained sprayed on asbestos material (Ex. 84262). The potential for asbestos exposure exists as long as asbestos remains in these buildings. Renovation work in the building such as relocating walls and electrical wiring^ will result in continued exposure for carpenters, painters, electricians, laborers, sheet metal workers and others. Therefore, OSHA dpes not believe that major exposure leveY reductions are about to occur in the near future. The trend towards reduction in asbestos use and asbestos exposure has bean a slow gradual process with the exception of instances where the government intervened. For example in 1973 the EPA banned use of sprayed on insulation containing more than 1 percent asbestos. Without OSHA acting at this time. OSI LA predicts that the grave danger Conditions witt persist, certainly during the period required for OSHA's issuance of a 6(b) standard, and probably for many yean thereafter. 5. The Need for the Specific Provisions of the ETS. The provisions of the ETS require the employer to use any in approximately one year, absent an U5 Consumption method of control to reduce employee ETS. exposure to 0.5 f/cc and'to train workers As shown above, the estimated risks Year Pattern (1000 tons) concerning the hazards associated with of developing asbestos-related cancer asbestos and procedures for reducing due to exposure for one year under the risk. These provisions must be current conditions, are still- implemented immediately. OSHA ' extraordinarily high. The additional 1975 552 believes they are greatly needed to risks of developing asbestosis due to one year's exposure under current 1976 afford employees meaningful protection. 659 As the benefit section explained in conditions, although quantified with less certainty, are also more than significant. 1977 detail, compelling exposure reductions 672 . through an ETS win resuR hi a great OSHA also believes that the risks of six months exposure, approximated by taking over half of the one year risks under current conditions, also are unacceptably high. OSHA emphasizes as stated above, that OSHA's experience shows that without an ETS. proceedings leading to a permanent health standard are unlikely to be completed within a six month period. The explanation of OSHA's capability to produce a standard within 6 months of an ETS lies in the urgency generated by OSi (A's finding of a grave danger, the existence of a specific statutory deadline to complete a rulemaking within 6 months and the need to prevent 1978 - 1979 1980. 1981 1982 619 561 360 349 248 Worker asbestos exposures wiM continue, however, for many years in the future because substitutes for some products do not appear immediately available. For other products, old replacement number of lives saved. In addition, as stated earlier OSHA believes that its estimates at the degree of compliance with respirator . requirements in the construction and shipbuilding industry may overstate the degree of present respirator use. Thus. OSHA may be understating the benefits to be derived from an ETS. and hence the need for such an ETS to be promulgated. In addition. OSHA's estimates of exposures in demolition and renovation segments of the construction industry are approximately 20 f/cc. Because under the ETS the PEL is lowered to 0.5 f/cc, the airborne concentration at which more protective ABS-008427 lam 025252 Fi^flealReffateW^olj^aj^te^lS^Friday^tovMribM^iaBa/RdMayBdR^uialiOTM^^SM# respirators (with a protaction factor exceeding 10) must'be worn drops from 20 f/cc (10 times the former PEL of 2 if cc) to & f/cc (10 limes the new PEL of 0.5 f/cc). Therefore OSHA believes that more workers will wear more protective respirators under the ETS than under the current permanent standard, that their actual exposure levels will be reduced to below 0-5 f/cc and consequently that their risks will be reduced even more than the calculations indicate^ OSHA Ends also that requiring a training program to be instituted 83 quickly as possible is one of the most effective methods of maximizing the beneficial impact of the exposure reduction and of all the protective provisions of the current standard. Thus, the content of this ETS has been chosen as the most effective short-term strategy to reduce asbestos risks which will be accepted and implemented. 6. Conclusion. OSHA finds that workers exposed to asbestos in the workplace at existing exposure conditions need this emergency temporary standard tuprotest them from the gpave danger presented by these conditions. OSHA finds that by compelling a reduction in exposure to 0.5 f/cc for those employees presently exposed aver that leveL many lives will be saved. Training hnposed by the ETS will enhance the risk reductions, although qeontificaShm of that additional reduction cannot be calculated. 0.3 f/cc is the lowest feasible level achievable thfpgigh this emergency action, where sfiori-term implementation of the controls is required. Only by issuing an ETS compelling reductions tn exposure levels below the current permissible limit of 2 f/cc can OSHA bring about adequate reductions in risks. The administrative action of stepping up enforcement of the current standard is an. ihadeqnate response to OSHA's fiadingihatcarrent condition* present a grave danger. OSHA does net believe that any sigpificant redaction will occur within an acceptable time period without this emergency standard action. OSHA has observed a gradual reduction in asbestos ana. buh tithe Agency's experience, significant and rapid exposure reductions usually occur in response to standard*. The provisions of the ETS wiH significantly reduce the risk and reduce it quickly. Therefore OSKA han detanakmd fast tike ETS is necessary fo redicethe gave danger. IV. Occupational Health Data A. Introduction 1. Asbestos-related Diseases. Asbestos exposure can cause a number of disabling and fatal diseases. Among these diseases are lung cancer, cancer of the mesothefial lining of the pleura and peritoneum, and asbestosis. it is also likely that asbestos increases the risk of gastrointestinal cancers. Of all the diseases caused by ashgstos. death from lung cancer constitutes the greatest health risk for American asbestos workers. Lung cancer has been responsible for over half of the excess mortality from asbestos exposure in some occupational cohorts. The relationship between lung cancer and asbestos exposure has been established by numerous epidemiologic, studies of diverse groups. Asbestosinduced lung cancer usually has a latency period in excess of 20 years and may ba diagnosed at an earlier age than for non-exposed persons (Craighead et al., 1982S Ex. 84-0331 Few cases of lung cancer are curable despite advances in medical and surgical oncology. Only 9% of hmg cancer patients survive five or more years after diagnosis (American Cancer Society, 1983; Ex. 84-160). Asbestos exposure acts synergistically with cigarette smoke to multiply the risk of developing lung cancer. Mesothelioma also has been' conclusively shown to be associated with asbestos by many studies. In some asbestos-exposed occupational groups, 10%-18% of deaths have been attributable to mafignant mesotheliomas. Malignant mesotheliomas of the pleura and peritoneum are extremely rare in . persons not exposed to asbestos.' Generally, a latency period of at least 25 to 30 years is required in order to observe mesotheliomas in an occupational cohort. Some victims of mesothelioma have had a latency period exceeding 40 years since their initial exposure to asbestos (Craighead et al., 1982; Ex. 84-033).This form of cancer is rarely curable and is usually fatal within a year after diagnosis. There is no evidence for a relationship between cigarette smoking and mesothelioma risk. Asbestos exposure can cause pleural and/or other pulmonary disease. Pleural plaques are one of the markers of exposure and may develop within 10 to 20 years after the initial exposure. Plaques are opeqne patches visible on chest X-rays that consist of dense strands of collagen (connective tissue protein) Hned by mesothefial cells. All commercial types of asbestos induce plaques. Plaques can occur even when fibrosis is absent and do not seem tu reflect the severity of pulmonary parenchymal disease. Pleural calcification is els* commonly found in persons who ham been exposed to asbestos (Craighead et al., 1982: Ex. 84033). Asbestosis is pulmonary fibrosis caused by the accumulation of asbestos fibers in the lungs. Adverse effects of asbestosis range from shortness of breath upon exertion to cyanosis, effusions of serous fluid, respiratory failure, cardiac decompensation, and death Often.asbestosis is a progressive disease, even in the absence of continued exposure. Symptoms of disease are shortness of breath, cough, fatigue, and vague feelings of sickness. When the fibrosis worsens, shortness of breath occur*even at rest. One clinical feature of early asbestosis as well as other lung diseases is end-inspiratory crackles (rale*). Diagnosis of asbestosis it based spots the presence oi characteristic radiologic changes, symptom*, rales, other clinical features of fibrosing lung disease and a history of exposure to asbestos. Cigarette-smoking asbestos workers may have an increased risk of asbestosis relative to non-smoking asbestos workers. (Craighead et al., 1982: Ex. 84-033). Some epidemiologic studies have observed increases in esophageal, stomach, colo-rectal, kidney, laryngeal, pharyngeal, and buccal cavity cancers. While the magnitude of increased cancer risk for these sites is not as great as for lung cancer and mesothelioma, the increased risk is nevertheless of considerable importance because of the high background rates of some of these tumors in (he general population. A 50% increase irr a common cancer such as colo-rectaf cancer results in many more deaths than a 50% increase in a rare cancer. Colo-rectal1 cancer, if detected and treated In an early localized stage, has a five year survival rate of about 70% (American Cancer Society 1983; Ex. 84-180). Surgical and medical treatment is less successful for the other sites listed above. Adverse effects frem exposure to asbestos have been observed in workers involved in asbestos cement pipes and shingles manufacturing (Enterline et al.. 1973a. 1973b: Weill at aL. 1979: Finkelstein. 1982.1983) (Exhibits 84-122. 84-123. 84-20S. 84-044. 84-240), asbestos mining and milling (Wagner at k I960; Liddell et aI-,1977; McDonald et al.. 1980; Hobbs et aL, I960; Nicholson at at.. 1979; Ruftino et aL 1979] (Exhibit* 2-21. 84-059; S4-06A Mr-Wt 84-072.84-088). asbestos textile.manufacturing (DoU. 1955: Peto et al. 18801 Berry et aL 1979; ABS-008428 lam 025253 51100 Federal Register / Vol. 48, No. 215 / Friday, November 4, 1983 / Jl.ules and RegulflliWM Dement et al., 1983) (Exhibits 84-040, 84189. 84-020,84-037). insulation work (Selikoff et al.. 1979) (Exhibit 84-090), shipbuilding (Selikoff et al.. 1979; Blot et al.. 1980; Tagnon et al.. 1980) (Exhibits 84-091. 84-109. 84-182), and in a variety of asbestos products manufacturing ' industries (JoneS et al.. 1980; Henderson and Enterline. 1979; McDonald and McDonald. 1978; Scidman et ai., 1979; Robinson et al.. 1979; Acheson et al-, 1981) (Exhibits 84-138, 84-048. 84-154, 84-087, 84-082. 84-103). It should be noted that 2 fibers per cubic centimeter, which is the current OSHA standard, is equivalent to 2 million fibers per cubic meter of air. Because humans breathe in about a cubic meter of air every hour, depending on physical exertion, the current OSHA PEL for asbestos allows workers to inhale 2 million asbestos fibers per hour during an 8-hour work day. For the sake of brevity, subsequent discussion in this preamble will express exposure in terms of fibers per cubic centimeter (f/cc) rather than fibers per cubic meter. Note.--The current OSHA standard includes asbestos fibers 5 micrometers or more in length, thereby excluding shorter fibers. Since up to 98% of airborne asbestos fibers are shorter than 5 micrometers,' workers may inhale up to 100 million asbestos fibers per hour during an 8-hour work day". Since OSHA's publication of a notice of proposed rulemaking in 1975, additional studies have confirmed that asbestos exposure causes a high risk of cancer. In addition, much more complete data on the nature of dose-response relationships for asbestos-induced diseases are now available. These studies generally indicate that the PEL set by OSHA in 1972 is inadequate to protect asbestos workers from either lung disease or cancer. The following agencies and organizations have reviewed the health data for asbestos; International Agency for Research on Cancer (LARC) (1977, Ex. 84-321), Organization for Economic Cooperation and Development (OECD) (1979. Ex. 84-337). NIOSH (1978.198a Exhibits 84-338 and 84-320). Advisory Committee of the Health and Safety Commission of the United Kingdom (1979. Ex. 84-218), the Chronic Hazard Advisory Panel on Asbestos (CHAP) (1983. Ex. 84-258). and the U.S. Environmental Protection Agency (1982, Ex. 84-180). All of these groups have concluded that there is a causal relationship between asbestos exposure and the development of cancer and non- mallgnant respiratory disease. NIOSH recommended reduction of the PEL for asbestos to 0.1 libers per cubic centimeter (0.1 f/cc) in 1978. In 198a e joint NIOSH/OSHA Asbestos Work Group stated that there was no level of exposure to asbestos below which clinical effects did not occur and recommended a PEL of 0.1 fibers per cubic centimeter (0.1 f/cc], based on the limitations of current technologies of measuring air concentrations of asbestos. The 1979 report of the Advisory Committee of the Health and Safety Commission of the United Kingdom, hereafter referred to in this section as the U.K. Committee, led to the reduction of the British standard for asbestos to 1 f/cc for chrysotile. 0.5 f/cc for amosite. and 0.2 f/cc for crocidolite. Currently, it appears that the United Kingdom may lower the PEL for chrysotile to 0.5. 2. Evaluation ofRisk. OSHA's first step in analysis of risk of disease from exposure to a potentially hazardous agent is a qualitative evaluation of scientific data. This evaluation involves reviewing human and experimental studies to consider such factors as overall study design, methods of data collection, biologic plausibility of findings, consistency of findings from different studies, temporal correctness of the association, and other factors as well as general scientific judgment Subsequently, after a specific agent has been judged to be hazardous, the quantitative exposure-response relationships between the agent and disease can be investigated. The available data on air concentrations of the substance or biological indices of exposure, such as fiber contents within lungs, can be reviewed for cohorts of workers demonstrated to have an increased risk of disease. If workers with an observed excess risk of disease have received cumulative exposures permitted by the current OSHA permissible exposure limit then a potential significant health risk from exposure to the PEL has been established. If the workers with observed excess risk received cumulative exposures above those p emitted by the current PEL. then risk from the current PEL may be estimated from risk observed at higher levels by using dose-response extrapolation models. The section. Epidemiologic Evidence on Risk from Exposure at die Current PEL. will discuss the extent to which excess risk has been observed from low exposures to asbestos. Section V,, Quantitative Risk Analysis, will discuss the prediction of excess risk from low asbestos exposures using dose- extrapolation models based on studies observing excess risk in humans. OSHA considers that both risks observed by studies and risks predicted by dose- extrapolation models are valid indicators of the existence ofsignificant health risks. Exposure data frequently are not available for workers exposed before 1970. Where historical exposure data are available, the data often have such limitations as having been collected and analyzed using industrial hygiene techniques no longer in use or having been collected in only some areas of the worksite or having been collected oh only a few occasions. Therefore, of necessity, estimates of dose-response based on epidemiologic studies will have a fairly broad range of uncertainty. OSHA must examine the best available data on exposure-response to arrive at a determination of significance of risk, despite inherent and inevitable uncertainties in the data. The current 8-hour time-weighted average PEL for asbestos of 2'fibers per cubic centimeter (2 f/cc) envisages that workers will not receive a cumulative exposure exceeding 100 f/cc-years (=2 f/cc X 50 years of occupational exposure). For asbestos, OSHA believes that a number of studies suggest that increased risk of lung cancer, asbestosis. and mesothelioma have occurred from cumulative exposures estimated as close to or below 100 f/cc-years. Note.--OSHA typically uses 45 years as the period of a full working lifetime for purposes of quantifying risk from exposure to toxic agents. For asbestos, meny scientists have used 50 years to represent e full working lifetime. Thus, both 45 yesrs of exposure and 50 years of exposure are used in this document for the puipoee of analyzing dose-response relationships for asbestos. In these studies, the cumulative exposures resulted from exposure levels greater than the current OSHA PEL of 2 f/cc for an 8-hour day. For example, workers who accumulated 100 f/ccyears could have been exposed to an average level of 5 f/cc for a period of 20 years. Section B.(3), below, discusses the epidemiologic evidence for risk from low exposures in more detaiL OSHA believes that a significant health risk has been observed for cumulative exposures that could be accumulated by workers exposed to no more than the present PEL of 2 fibers per cubic centimeter (2 f/cc). Estimates of cumulative exposure are approximations of total dose received by a worker during the period of employment involving exposure to asbestos. Cumulative exposures generally are estimated by multiplying the varying intensities of exposure, s\u& as the 8-hour time-weighted averages, by the number of year's exposed. Most theories of the mechanism of ABS-008429 025254 Federal Resistor / Vat. 48; Nd. 215 / Friday. November 4. !Bb3 / Rules and Regulations 51101 carcinogenesis assume that the risk of transformation of target cell* to cancer cells increases with increases ia the total dose. Exposures received after such a transformation has taken place may aid the development of cancer, but have not caused the cancer. Thus, in some instances, using total cumulative exposure may overstate the exposures sufficient to produce increased cancer risk. For this reason, some epidemiologists omit the later years of exposure to a carcinogen in analyses of mortality in relation to cumulative exposure. In addition to cumulative exposure. intensity of exposure can be examined in relation to disease.Intensity of exposure is often approximated by 8hour time-weighted averages (TWA's). The OSHA. PEL of 2 fibers per cubic centimeter (2 f/cc) as an 8-houf TWA has only been m effect since 1978. At this time, it does not appear possible to determine whether intensity of exposure has an effect on disease risk separate from that of cumulative exposure. This is because the current OSHA PEL has been in effect only since 1978. which is an insufficient period to observe asbestos-related diseases, which characteristically have long latency periods (in excess of 25 years). As discussed in the section. Quantitative Risk Analysis. it appears that duration of exposure may hove an independent effect on mesotbefioma risk. The following sections analyze recent epidemiologic and experimental studies and discuss important aspects of the occupational health data concerning asbestos. B. Epidemiologic Evidence on Riskfrom Exposure at the Current PEL 1. Comrereren ofParticle Comets to Fiber Caaatr Cmranfty. personal asbestos samples of exposed workers are coHeeted with a membranefilter, and fibers am coasted Bring epkaaa eyepiece grattods. is die pact, particles were counted ratter than fibers, area samples were taken lather than personal samples, and samples wen collected using thermal precipitators or midget impfisprfc Thaae past hslestrlal in miUkmj of particles per cubic foot, (abbreviated as mppef or mpcf). whereas current nsmmtnti are expressed m fibers per cubic centimeter or per milliliter (f/ce or E/ml). to i/cc ieaot retapfe metierof applying a siatgje mrilipiicativ* factor dam to the differing wodh eevwoament* in which samples were token end differing sampliag methodologies. The factee fee converting mpcf to f/cc has been suggested as ranging from 1:1 to 1:5 depending on the industry studied by the scientists (Kang and Chn. Ex. 84-1: Hammad et alL Ex. 84-277*, McDonald. Liddell, Gibbs. Eyseen. and McDonald. Ex. 84-085} Far example, using conversion factors of in to 1:5 a cumulative exposure of 20 mpcf-years could range from 20 f/cc-years to 100 f/ The British Occupational Hygiene Society Committee an Asbestos conducted a study of deaf concentrations measured by current methods and historic*! methods (Ex. 84024} Conversion factors for historical methods to current methods were 1: 0.07, 1: 2.2, and 2, depending on the historical method. For asbestos mining and stilting, this section wiif use the midpoint ef the range of conversion factors: 1 mpcf=3 f/ cc. For asbestos production and asbestos cement production this section will use Hammad et al.'s suggested conversion factor of 1:1.4 for mpcf to f/ cc. 2. Epidemiologic Studies. The epidemiologic studies that OSHA interprets a* suggesting that significant health risk ha* baas observed at low asbestos exposures toll tote 3 categories: (i) Studies of hoasehold contacts of asbestos worker* (2) studies of workers whose exposure* were short term. and (3} studies ei workers with cumulative exposures estimated to be close to or betow ISO f/cc-ye&rs. This section summarises these studies and presents OSHA * analysis of these studies. A auiuher of studies have observed meaodwkomas and aa increased prevalence of cheat X-ray abnormalities among families of asbestos workers (Aatmoe et eL. Exhibits tit-808 and 84-017; Vienna and Man. Ex. tit-188. Li et al.. Ex. 84-149). In addition, mesotheliomas have been observed in community member* livmg near asbestos mtoee and factories (Wagner et at.. Ex. 2-21; Newhouse and Thompson. Ex. 84-070} Anderson et aL (Ex. 84-018) observed a 35.9 percent prevalence of chest X-ray abnormalities. including pleural thickening: plaques, pleural calcification, and irregular opacities, among 8ti& hoaseholdcontact* of amosito factory workers compared to a 4.8 percent prevalence ei cheat X-ray abnormaUtiee mnaag 328 catrfrols drawn from to same raawinity as the amosite workers. Controls were matched to the case* by age and sesc As of lata, 4 cases of mesothelioma bed been diagnosed among to 828 family contacts of the amosite workers. Presumably, family contacts received their exposure to asbestos from dust carried home on the worker's clothing, especially during the laundering of duslv clothes. About a 10-fold increase In prevalence of pleural thickening compared to controls was observed in family contacts of workers with only one year of exposure within the amosite factory. Estimated asbestos exposure levels of family contacts and community members observed to be at risk of asbestos related disease have not been reported. OSHA considers it very likely that the curaafative exposures of the family contacts and community members were teas than 100 f/cc-years. Seidman. Setikoff. and Hammond (Ex. 84-067) studied the mortality of 820 amosite production workers employed sometime during 1941-45. Seidman et al. reported that dust concentrations had never been measured in this plant and that the plant was known to have deficient ventilation systems. Workers were classified as-having worked tess than 1 month. 2 months. 3-5 months. 611 months. 1 year, or 2 or more years. Workers in all categories of length of employment had excessive mortality from lung cancer. For example, men employed less than one month had a lung cancer standardized mortality ratio (SMR) of 287. based on supplemental autopsy, clinical, and surgical information. (The standardized mortality ratio is calculated as the observed number of deaths in the exposed population divided by the number of deaths that would be expected in die exposed population, based on mortality rates of an appropriate comparison population. The SMR is frequently used as an approximation to die relative risk.] Hence. OSHA considers that Seidman et al. demonstrated that excess lung cancer risk could be experienced even by workers with exposures of relatively short duration. Berry, Gilson. Holmes. Lewinsohn. and Roach (Ex. 84-020} studied the workers of an asbestos textile factory in the United Kingdom. Berry et al. reported in 1978 that the average exposure level was 5 f/cc and that 8.8 percent of men employed after 1950 had "possible asbestosis." Among men employed after I960, Berry et al. observed 1 percent prevalences of crepitation* "posaibfe asbestosis," or "certified aebestests" at 37,46, and 83 If cc-ytan respectively. Berry ef al. eonduded: "In view of these fintogs there is no room for complacency about the 2 f/cm' standard ABS-008430 lam 025255 511M^^Fodora^Re2ste^^VoL4ajNa215/Fnday^Novernber4^198^/RuIle3an^^eg\dtiorw and efforts should be continued to reduce asbestos dust to as low a level as possible. At this stage it is impossible to state definitely that the standard is inadequate, because its introduction is so recent, and it is essential to follow up groups exposed to low levels in order to improve the data necessary for the formulation of better standards" (Ex. 84n20. p. 109). Finkelstein (Ex. 64-044) studied the development of compensable (certified) asbestosis among 157 Ontario asbestos cement production workers in relation to cumulative exposures (f/cc-years). All workers in the study cohort had at least 15 years of exposure. Ontario criteria for certifying asbestosis. which results in the award of disability pensions, are not inflexible and involve consideration of such factors as history of occupational exposures, dyspnea, crepitations, clubbing of Angers, radiographic signs of pulmonary fibrosis, and abnormal lung function. Certiflcation of asbestosis hence would occur at an advanced stage of disease. Each production worker was classified us having received 0-49. 50-99.100-149, 150-199, or 200-249 f/cc-years of cumulative exposure within the 18 years following initial exposure. Finkelstein calculated incidence rates consisting of number of new cases of certified asbestosis per 100 person-years at risk. [Incidence is the rate at which new cases of asbestosis develop in a given period of time. It is a direct measure of the risk of developing the disease.) Incidence rates were 0.5, 3.4, and 6.5 for workers who received 0-49, 50-99, and 100-149 f/cc-yeara respectively. Finkelstein also calculated the cumulative probability of having developed certified asbestosis by the end of 32 years of latency, and observed that men in the 0-49,50-99, and 100-149 f/cc-years categories had about 10 percent. 55 percent, and 70 percent probabilities respectively. Due to small numbers of men in each category, these estimated probabilities had much statistical uncertainty. Finkelstein stated that the uncertainties in exposure assignments and the selection bias arising from exclusion from the study of workers with less than 15 years of employment may have led to overestimation of risk at low exposures, which was "to some extent balanced by the fact that the end point we studied was certified disability, an advanced stage of asbestosis" (Ex. 84-044, p. 501). OSHA considers that Finkelstein's findings of excess risk from low cumulative exposures are very Important because the outcome, certified asbestosis. was based upon substantial medical 1941-1987. Mortality was observed evidence of severe disability from the during the period 1941-1973. Cumulative disease. exposures, expressed in millions of Dement and colleagues conducted a particles per cubic foot of air times detailed study of plant processes and years exposed (mpef-years). were dust control methods at a chrysotile estimated for each man included in the textile plant during 1930-1875 (Ex. 84- study cohort. Mean exposures for 8 038, 84-037). Exposure histories were cumulative exposure categories were 62. constructed for each worker and 182, 352. 806, and 978 mpef-years. estimates of individual cumulative Hammad et al. (Ex 84-277) suggested exposures in terms of f/cc-days were made. After 1940. exposure levels a mpef to f/cc conversion factor of 1:1.4 for a cement plant. Because cement and usually were in the range of 5-10 f/cc. Workers were categorized as receiving other mineral particles have been extensively used in asbestos products exposures of less than 1000; 1000-10,000; manufacturing, using a conversion factor 10.000- 40,000; 40000-100,000; and greater than 100,000 f/cc-days. Because Dement et aL included holidays and weekends in their estimates of f/ccdays. their estimates of cumulative of 1:1.4 does not appear to be unreasonable for Henderson and Enterline's study cohort Using a factor of 1:1.4,62 mpef-years is roughly equal to 87 f/cc-years. exposure are likely to be overstated. OSHA calculated that Dement et al.'s categories of f/cc-days are roughly equivalent to less than 2.7; 2.7-27.4; 27.4109.8; 109.8-274; and greater than 274 f/ cc-years of occupational exposure. A respiratory canCer SMR of 197.9 was observed in the cumulative exposure category with a mean exposure of 82 mpef-years. Thus. OSHA views this study as having observed excess risk of dying from respiratory Note.--Because Dement et aL counted cancer among men receiving cumulative holidays and weekends for their calculations exposures permitted by the current of f/cc-days, OSHA divided f/cc-days by 365 OSHA PEL. Using a conversion factor of to arrive at estimates of f/cc-years. 1:3 would result in an estimate of Dement et al.'s first 3 exposure categories fall within the range of cumulative exposures permitted by the current OSHA asbestos standard.-------- Among white males with 15 at more years of latency, lung cancer SMR's were 140,279, and 352 in the categories of less than 1000,1000-10,000, and 10.000- 40,000 f/cc-days respectively. The latter 2 categories' excesses of lung cancer were statistically significant. For other non-malignant respiratory disease (excluding both infectious respiratory diseases and bronchitis), the following SMR's were observed for white males with 15 or more years of latency; 382 for men with less than 1000 f/cc-days. 84 for cumulative exposure for the low exposure category that is about 90% higher than that envisaged by the current OSHA PEL Dement et al. commented on the marked differences between the studies of Dement et aL and Henderson and Enterline with regard to risk observed from low cumulative exposures (Ex. 84- . 037). Dement et aL suggested that these differences may be attributable to the fact that Henderson and Enterline studied retirees 65 years or older, with these retirees constituting a select group of survivors. Only 8 of the 35 lung cancer deaths observed by Dement et al. occurred in persons 65 or older. men with 1000-10.000 f/cc-days, and 879 for men with 10.000-40.000 f/cc-days. The excesses were statistically McDonald. Liddell, Gibbs, Eyssen, and McDonald (Ex. 84-085) studied the mortality of 11,379 workers exposed to significant for both the first and third ' category of cumulative exposure. Dement et aL concluded: "Based on chrysotile mining andmilling bom during 1881-1920. Mortality was observed during the period 1910-1975. data from this study, significantly Each worker was classified as having elevated mortality risks are predicted accumulated less than 30, 30-299, or 300 for lung cancer and for asbestosis at ,or more mpef-years by age 45. Using a cumulative exposures of 100 fibers/ conversion factor of 1:3 for mpef to f/cc ccx years in the textile industry" (Ex. these groupings would correspond to 84-037, p. 432). less than 90.90-699, and 900 or more f/ OSHA considers that Dement et al.'s cc-years. Lung cancer SMR's were 93, observations of excess risk from low 118 and 225 for the 3 categories starting cumulative exposures are well- with less than 30 mpef-years supported because of the careful respectively. SMR's for pneumoconiosis estimation of exposure histories for the were 298,1081, and 5400 for cumulative cohort. exposure categories of less, than 30,30- Henderson and Enterline (Ex. 84-048) 299. and 300 or more mpef-years. Hence, studied the mortality of 1075 asbestos McDonald et aL did not observe much production workers who retired during lung cancer risk but did observe an ABS-008431 LAM 025256 Federal Register / Vol. 48, No. 215 / Friday, November 4, 1983 / Rules and Regulations 51103 increase in pneumoconiosis risk in the low exposure category. A case-control analysis used similar exposure categories: however, the analysis was intrinsically incapable of observing an excess risk in the less than 30 mpcf-year category because the controls had also been chrysolite miners and millers. Regarding the different findings on lung cancer risk from low exposures between the studies by McDonald et al. and Dement et al., Dement et al. suggested imprecision of exposure estimates and differences in airborne fiber characteristics as possible reasons for the differences. Dement and Harris previously had found that textile processing produced a higher proportion of long thin fibers falling within the ranges delineated as carcinogenic by the animal studies of Stanton and colleagues (See Section E(3) below. Experimental Data). Finkelstein (Ex. 84-240) also studied mortality from lung cancer, mesothelioma, and other diseases among workers at an Ontario asbestos cement factory established in 1948. The study cohort consisted of 339 men hired' before 1960 who haddjeen employed at the factory for'9 years or more. Each cohort member was classified as having accumulated 8-69,70-121, or 122-420 f/ cc-years of asbestos exposure within the 18 years following initial exposure. These exposure categories, hereafter referred to as Groups A. B. and C. had the following mean cumulative exposures respectively: 44, 92, and 180 ff cc-years. Groups A and B are tow exposure groups. Mortality by cumulative exposure was analyzed starting 20 years after initial exposure, so that the calendar period during which deaths were observed spanned 1968-1980. Cohort mortality was compared to that of Ontario men during 1970-1974. Age-standardized mortality rates per 1000 man-years for specific causes of death were computed for Croups A. B. and C and for Ontario men. Thus, this study did not have the biases from confounding effects of age that can occur when SMR's are being compared among different groups of workers (See Ex. 84-335). One potential problem is that lung cancer mortality may have risen in Ontario during 19751980; this would result in some underestimation of comparison values and hence some overestimation of lung cancer risk. However, this overestimation of risk likely would be slight. Mesothelioma mortality rates per 1000 man-years for Groups A. B. and C were 1.9. 4.9, and ltd respectively. Lung cancel mortality rates per 1000 man: years were 13.6, 26.1,11.9, and 1.8 for Groups A. B. and C and for Ontario men respectively. Approximate relative risks for Groups A, B, and C. as compared to Ontario men. are 8.5.16.3. and 7.4 for lung cancer mortality. Gastrointestinal cancer mortality was also elevated in Groups B and C relative to Ontario men. The lung cancer mortality rates did not consistently increase with increasing estimates of cumulative exposure in that Group C had the lowest lung cancer excess. Finkelstein suggested several potential . explanations, including inaccuracy of exposure classifications, statistical fluctuations resulting from the small size of the cohorts and confounding effects of smoking if there were differences in smoking habits among Groups A. B, and C. Because a consistent dose-response was observed both for mesothelioma mortality and asbestosis in the study group. Finkelstein suggested that the exposure classifications may have been correct. Finkelstein concluded that lung cancer mortality rates "may be raised several fold" (Ex. 84-240, p. 143) at cumulative exposures of 100 f/cc-years. With regard to gastrointestinal cancer, Finkelstein judged that no firm conclusions could be drawn due to the small number of deaths, although there was a trend of increasing risk with increasing cumulative exposure. Mesothelioma death rates were considered by Finkelstein to be related to cumulative exposure. OSHA believes that Finkelstein's study presents evidence of excess lung cancer and mesothelioma risk from relatively low cumulative exposures, namely 44 and 92 f/cc-years. Lung cancer risk may have been somewhat understated by Finkelstein's exclusion of lung cancer deaths which occurred before 20 years of follow-up. Rubino. Piolatto, Newhouse, Scansetti, Aresini. and Murray (Ex. 64-086) studied the mortality of 952 male Italian chrysolite miners and millers during 1946-1975. Mortality from respiratory disease and other causes of death, but not from lung cancer, was excessive in this cohort. Criteria for inclusion in the study cohort were survival until January 1.1946. and at least one month of employment during 1930-1965. Comparison of mortality was made to age. calendar period, and cause-specific mortality rates of Italian males. In addition, mortality rates were compared among the cohort members using a casecontrol analysis and a historical prospective analysis. The cumulative fiber exposure was estimated for each worker in terms of f/cc-years. The investigators.simulated past working conditions and measured dust concentrations during these simulations in order to make more accurate estimates of exposure during periods (pre-1969) in which exposures were not measured. Mean concentrations of asbestos were estimated to range u|> to 50 f/cc before 1950. Compared to Italian males, the overall cohort had statistically significant excesses of mortality from laryngeal cancer, non-malignant respiratory diseases, tuberculosis, cardiovascular diseases, cirrhosis of the liver, and accidents. Lung cancer mortality was elevated only slightly (SMR = 106): however, there was some trend of increasing lung cancer risk with increasing length of follow-up. For example, the SMR for lung cancer was 206 during 1971-1975. Because all analyses of mortality in relation to cumulative exposure consisted of comparisons among the cohort members, rather than between the various exposure subgroups and Italian males, these analyses were incapable of detecting excess risk in the lowest cumulative exposure category. These analyses were capable only of detecting trends toward increased risk with increased exposure. Although the excess lung cancer risk (odds _________ ratio = 2.89) in the high exposure group was not statistically significant. Rubino et al. considered it "likely that the increasing mortality truly reflects the effects of higher exposure" (Ex. 84-086. p. 193). (The odds ratio is an estimate of the relative risk.defined as the ratio of the odds of dying from cancer in the exposed population to the odds of dying from cancer in an unexposed population.) Rubino et al.'s study is unusual in not finding a higher lung cancer risk among workers exposed to relatively high levels of asbestos. OSHA believes that the trends of increasing lung cancer risk with both increasing length of follow-up and increasing exposure are consistent with an asbestos etiology, however. OSHA also believes that the high risks of mortality from other causes such as non-malignant respiratory disease might have obscured lung cancer risk, especially since these high risks appeared within the first 19 years following exposure and asbestos-related lung cancer generally has a longer latency period. Also. OSHA calculated that this study only had 33.5% power to detect a 50% increase in lung cancer risk among workers with 20 or more years of follow-up. Note.--Statistical power quantifies the ability of a study to detect a true increased risk of a specified magnitude and refers to the ABS-008432 LAM 025257 51104 Federal Register / Vol. 48. No. 215 / Friday, November 4. 1983 / Rules and Regulations probability of not misting a true risk. Generally, it is considered desirable for studies to have at least 80% power. Because lung cancer is a relatively common cancer, epidemiologic studies should have at least 80% power to detect a 50% increase in risk of lung cancer. See Ex. 84-338 for a description of OSHA'i method of calculating power. Weill. Hughes, and Waggenspack (Ex. 04-206) investigated exposure-response relationships between respiratory cancer risk and exposure in an asbestos cement plant. Weill et al. did not observe excess risk among men with low cumulative exposures. A total of 5,645 men with at least 20 years of latency since first exposure in either of 2 asbestos cement plants were studied. All men had at least one month of employment before 1970 and their vital status was determined as of December 31.1974. Each worker's cumulative dust exposure during the 20 years from initial exposure was estimated in terms ofmpcf-years. Men were classified in one of 5 different cumulative exposure categories: 10 or fewer, 11-60, 51-100, 101-200. and 201 or more mpcf-years. Using the conversion factor of 1:1.4 suggested by Hammad et al., the 5 cumulative exposure categories would be equivalent to 14 or fewer, 15-70, 71140,141-280, and 281 or more f/cc-years. Respiratory cancer SMR's were as follows, starting with the lowest category: 77, 70. 26. 290, and 226. None of the other causes of death were in excess for workers in the 3 lowest categories. Concerning their failure to detect excess respiratory cancer mortality in their lowest categories of exposure, Weill et al. stated: "Such findings are not necessarily incompatible with a linear response curve at low doses because of the relative insensitivity of currently used epidemiologic methods in detecting slight increases In risk when compared to background. They do indicate, however, that any excess risks at low degrees of exposure are small" (Ex. 84-206, p. 353). Weill et al. noted that the relatively high proportion (25%) of the cohort who were lost to follow-up and assumed alive through 1974 may have led to underestimation of respiratory cancer risk. The upper limits of the 95% confidence intervals of the respiratory cancer SMR's for the 3 lowest exposure categories ranged from about 115 to 150, indicating, in OSHA's opinion, that excess risk could not be ruled out for these categories. Berry and Newhouse (Ex. 84-021) studied mortality during 1941-1976 of a cohort of friction material production workers whose exposures were relatively low. Levels of exposam ranged from less than 1 f/cc to 5 f/cc after 1931. and cumulative exposures for the cohort averaged less than 50 f/ccyears. Although excessive mortality from mesothelioma was observed, there appeared to be little excess mortality from lung cancer. Most of the mesothelioma cases had been exposed to asbestos levels exceeding 5 f/cc. The cumulative exposures to asbestos of the mesothelioma cases were not reported by the authors. Unexpectedly, this study observed excessive mesothelioma mortality but only a non-significant excess of lung cancer mortality. A suable portion of the study cohort had a short follow-up period between their initial exposure and the study cut-off date. For example, 33% of the men had follow-up periods of less than 20 years. Lung cancer risk any be expected to increase in this cohort as the members are followed for a longer period. 3. Summary. A number of epidemiologic studies have examined exposure-response relationships for asbestos and asbestos-related diseases. OSHA recognizes that there are many inevitable uncertainties associated with epidemiologic studies of exposureresponse. Rarely, if ever, are personal samples of asbestos concentrations for exposed workers during theentire period of exposure available. Investigators typically have developed individual indices of exposures from reconstructed occupational histories, recent industrial hygiene data, and assumptions about past working conditions for which exposures were not measured. Another question concerns the factors for-conversion fronthistorical methods of measuring dust concentrations to current methods of measuring fiber concentrations. The ratio of millions of particles per cubic foot (mpcf) to fibers per cubic centimeter (f/cc) has been suggested as ranging from 1:1 to 1:5. Conversion factors of 1:3 for mpcf: f/cc in asbestos mining and milling and 1:1.4 in asbestos production and asbestos cement operations appear reasonable to OSHA. OSHA also recognizes the statistical variation associated with estimation of Standardized Mortality Ratios (SMR's). Such statistical variation, as well as differences in study design, statistical power, and length of follow-up. may account for some of the divergent findings among studies. Because the present OSHA PEL of 2 f/ cc was effective in 1976, there are few, if any, occupational cohorts exposed solely to 2 f/cc or less who have follow up intervals sufficient for the appearance of diseases related to asbestos exposure. At present, OSHA is not aware of any evidence suggesting that intensity of exposure will affect excess risk in a manner different from cumulative exposure, which encompasses both duration of exposure and intensity of exposure. Most epidemiologic studies have observed increasing risk with increasing' cumulative exposure. Duration of exposure may have an independent effect on mesothelioma risk (see Section V., Quantitative Risk Analysis). A worker exposed to the OSHA PEL of 2 f/cc for 50 years would have a cumulative exposure of 100f/cc-years. Hence, studies of workers estimated to have received close to or less than 100II cc-years provide evidence concerning risk from exposure to the OSHA standard, even though the past intensities of exposure may have exceeded 2 f/cc. OSHA considers that the following studies have observed increased risk close to or below 10O f/cc-years. Berry et aL (Ex. 84-020) observed# 1% prevalence of crepitations, possible asbestosis. and oertified asbeatoaie at 37,46, and 83'f/cc-years respectively among asbestos textile workers. Finkelstein (Ex. 84-044) observed a 10% probability of having certified asbestosis among asbestos cement production workers with cumulative exposures of less than 50 f/cc-years who had been observed 32 years since their initial exposures. Finkelatain (Ex. 84240) also observed excessive lung cancer mortality among workers with average cumulative exposures of 44 and 92 f/cc-yeare. Among asbestos textile workers. Dement et aL (Ex 84-036) observed excessive mortality, from lung cancer and non-malignant respiratory diseases at cumulative exposures of less than 1000 f/cc-days and 1000-10,000 f/ccdays (equivalent to less than 2.7 and 2.7-27.4 f/cc-years). Henderson and Enterlina (Ex 84-048) observed about a 2-fold increase in lang cancer mortality among retired asbestos production workers receiving an average cumulative exposure of 02 mpcf-years (equivalent to 87 f/cc-years using a conversion factor of 1:1.4). At relatively low cumulative exposures, the following studies did not observe excess lung cancer risk or observed only a small increase in lung cancer risk. Berry and Newhouse (Ex. 84-021) observed little excess hing cancer mortality among friction material production workers whose cumulative exposures averaged leu than 50 f/ccyears. Mesothelioma cases were observed by Berry and Newhouse; ABS-008433 lam 025258 Federal Register / Vol. 48, No. 215 / Friday, November 4. 1983 / Rules and Regulations 51105 however, their cumulative exposures Francisco and Oakland (Kanarek et al.). were not reported. Weill et al. (Ex. 84- whije other studies did not observe a 206) did not Find increased mortality relationship between asbestos in from any cause of death below 101 drinking water and cancer incidence mpcf-years (equivalent to 141 f/cc-years (Harrington et al.. Meigs et al.. Levy et using a conversion factor r.f 1:1.4) in a al.) (Kang. Ex. 84-139). OSHA is aware cohort of asbestos cement workers. that ecologcal studies generally have Rubtno et al. (Ex. 84-086) observed only certain limitations for determining the a slight excess of lung cancer mortality effects of long-term environmental (and a large excess of mortality from exposure to specific substances, other causes, such as non-malignant including confounding variables such as respiratory disease) despite high migration into and from communities exposure levels. Data were not analyzed and multiple exposures to other by Rubino et al. in such a way as to carcinogens and toxic chemicals. For permit estimation of risk from low example, increased cancer mortality in a cumulative exposures. particular community may result from McDonald et al. (Ex. 84-065) observed occupational exposures rather than from little or no-excess lung cancer mortality carcinogens in the drinking water or among asbestos miners and millers ambient air. Because studies of receiving low cumulative exposures. occupational cohorts do not have as Excess mortality from pneumoconiosis many limitations as ecological studies, was observed in the low exposure occupational studies have the potential group> however. Workers receiving less to be more determinative concerning than 30 mpcf-years by age 45 had an carcinogenic risk to humans, depending SMR of 298 for pneumoconiosis (30 on size of the cohort, length of mpcf-years equals 90 f/cc-years using a ' observation of the cohort, and other conversion factor of 1:3). pertinent factors. Therefore, because As is commonly observed among well-conducted epidemiologic studies of epidemiologic studies of etiologic agents asbestos workers are available and for disease, there are some because inhalation rather than ingestion inconsistencies in the Findings of is the primary route of workplace different studies of workers exposure. OSHA has based its accumulating relatively low exposures. conclusions on the potential Statistical variation, differences in the carcinogenicity of asbestos for sites size distribution of airborne Fibers, other than the lung and mesothelium on imprecision of exposure estimates, and - epidemiologic studies of asbestos competition among different causes of workers. death might explain some of the 2. Epidemiology Studies. Elmes and inconsistencies. Nonetheless. OSHA Simpson studied the mortality of Belfast considers that many well-conducted insulators employed during 1940, mostly studies observed substantially increased 1 in shipyards (1971,1977) (Exhibits 84- risk of death from lung cancer and non- 041 and 84-042). Vital status was malignant respiratory disease among observed during the period 1940-1975. workers receiving cumulative exposures Excess mortality was observed from permitted by the current OSHA asbestosis. lung cancer, mesothelioma. standard. OSHA concludes that these and gastrointestinal cancers (stomach, study results provide evidence of grave jejunum, pancreas, colon and rectum). danger from low cumulative exposures Elmes and Simpson stated that "both to asbestos. in respiratory and in gastrointestinal C. Carcinogenicity ofAsbestos for Sites Other Than the Lung and Mesothelium cancers there is the difficulty of differentiating between mesothelial and epithelial tumours" (Ex. 84-042, p. 176). 1. introduction. A number of studies of Of the total 122 deaths in this cohort, asbestos workers have observed either autopsy information or clinical excesses of cancer at sites other'than information with biopsies and/or the lung and mesothelium. These sites radiographs supplemented the death include colon and rectum, esophagus, certificates for all but 22 deaths. By stomach, larynx, pharynx, buccal cavity, 1966.13 deaths had been coded as kidneys, and ovaries. Based on these gastrointestinal cancers on death studies. OSHA has concluded that certificates. 12 of which continued to be gastrointestinal malignancies appear to classified as gastrointestinal cancers, have been producedby asbestos inhaled after supplemental information had been in the workplace. obtained. Using the expected value of A variety of community-based 5.16 deaths for all non-respiratory epidemiology studies have investigated cancers that were reported by Elmes the effects of ingestion of asbestos in and Simpson. OSHA calculated that the drinking water. One-study suggested gastrointestinal cancer excess was that asbestos In drinking water statistically significant at the 0.06 level increased cancer Incidence in San (one-tailed Poisson test). Elmes and Simpson concluded in their 1971 paper "Cancer of the lung and mesotheliomas do not account for all the excess of deaths: a significant excess of other cancers remain . . . and most of these appear to be in the gastrointestiriHl tract" (Ex. 84-041. p. 235). In their 1977 paper. Elmes and Simpson reported a decline in both asbestosis and gastrointestinal cancer as major causes of mortality among the survivors followed from 1967-1975. Selikoff. Hammond, and Seidman studied a cohort of 17.800 U.S. and Canadian insulation workers (Ex. 84090). Comparison of insulators' mortality during 1967-1976 was made to age and calendar period-specific mortality rates of U.S. white males. The large size of this cohort resulted in very high statistical power to detect increased mortality bom specific causes. Therefore, OSHA considers that this study carries much weight with regard to the question of asbestos-induced malignancies. The investigators sought supplemental clinical, surgical, and. autopsy information in order to determine the extent of misclassification of cancers. Based on death certificate information alone, significant excess mortality was observed from lung cancer (SMR=406). mesothelioma (104 deaths), esophageal cancer (SMR=253), stomach cancer (SMR=128). colon-rectal cancer (SMR=152). laryngeal cancer (SMR=191), pharyngeal and buccal cavity cancer (SMR=159). kidney cancer (SMR=223). all other cancers (SMR=191). and non-infectious respiratory diseases (SMR=319). including 78 deaths from asbestosis. With the exception of deaths listed as "all other cancers", reclassification of deaths based on supplemental clinical, autopsy, and surgical data resulted in slightly higher SMR's for the causes of death listed above. OSHA believes that Selikoff et al.'s reclassification of causes of death is justifiable because misdiagnosis of mesothelioma and asbestosis can occur due to these conditions' resemblance to more common diseases. Furthermore, it is possible that metastases from primary lung cancers could be misdiagnosed as primary cancers of other sites, and viceversa. One disadvantage of reclassification of deaths, however, is that the extent of misclassification in the comparison population of U.S. white males remains unknown. Nonetheless. OSHA believes that, for this particular cohort, the advantage of reclassification in terms ofimproving the certainty about the causes of death outweighs the disadvantage. ABS-008434 LAM 025259 5110^^^FederalReglt*^/VoL4&Na215/Friday^Novhber4^ 1863 /Ru!e3andR2ulatipn^ In the category listed as "all other using a one tailed Poisson test). workers), McDonald and McDonald cancers." 252 deaths were observed Nicholson. Selikoff. Seidman, and (asbestos gas mask workers], Peto et al. compared to 131.8 deaths expected. The Hammond (Ex. 84-251) also studied a (asbestos textile workers), Thomas et at. excess in this category was mostly due cohort of U.S. asbestos factory workers (asbestos cement workers). RossRefrand to 'ncreased mortality from pancreatic and observed a statistically non Coles (shipyard workers), and Jones et cancer (SMR = 281). liver and biliary significant 1.5-fold excess of al. (asbestos gas marie workers) (Ex. 88- tract cancer (SMR=285), prostate gastrointestinal cancer. Newhouse and 256). r.incer (SMR = 137), and brain cancer Berry (Ex. 84-330) observed a When non-gastrointestinal cancers at (SMR = 163). When reclassified gastrointestinal cancer SMR of 138 m a sites other than the hing and according to supplemental information. ' cohort of London asbestos factory mesothehum are considered, at least 4 the pancreatic cancer excess declined workers (80 deaths observed versus 44.2 studies observed excesses at these other greatly and the liver and biliary tract expected) (OSHA calculated that this sites: Selikoff et al., Shettigara and cancer excess disappeared. excess was statistically significant at Morgan. Puntoni et al.. and Stef! and Selikoff, Hammond, and Seidman concluded: "Asbestos insulation workers in the United States and Canada suffer an extraordinary the 0i)5 level using a one-tailed Poisson test). Puntoni et al. (Ex. 84-248) studied the mortality of shipyard workers exposed to asbestos in Genoa, Rafy. McGill. Other studies have not observed excesses of cancers at oflier sites. One issue which (he Agency had to address was how to weigh thejxjaitive increased risk of death of cancer and These workers were also exposed to epidemiologic data versus the nan- asbestosis associated with their other toxic substances, mclading silica, positive epidemiologic data. One employment. This includes increases in benzene, carbon tetrachloride, and consideration is statistical power, the death from lung cancer, pleural polycyclic aromatic hydrocarbons. ability to detect a true risk if such a risk mesothelioma, peritoneal mesothelioma, Compared to the age-specific mortality exists. The Chronic Hazard Advisory cancer of the esophagus, colon and rates of male Genoans. the shipyard Panel on Asbestos (CHAP) convened by rectum, cancer of the larynx, oro workers had statistically significant the U.S Consumer Product Safety pharynx. Kidney, and perhaps stomach. increases in mortality from colon cancer Commission did not find a consistent Some Increases were seen in cancer of (relative risk=1.81). All of the studies relationship between having a higher several other sites, as well, but data are listed in this paragraph also observed degree of statistical power and finding ^ inadequate at this time to permit characterization of their significance, although attention is called to such wider increase" (Ex. 84-080,114). Dement and collegues (Ex. 84-037) observed a statistically non-significant increase in digestive cancer mortality (SMR=13l). When Dement et aL confined their analysis to white males with latency intervals of 15 or more years and examined exposure-response relationships for digestive system cancer, SMR's increased with increasing cumulative exposure, ranging op to 390. Because of the small numbers of deaths, none of the excesses of digestive cancer in any of the exposure categories were statistically significant. McDonald and colleagues (Ex. B4-0B5) observed some increases (not statistically significant) in mortality excessive mortality from King cancer. In addition, other investigators have observed excesses of digestive sytem cancer. Robinson et al. (Ex. 84-082) observed an SMR for digestive system cancer of 121 (not a statistically significant excess) among asbestos production workers. Kletefeld et al. (Ex. 84-140. 84-141} observed an SMR of400 (statistically significant) for digestive system cancer among tremolite and anthophyBite exposed workers mining talc. Mancnso and Coulter (Ex. 84-224) also observed a significantly elevated digestive system cancer SMR among insulation workers. Flnkelstein observed a statistically non-significant 3-fold increase in mortality from digestive system cancer among asbestos cement workers (Ex. 84-240). A number of studies have not excessive mortality from these other cancers (Ex. 84-256], it should be noted that the study that had tha highest statistical power by virtue of studying 17,800 workers, namely that of SeUkeff et al.. observed excesses of gastrointestinal, laryngeal, kidney, and pharyngeal and buccal cavity cancar. Another consideration is the relationship between the magnitude of excess hing cancer risk and excess gastrointestinal cancer risk. CHAP observed that studies with high long cancer excesses were also likefy to hare found gastrointestinal cancar excesses- > Because high lung cancer risks would be expected to have resulted from relatively high exposures, thin observation suggests that gastrointestinal cancar excesses are from esophageal and stomach cancer observed excessive gastrointestinal found where exposures are greater, and from colo-rectal cancer among chrysotile miners and millers in Quebec. Gastrointestinal cancer SMR's increased cancer mortality among workers exposed to asbestos. Nicholson, Selikoff. Seidman. Libs, and Formby (Ex. 84-072). which increases the plausihility of there being a true association between asbestos and gastrointestinal cancer. with heavier cumulative exposures. who studied the mortality of544 in summary.-et least 12 different Seidman. Selikoff, and Hammond (Ex. chrysotile miners and miners from occupational cohorts exposed to 84-087) studied-a cohort of 820 amosite Quebec, observed a gastrointestinal asbestos have been observed to have insulation production workers employed cancer SMR of only 10S. Weill. Hughes, excesses of mortality from at a New |ersey plant during 1941-1845. and Waggenspack (Ex. 84-208), who gastrointestinal cancer, 7 of which were A statistically non-significant excess of studied a cohort of 5.645 asbestos statistically significant. OSHA considers gastrointestinal cancer (esophagus, cement production workers, alao did not that these findings constitute substantial stomach, and colon-rectum) was observe an excess of deaths from evidence for an association between observed (SMR=121). Henderson and gastrointestinal cancar. Other asbestos exposure and gastoointestinal Enterline (Ex. 84-048) studied the investigators who did not observe an cancer risk. The evidence for a mortality during 1941-1973 of 1348 increase in gastrointestinal cancer relationship between asbestos exposure retired asbestos factory workers. Fifty- mortality include Berry and Newhooae and cancer at other sites Is noteworthy five deaths from digestive system cancer (friction materials production workers). yet not as substantial as that regarding were observed compared to 39.9 deaths Rubino et al. (chrysotile miners), gastrointestinal cancer. / expected (SMR=1378) (OSHA Meurman et ail. (asthophyilite miners). Goldsmith (Ex. 84-235) reviewed the calculated that this excess was Brown et aL (talc workers exposed to .evidence for a causal relationship statistically significant at the 0.05 level. asbestos). Weiss (chrysotile factory between asbestos exposure and non- ABS-008435 LAM 025260 Federal Register / Vd. 48, No. 215 / Friday. November 4, 1983 / Rufes and Regulations 51187 pulmonary cancer in 8 *todies of 11 results suggest that additional sensitive occupational cohorts fallowed for at animal experiment* are needed" (Ex. least 20 years since initial exposure. He 84-200, jx 311). concluded that the data "cast doubt on Bolton, Dfcrvi*, and Lamb (Ex. 84-214) whether these ia site-specificity of administered amosite. crocidolite. and asbestos-related cancer" and that "a chrysotile in diet supplements to Wisfar systemic carcinogenic role is more likely rats for periods of up to 25 months. [for asbestos]" (Ex. 84-235, pp. 346-347). Malignant tumors, including GI tumors, 3. Toxicology Studies. A number of were not increased in the asbestos- toxicology studies have been conducted exposed animals compared to the to determine the carcinogenicity of controls; however, chrysotile-treated ingested asbestos. Inhaled asbestos is rats had a statistically significant excess thought to enter the digestive tract when ofbenign tumors. The excess of benign asbestos fibers caught in the mucous tumors in chrysotile-treated rats was lining of the lung are brought up to the largely due to an excess of mesenteric throat and swallowed. Also, fibers hemangiomas. Bolton et af thought that - caught in the nose may travel down to the observed excess of benign tumors in the pharynx (Kang, Ex. 84-139). Evans et chrysotile-treated animals was not al. (Ex. 84-236) conducted an inhalation likely to be doe to asbestos because study in rats of radio-labelled they did not observe asbestos fibers in crocidolite. The radio-labelled OnmWVITIv WWVW if<v\ iVVViav lIaWwf fwv the mesenteric lymphatic tissues of (he aruraIs* esophagus, and gastrointestinal tract Donham. Berg, Will, and Leininger immediately after exposure, and was (Ex. 84-222) fed F344 refs a diet largely excreted in the feces within the consisting of 10% chrysotile. A total of 30 days following inhalation. 189 asbestoa-fed rats and 312 control The National Toxicology Program rata were studied. Regarding thpir (NTP) has administered chrysotile. findings. Donham et al. stated: amosite, taemolite. and crocidolite in "Although the risk differences for feed (1% of diet) to laboratory animals. development of colon tumors Increased incidences of tumors have not specifically is not statistically been reported in any of the NTP studies significant at the 5% level, we feei there (Exhibits 84-225.84-228, 84-227, 84-228). - is suggestive evidence that ingested fit some of the studies, relatively short asbestos may have some role in colon fibers were administered. The asbestos carcinogenesis..(Ex. 84-222. p. 1080). dose of 1% in diet was not the Smith. Hubert Sobel. Peters, and maximally tolerated dose (MTD), unlike Doerfler (Ex. 84-193} administered most carcinogenesis bioaasays. amosite asbestos in drinking water to . Ward. Frank. Wank. Devor. and Syrian hamsters of the Lak: LVG strain Tarone (Ex. 84-300) exaaiiwsd (he effect for periods ranging up to 22 months. of oral exposure tar amoatta or chrysotile There was soma clustering of malignant on gastishtaeetiaal (O) caw;biogenesis tumors, including a peritoneal among PM*rats whohad been injected mesothelioma. a pnhnonmycascinoana. with azoxym*thane. Azoxymethan* is a and 2 sqwanwws eaM carcinomas of the well-recognized intestinal carcinogen for forestomadt. in the hamsters exposed to laboratory animals. Rats were also amosite [4 of 180 animals). None of the exposed to amosite atone and control animals developed these types azoxymethane aleae. The exposues of malignant tumors. Smith et al. did not period was 10 weeks and the rats were consider (he cluttering of malignant observed foe 95 weeks or more (some tumors in amosite exposed animals to rats did not survive as tong ee 95 be related to their ingestion of amosite weeks). Although amosite and chrysotile because these types of tumors have exposure did not significantly increase been observed m central Syrian the incidence of G1 cancer among rats hamstess of the Laic LVG strain by other given azoxymethane injections, the 49 investigators, rats receiving amosite alone had an . OSHA considers feat there is some unusually high incidence (32%) of colon evidence that ora! ingestion of asbestos carcinoma compared to the incidence is carcinogenic to laboratory animals: among historical control F344 cate however, this evidence is rather maintained in the same Laboratory. inconsistent. The generafizability of the Ward et al commented on these non-poeitrse NTP madias is somewhat findings: "The results of these limited by the hw dotes and short fibers experiments suggest that oral asbestos that were administered to the animals. exposure may have caused an increased 4. Summary. In summary, at least 12 Incidence of intestinal tumors in male epidemiologic studies have observed F344 rats .. Although our findings did increased mortality from not conclusively demonstrate the eo- gastrointestinal cancers among workers carcinogenic or carcinogenic effect of exposed to asbestos. 7 of which were asbestos for the intestinal tract, our statistically significant. Also, at least 4 epidemiologic studies have observed excesses of cancer at sites other than the respiratory tract, mesolhelium. and gastrointestinal tract. A number of other studies have not obaerved increases in mortality from cancers at sites other than (he fang and mesotheiium. It is possible far physicians to misdiagnose peritoneal mesothelioma as gastrointestinal or other cancers: however, the excess of gastrointestinal cancer persisted even after Elmes and Simpson and Selikoff et al. reclassified causes of death using supplemental ' autopsy, surgical, and clinical information. OSHA believes that the results of these reclassifications constitute additional evidence for an association between asbestos exposure aqd development of gastrointestinal cancer. OSHA regards the numerous epidemiologic studies indicating increased risk from gastrointestinal cancer ae outweighing non-positive epidemiologic studies and non-positive and equivocal findings in animals ingesting asbestos. Dose-response relationships far gastrointestinal cancer are characterized less welt than for respiratory system cancers. Nonetheless. OSHA concludes thet gastrointestinal malignancies should be included in quantitative analyses of excess cancer risk horn asbestos exposure because such cancers have made substantial contributions to the increased mortality of many cohorts of asbestos workers. Otherwise, excess cancer risk from asbestos exposure would be understated. The excesses of malignancies at sites other than the fang, mesotheiium. and gastrointestinal tract observed by SeBkoff et al. are particularly noteworthy because of the large size of Selikoff et al.'a cohort. The large size of their cohort resulted in a high degree of statistical power and a high degree of stability for the observed SMR's. OSHA considers that asbestos might induce cancer at sites other than the lung, mesothefium. and gastrointestinal tract based on the studies finding cancer excesses at these other sites, especially the study of Selikoff et al. OSHA will not attempt to quantify the excess risk at these other sites in relation to exposure. OSHA views the evidence fur asbestos inducing gastrointestinal tract cancer as stronger and more consistent than the evidence for asbestos inducing cancer at these other sites. D. Effects of Cigarette Smoking 1. Introduction. A multiplicative effect of asbestos exposure and cigarette smoking with regard to producing ABS-008436 lam 025261 51108 Federal Register / Vol. 48. No. 215 / Friday. November 4, 1983 / Rulea and Regulations increased lung cancer risk was shown in a 1908 paper by Selikoff. Churg. and' Hammond (Ex. 2-5). Subsequently, other studies of occupational cohorts confirmed this finding (Selikoff, Scidman. and Hammond. Ex. 84-190; 1 lammond. Selikoff. and Seidman. Ex. 84-047). Cohen et al. (Ex. 84-031) observed poorer particle clearance from the lungs of smokers than from the lungs of non-smokers. This finding may help to explain the higher lung cancer risk of smoking asbestos workers. In addition, smoking asbestos workers have been reported-to be at higher risk of asbestosis and chest X-ray abnormalities, including pleural plaques (Hammond et al., Ex. 84-047; Weiss, Ex. 84-097; Weiss, Levin and Goodman, Ex. 84-099). There is no evidence for an association between cigarette smoking and either mesothelioma risk or gastrointestinal cancer risk (Hammond et al., Ex. 84-047). 2. Respiratory Cancer. Hammond et al. (Ex. 84-047) collected smoking histories from 8220 of the 12051 insulation workers with a follow-up period of 20 or more years since initial exposure who had been studied by Selikoff and colleagues. The mortality experience of these workers was ubserved during 1907-1978. Of the 8220 workers who answered the questionnaire on smoking habits in late 1968, 8841 were either current or past cigarette smokers. 488 had a history of pipe or cigar smoking, and 891 had never smoked regularly. The comparison population was drawn from the American Cancer Society's long-term prospective epidemiologic study conducted by volunteers, and consisted of 73,763 white men who had no more than a high school education, were not farmers, were alive as of January 1. 1987, and had a history of occupational exposure to dust, fumes, vapors, gases, chemicals, or radiation. The major advantage of this comparison population was the availability of agespecific mortality rates by smoking status. Also, men with the above described education and occupational histories likely would resemble the insulation workers more than the general U.S. white male population. Age-standardized lung cancer mortality rules for controls and for asbestos workers are given in Table 5. As shown in Table 5. non-smoking asbestos workers had a mortality rate from lung capcer that was 5 times higher than that of non-smoking controls. The lung cancer mortality of smoking, asbestos workers was also 5 times higher than the controls with a history of cigarette smoking. Thus, for both smoking and non-smoking asbestos workers, the relative risk of death from lung cancer was about 5-fold. Hence, the relationship between cigarette smoking and asbestos exposure can best be described as multiplicative in nature. Taole 5 Age-StaiuUrdlzeo Lung Cancer Oeatn Rates by Sacking Status ana Occupational History of Asbestos Exposure broup Exposure to Asbestos? History of Cigarette issixing^ Oeatn Rate* Mortality Difference Mortality Ratio Controls Asbestos worxers .Ho res Ho 11.3 0.0 1.00 ho 58.4 47.1 5.17 Controls Ho Yes 122.0 *111.3 10.65 Asbestos workers Yes Yes 601.6 590.3 53.24 Rata per 100,000 aan-years standardized for age on the distiiout>on of the sun-years of at) the asbestos workers (based on death certificate Information) From haassond et at.. Ex. 84 047 Regarding these data, the NIOSH/ OSHA Asbestos Work Group stated: "The combined effect of smoking and asbestos exposure appears to be more than simple addition. If the combined effect were additive, one would expect death rates of 169.7 per 100.000 man-years among asbestos workers who smoked. This rate was derived from the sum of the baseline rate (11.3) plus the excess over that baseline due to asbestos (58.4-11.3 =47.1) plus the excess due to smoking (122.8-11.3=111.3). The data seem rather to satisfy a multiplicative model. It was shown that smoking alone increased the death rata about 11 times, and asbestos alone increased it 5 times. Therefore, for a multiplicative model the mortality ratio for those exposed to both asbestos and smoking would be 55 (5 times 11) times greater than. those who were exposed neither to asbestos nor to smoking. The mortality ratio for those exposed to asbestos and to cigarettes was actually 53.24" (Ex. 84-320, p. 27). Selikoff, Seidman, and Hammond examined the effects of cigarette smoking and asbestos exposure among 582 amosite production workers, 567 of whom had smoking histories (Ex. 84- 190). As in Hammond et aL's study, comparison was made.to age and causespecific mortality rates within each smoking status category of the American Cancer Society (ACS) cohort. Non smoking amosite workers had a greater than 5-fold increase in lung cancer risk compared to non-smokers in the ACS cohort, while smoking amosite workers had an almost 5-fold increase compared to their smoking ACS counterparts. Based on the excess lung cancer risk in non-smokers. Selikoff et al. stated that asbestos exposure alone produced an increased risk of lung cancer, although the increase among non-smokers was limited in terms of total numbers of additional deaths. For cigarette smoking asbestos workers, Selikoff et al. described the increased numbers of lung cancer deaths as "devastating." Selikoff et al. observed no increase d risk of death from mesothelioma. ABS-008437 LAM 025262 Federal Register / Vol. 49. No. 21S / Friday. November 4. 1983 / Roles and Regulations 51109 gnstrointenstinal cancer. andasbestosis among smoking aaasile workers compared to non-smoking arnosite winkers. 3 Lang Disease and Chest X-ray 1 ^turmalities. Hammond et a!, reported that asbestosia workers who smoked one or mare pecks of cigarettes per day nail an asbestos mortality rate 2.8 times higher than that of asbestos workers who had never smoked regularly (Ex. 84-047). Hammond et aL also reported that ex-smokers who were asbestos workers bad substantially lower death rates than asbestos workers who had not quit smoking. Weiss (Ex. 84-009) conducted a cheat X-ray and questionnaire survey of 100 aabratns textile workers. Wetss reported that smoking workers had a 40% prevalence of pulmonary fibrosis and non-smoking workers had a 24% prevalence of puknenary fibrosis. A gradient in risk of fibrosis waa observed for both increasing age asd increasing dura taon of exposare. The age distributions of kh* smokers and non- smokers were similar, as were the median durations of exposure to asbestos. Weiss concluded that both asbestos exposure and cigarette smoking were associated with pulmonary fibrosis and that smoking asbestos workers had a higher prevalence of fibaosir relative to non smoking asbestos workers. Weiss did net indicate whether the difference in puhnenery fibrosis prevalence between smokers and nonsmokers was staflstioally significant. OSHA tested the sigoificaace af (be difference in prevalence at pnknonary fibrosis between smokers and nonsmokers sting a chi-squared mat of proportions and did not find n significant riiffrrcrs (p greater than 0.1). In tsn. Weiaak. Levin, end Goodman (Ex. 84-099) reported the reaultn af a survey of 45 men aged 40 or more whs had worked 5 or more years in an asbestos manufacturing plant. The prevalence of pleural plaques was greater tn cigarette smokers; however, there was soma confounding of this relationship by cmmdklive asbestos exposure, which alsnseemed'to influence the prevakm& of pleura1 plaques. Weira et aL rtaSerfc "Conclusions are restrained by the small number of workers in this investigation. A clcur-cut answer to the question as to whether the association between plaques and smoking is spurious will require a much larger research effort, preferably In h cohort Study" (Ex. 84-099. p. 429). Pearle (Ex. &4-4J79)surveyed 131 asbestos-exposed shipyard workers to determine the relative contribution of asbestos exposure and smoking to lung function decrements and chest X-ray abnormalities. Pearle found that both cigarette smoking and asbestos exposure reduced FEVl (forced expiratory volume in 1 second) and FVC (forced vital capacity), with combined exposure having a "cumulative or possibly synergfstic effect" (Ex. 84-079. p. 39). Pearle also reported that both smoking and asbestos exposure produced pleural and interstitial abnorroalties and that smokng appeared to be the primary factor in airways obstruction and diffusion impairment. Berry. Gilson, Holmes. Lewinsohn, and Roach (Ex. 84-020) studied 379 men employed at an asbestos textile mill as of June 30.1966. Smoking histories were available for 376 men. Men were classified as having never smoked, as ex-smokers, or as current smokers (light, medians, or heavy cigarette consumption). The mean cumulative exposures (fibers/cubic centimeteryears) were similar among the smoking status gcoaps. and age adjustments were mada to account tor the younger ages of non-smokers. Significantly greater prevalences of crepitation* and small radiological opacities were observed in heavy smokers and ex-smokers compared to non-smokers and light smokers. For example, 16% ofheavy smokers employed after 1950 had small . radiological opacities compared to 4.8% of never-smokers employed after 1950. Kilbum (Ex: 84-237) has reviewed the studies of the relationship between cigarette smoking and X-ray evidence of pulmonary fibrosis. Kilbam criticized Weiss's use of a definition for pulmonary fibrosis other than that of the ILO in his 1971 study, as well aa the mail number of workers reported on in Weiss's 1971 paper and his other papers. A study by Samet et al. which included a relatively large number of subjects and which dtd not find an effect of cigarette smoking on radiologic abnormalities characteristic of asfcestosfs was cited. Kffburn concluded that cigarette smoking neither produced X-ray appearance of pulmonary fibrosis nor contributed to fibrosis resuming frnm asbestos exposure. in summary, there is some evidence that smoking asbestos workers have ii hither risk of mortality from asbestosis. as well as a higher prevalence of crepitations, lung function decrements, and small radiological opacities. This evidence is inconclusive. 4. Aftribtthng Probable Etiologies to Lung Cancer Among Individual Asbestos Workers. Enteriine (Ex. 84126) analyzed the probability that any single case of lung cancer in a person with known exposure to asbestos could be attributed ta the asbestos exposure. His paper emphasized that it cannot be stated wikh certainly that a lung cancer in an individual worker was due to asbestos exposure: rather, statements can only be made concerning probabilities of cause and effect. He converted observed relative risks into probabilities using the following formula: l*dtnm#l fmmin; et it** Cmsttr mh^ Out U MMM UmitI' mimwim if t m tifk-------" " Usii^ Seliurff et al.'s data on insulators, where a relative risk of about 4.5 was observed for smoking asbestos workers and non-smoking asbestos workers abkciEnterliae estimated a probability of 75% that lung cancers were attributable to asbestos exposure for both smoking and non-smoking asbestos workers (3.5K5 \19G i about 9.751 Enterline's paper is important io exploring the extent to which asbestos can be considered the etiologic agent for lung cancer in exposed workers. However, in the case of smoking asbestos workers, dichotomizing causation as either due to smoking or due to asbestos does not seem appropriate to OSHA because of the factor of synergism between cigarette smoke and asbestos. Using Seiikoff et al.'s data, probabilities of causation were estimated by OSHA for each etiologic factor (See Tables 8 and 7). ABS-008438 lam 025263 SHIP Federal Register / Vol. 48. No. 215 / Friday. November 4, 1983 / Rules and Regulations Table 6 LUN6 CANCER MORTALITY BY SMOKING STATUS^ Lung Cancer Rate per 100,000 per Year Smoking Asbestos Insulators: 362.0 Non-Smoking Asbestos Insulators: 40.4 . Smoking U.S. Males: 74.4 Non-Smoking U.S. Males 9.2 1/From Sellkoff et al., 1979 (Ex. 84-090) Table 7 PERCENT ATTRIBUTABLE RISK FROM AS8EST0S EXPOSURE ANO SM0KIN6 Lung Cancer Risk Attributable To: Background Asbestos alone Smoking alone Interaction of Smoking and Asbestos Formula 9.2 40.4 - 9.2* 31.2 74.4 - 9.2* 65.2 362 - 9.2 31.2 - 65.2 - 256.4 X Attributable Ri'sk for Asbestos Workers: Smoking Non-Smoking 2.5% 22.8% 8.6% 77.2% 18.0% 70.8% 90S 100% While OSHA's calculations differ . from Enterline's calculations of attributable risk by including the synergism factor, the probability estimates do not differ much. According to OSHA's calculations, asbestos exposure contributes to 78.4% and 77.2% of lung cancer deaths among smoking asbestos workers and non-smoking asbestos workers respectively. 5. Summary. OSHA considers that asbestos exposure by itself can cause lung cancer, asbestosis, mesothelioma, and gastrointestinal malignancies. thereby presenting a significant health risk to exposed workers and a grave danger under current exposure conditions. Cigarette smoking by itself can cause lung disease and lung cancer. Cigarette smoke and asbestos exposure appear to have a multiplicative . relationship for causation of lung cancer. Mortality from asbestosis has been reported to be increased in cigarette-smoking asbestos workers compared to non-smoking asbestos workers, although this evidence is inconclusive. Pleural plaques and the radiologic abnormalities characteristic of pulmonary fibrosis have also been reported to be more prevalent in cigarette-smoking asbestos workers; this evidence is also inconclusive. Ex smoking asbestos workers have been reported to have decreased lung cancer risk relative to smoking asbestos workers. There is no evidence for.any relationship between cigarette smoking '' and induction of mesotheliomas and gastrointestinal malignancies. E. Relative Carcinogenicity and Toxicity ofDifferent Fibers 1. Introduction. OSHA has reviewed the data concerning the'relative carcinogenicity and toxicity of different asbestos fiber types and has decided not to make distinctions in this emergency temporary standard by asbestos fiber type with regard to regulatory provisions designed to protect workers from the harmful effects of asbestos exposure. Mainly, this is because lung , cancer is the leading cause of death associated with asbestos exposure and there do not appear to be differentials in lung cancer risk by fiber type. Some investigators and committees have suggested that different fiber types of asbestos have differences in their carcinogenic potency, including Enterline and Henderson (Ex. 64-122), McDonald and McDonald (Ex. 64-154). Weill et al. (Ex. 64-206). Acheson and Gardner (Exs. 64-015 and 84-243), Muir (Ex. 64-350), and the Advisory Committee on Asbestos (Ex. 84-216). These scientists generally believe that crocidolite and amosite are more carcinogenic than chrysotile and anthophyllite. Based on these scientists' recommendations, both the United Kingdom and the province of Ontario, Canada have promulgated the following standards for asbestos: 0.2 fibers per cubic centimeter (0.2 f/cc) for crocidolite 0.5 fibers per cubic centimeter (0.5 f/cc]". for amosite 1 fiber per cubic centimeter (1 f/cc) for chrysotile and all other forms of asbestos i ... Recently, the United Kingdom has announced reductions in the limits for chrysotile and amosite. It is important to note that the question of differentials in potency by fiber type primarily concerns Induction of mesotheliomas, not asbestosis and lung cancer. For mesothelioma, Acheson and Gardner in 1979 described "a powerful case. * * * that crocidolite has been more dangerous than chrysotile and anthophyllite" (Ex. 84-216, p. 11). whereas they characterized evidence for ABS-008439 LAM 025264 Federal Register / Vol. 48, No. 215 / Friday. November 4, 1983 / Rules and Regulations 51111 fiber type differentials in risk from lung cancer and asbestosls as "inconclusive" and "slight" respectively (Ex. 84-210, p. IDIn 1963. Acheson and Gardner stated: "Peritoneal mesothelioma has an almost exclusive relationship with exposure to the amphiboles crocidolite and amosite" (Ex. 84-243, p. 8). Concerning lung cancer, they noted that although one study (Weill et al. Ex. 84-206) had found a higher risk among workers exposed to crocidolite. no "clear distinction" of risk by fiber type could be found when the slopes of dose-response curves were compared. Hence, they concluded that available evidence did not support the assumption that chrysotile was a less potent lung carcinogen (Ex. 84-243. p. 8). In 1983. the Chronic Hazard Advisory Panel on Asbestos (CHAP) convened by the U.S. Consumer Product Safety Commission also reviewed the evidence for fiber type differentials in carcinogenic risk (Ex. 84-258). They concluded that there were inconsistent and inconclusive findings with regard to lung cancer differentials by fiber type,. For peritoneal mesothelioma. CHAP concluded that epidemiologic studies suggested diet this disease was most common in amosite workers, teak common in crocidolite workers and rare or non-existent in chrysotile workers.' Evidence for a fiber type differential in pleural mesothelioma risk was not considered by CHAP to be as substantial as for peritoneal mesothelioma. Some health scientists who believe that evidence links crocidolite exposure to a substantial increase in mesothelioma, risk have suggested separate regulatory treatment for crocidolite (Ex. 84-218). OSHA has been urged to consider requiring that workers exposed to crocidolite or asbestos mixtures containing crocidolite wear appropriate respirators, irrespective of airborne concentrations of asbestos and irrespective of,compliance with the permissible exposure limit. Furthermore, OSHA has been urged to consider requiring that employees exposed to crocidolite be informed of the potential greater risks associated with crocidolite and the reasons for requiring respirator usage whenever crocidolite exposures occur. OSHA invites comments on this suggestion and on other alternative approaches, to separate regulatory treatment of crocidolite. There are also scientific questions concerning the relationship between fiber dimensions and ability to cause disease. OSHA believes that asbestos fibers longer than 5 micrometers (>un) cause lung disease and cancer, provided that the ratio of fiber length to fiber width is 3:1 or greater. Evidence for risk from fibers less than 5 pm in length is inconclusive. A critical analysis of the evidence concerning risk differentials by fiber type and fiber dimension follows. 2. Epidemiologic Data--Introduction: Commercial asbestos fiber types including amosite. chrysotile, and ,crocidolite. have been observed to be associated with elevated risks of asbestoeis, lung cancer, and mesothelioma whether exposures occurred to a single fiber type or to various combinations of fiber types (NIOSH/OSHA 1980. Ex. 84-320). Excess lung cancer risk and asbestosis have been observed in anthophyllite asbestos workers; however, no cases of mesothelioma have been reported among anthophyllite workers. For amosite. an association between exposure and disease has been pbserved by the following investigators: Seidman et aL (Ex. 84-087), Anderson et al. (Ex. 84-017). and Murphy et al (Ex. 84-311). For chrysotile, the following investigators have obtained positive findings: McDonald et al. (Ex. 84-069), McDonald and Fry (Ex. 84-084), Liddell et al. (Ex. 84-059), Nicholson et al. (Ex. 84-072). Rubino et al. (Ex. 84-068), Dement et al. (Ex. 84-037), and Acheson and Gardner (Ex. 84-015). For crocidolite. positive findings were observed by the following investigators: Jones et al. (Ex. 84-138). Hobbs et al. (Ex. 84-132) and McDonald and McDonald (Ex. 84-154). Meurman et al. (Ex. 64-181) observed an excess lung cancer risk among anthophyllite miners in Finland. In addition. Aumerous studies have observed asbestosis, lung cancer, and mesothelioma among workers exposed to mixed fiber types (Hughes and Weill, Ex. 84-135; Weill et al.. Ex. 84-206; Jones et al* Ex. 84-138; Berry et al.. Ex. 84-020; Elmes and Simpson. Ex. 64-042; Peto et aL 84-80; Lacquet et al, Ex. 84-144; Selikoff et al.. Ex. 84-089; Robinson et al.. Ex. 84-082; and Balselga-Monte and Segarra. Ex. 84019). Also, several studies of. talc miners and millers, where the talc contained tremolite and anthophyllite. have observed excesses of lung cancer and lung disease, and reductions in pulmonary function (Klelnfeld et al.. Ex. 84-140. 84-141: Brown et al.. Ex. 84-025; Gamble et al.. Ex. 84-181). In addition, zeolite, an asbestiform mineral found in the soil and water in two villages in Turkey, may be responsible for those villages' high mortality rates from mesothelioma (Artvinii and Baris, Ex. 84-018; Baris et aL, Ex. 84-110). OSHA is not aware of any epidemiologic studies which compare workers exposed to different fiber sizes within the same industrial process. Undoubtedly, this is due to the scarcity of such occupational groups since most occupational populations have bean exposed to h mixture of long uiul short fibers. Therefore, most of the d.ilu on differential risk by fiber size are from experimental studies rather than epidemiologic studies and will be discussed in section (D)(3) Experimental Data. Evaluating the relative carcinogenicity of the different asbestos fiber types involves comparison of results from different epidemiologic studies. OSHA believes that it is important to note the potential difficulties of comparing different occupational cohorts. Standardized Mortality Ratios (SMR's) are risk measures that are dependent on the age distribution (as well as other factors) of the particular study population. This is because the values used for comparison, referred to as the expected values, are derived from mortality rates specific for age. race, sex, and calendar period in the comparison population. Cancer risk rises with Increasing age. so that an older population would have higher expected values for cancer mortality. Misleading results may be obtained when comparing risk measures for cancer among study populations with different underlying age distributions. An example of the misleading results that may be obtained in such an instance is given in Ex. 84-335, which illustrates that SMR's of diverse study populations may be the same while the actual excess risk of mortality may vary greatly because of differences in age distribution. Sex and calendar period of observation also affect expected values, because women have lower mortality rates from lung cancer and lung cancer has risen greatly since the 1940's. Length of follow-up is another pivotal variuble because of the long latency periods necessary for development of lung cancer and especially mesothelioma. Epidemiologic Studies of Lung Cancer. Mesothelioma, and Absestosis: The following studies examined the question of differential lung cancer risk by asbestos fiber type by comparing the mortality experiences of different occupational cohorts: McDonald and McDonald (Ex. 84-154). Henderson and Enterline (Ex. 84-048), Weill et al. (Ex. 84-208). McDonald and Fry (Ex. 84-064). and Acheson et al. (Ex. 84-015). Some of these studies also addressed the question of differential mesothelioma risk by fiber type. Some additional studies of workers exposed to a single fiber type are also pertinent to evaluation of risk by fiber type. ABS-008440 LAM 025265 51112 Fedora) Register / Vol. 48. No. 215 /^Friday,, November 4, 1963 J Rules and Rggufy'tiona including McDonald et al (Ex. 84-086), Hobba at al. (Ex. 84-182), Weiss (Ex. 84098), Dement et al. (Ex. 84-037}, and Berry and Newhousa (Bx. 84-021]. McDonald and McDonald (Ex. 84-154} suggested that crocidoiite wire more carcinogenic than chryaotite baaed on their comparison of the mortality of Canadian chrysotile miners and millers with that of Canadian workers manufacturing crocidolite-containing gas masks. For gas mask workers, 13% and 18% of the observed deaths were from tang cancer and mesothelioma respectively. In contrast, 6% and 0.28% of die observed deaths among chrysolite miners and millers were from lung cancer and mesothelioma respectively. One potential problem was that the personnel records of the crocidotite workers were incomplete, which could result in biases toward observing greater or lesser risk In this cohort. An unpublished report by Dr. Han K. Kang (Ex. 64-139) commented as follows on this comparison: (1) Two major potential confounding factors, fiber site distribution end fiber concentration, were not analysed by the authors. If the gas mask plants had higher fiber concentrations and/or a higher proportion of long ttyta fibers than the minee, a higher cancer risk would be expscted in the plants. (2) Comparison of riek measures such as Standardized Mortality Ratios (SMR's) and Proportionate Mortality Ratios (PMR's) might not be appropriate between the two occupational cohorts because of potential differences in age distribution and length of follow-up period. Henderson and Enterline (Ex. 84-048} compared mortality of retired workers exposed to different asbestos fiber types. After adjusting for differences in cumulative dust exposures, Henderson and Enterline reported the following respiratory system cancer SMR's for . workers in an asbestos factory and workers manufacturing asbestos cement 247 for workers receiving chrysotile exposures only, 364 for workers receiving amoeite exposures only, and 461 for workers receiving combined exposures to chrysotile and crocidoiite: Asbestos cement pipe workers had a very high hmg cancer SMR of 522. Henderson and Enterline stated: "Men who worked in the production of asbestos cement ptpe exhibited a higher risk of respiratory cancer, as did men with tome crocidoiite asbestos exposure. Because these two groups overlap, we could not be certain that oroddollte asbestos was responsible for the Increased risk" (Ex. 84-064). Concerning Henderson and Entertine'i one or more years of employment and at findings. OSHA notes that the asbestos least IS years of follow-up bad a lung cement pipe exposures may have been cancer SMR that was approximately dustier than earlier estimates indicated twice as high as that of Quebec miners and that some of the workers thought to with one or more years of employment have been exposed only to chrysotile and at least 20 years of follow-up. may have had Indirect exposure to . Comparison of these two groups of crocidoiite. Also, workers in the miners is problematic because of . asbestos cement pipe operation potential difference* in age distribution, probably were exposed mostly to fiber concentrations, and length f chrysotile (Kang, Ex. 84-139). Another employment [Kang. Ex. 84-139). potential confounding factor was that Acheson. Gardner. Fippard. and exposure levels were uncertain for each Grime [Ex 84-015) studied mortality af group of workers. female workers manufacturing mostly Weill, Hughes, and Waggenspack (Ex.. chrysotile-containing gas masks in ona 84-206) studies the mortality of asbestos plant (Plant 1) during 1939 and female cement pipe workers in two plants. In workers mostly the first plant, workers were exposed to crocidolite-containinggas masks and both crocidolrte'(3% of product] and some chrysotile-containing gas masks in chrysotile. In the second plant workers a second plant in 1950 (Plant?). Plant 2 were exposed to chrysotile, amosile.(l% had a larger excess of mortality than of product), and crocidoiite Plant L Ages at initial erapkiymant ware; (infrequently). very similar at the 2 plants. At Hast 3,7 .. Weill et al reported respiratory lung cancer deaths ware observed system cancer SMR's as follows for versus 4.8 expected (SMR=145, workers receiving cumulative asbestos statistically non-significant). One death exposures greater than 200million front pleural masathaHomi was , particles per cubic foot-months (mpef- observed in female worker from Plant - ' months) during the first 20-years of 1. employment: At Plant 2,15lung canoer deaths were No croddolits exposure --............SMR--Mg Intermittent exposure to crocido iite_______________________ SMBs337 Steady exposure to ouddbBte__ _SMR=241 observed versus (L2 expected (SMR241. p teaa-OuM). Msaotheboma was mentioned on 5 women's death certificates. 3 at which coded mesothelioma aa tha.audariyvag oansa at death. A atatiihcaify significant OSHA notes that the apparent differences in limg cancer risk by fiber type observed by Weill et aL are based on relatively small numbers of deaths. Hence, statistical variation may be considered a poaaible explanation for the observed differences. It is not known if the 75% bfiow-op rate may have contributed to these apparent differences. Weill. Ziskind. Waggenspack. and Rosaiter (Ex. 84-207} suggested that crocidoiite exposed workers had poorer pulmonary function than workers exposed to other forms of asbestos. They reported that cnxadolite workers increase in ovarian cancer asortaltty was also observed at Plant 2 (12 deaths . observed venae 4.4expected, SVffis273, p Isas titan (MU). Possibly, the deaths certified as das to ovarian cancer may have actually bean misclaasified peritoneal Mesotheliomas. Achaaoo et aL concluded that "the pattern of mortality is consistent with the view that mesothelioma (and possibly ovarian canaar) is particularly associated with exposure to crotidofitd'* (Ex. 84-015, p. 347}. ' Acheson et al.'s study has 2 advantages relative to other studies of populations exposed to different had smaller hmg volumes, a higher asbestos fiber types: both the age prevalence of X-ray changes. Tower distributions and lengths of follow-up FEVTs. and reduced diffusing capacity. were similar for Plant 1 and Hant 2. Nate, FEVl reins to ths valum of sir forcibly expired in on* soared, and diffusing capacity refers to the ability at oxygon to leave the sfesoh and enter the pulmonary Neither air concentrations nor durations of employment were reported for die 2 plants, however. Because Plant 2 had been operating longer than Plant 1. Plant bloodstream Redactions in FEVl and diffusing capacity are tadtestirv of Impaired lung fenctioa. 2 workers may have had a greater median length ofemployment Manual manufacturing processes were employed A comparison between CanaJan at Plant 2. white mechanical processes chrysotile miners studied by McDonald were employed at Heat 1. Perhaps die et al (Ex. 84-065} and Australian manual procsss led to higher dust croddolito miners studied by Hobbs at exposures for the Individual workers at al. (Ex. 84-132] has been made by Kang Plant 2. Yat It te alee possible that the (Ex. 84-139). The Australian miners with mechanical process at Plant 1 reduced ABS-008441 lam 025266 Federal Register / Vol. 48. No. 215 / Friday. November 4. 1983 / Rules and Regulations 51115 Table 8 MESOTHELIOMA DEATHS AMONG COHORTS OF CHRYSOTILE WORKERS Study X of deaths from mesothelioma McDonald at al. (Ex. 84-065.) - miners Nicholson et al. (Ex. 84-072) - miners Dement et al. (Ex. 84-037) - textile workers' Rubino et al. lEx. 84-086) - miners 0.3 0.5 0.5 0 Hammond et el., however, observed a marked as for peritoneal mesothelioma.. higher percentage of deaths from according to CHAP (Ex. 64-256, p. 126). mesothelioma (Ex. 84-047). Hammond et Summary ofEpidemiologic Data: In al. studied insulation workers exposed summary, there is evidence that al) predominantly to ehrysotile and to some commercial asbestos fiber types, except amosite, but not to crocidolite. From for anthopbyllite. cause asbestosis. death certificates and other clinical mesothelioma, and lung cancer in material. 175 deaths (7.7%) were humans. Anthophyllite has been shown attributed to mesothelioma out of a total . to cause lung cancer and asbestosis, but of 2271 deaths. Peto (Ex. 84-168) Studied no mesotheliomas have been reported a factory which used mostly ehrysotile among workers exposed to and some crocidolite. and found a high anthophyllite. For long cancer, OSHA mesothelioma risk. Robinson et aL (Ex. views the evidence for differentials in 84-082) observed 4% of deaths from risk by fiber type as inconclusive and mesothelioma in a cohort where the inconsistent Some studies have found factory used the following percentages that ehrysotile workers have roughly the of asbestos fiber types: ehrysotile--99%. same or higher risk of lung cancer amosite--0.9% and crocidolite--0.07%. Among crocidolite workers. compared to workers exposed to amphibole fibers, while other studies McDonald and McDonald (Ex, 84-154) have found that ehrysotile workers had observed 16% of jdeaifcsfram - - somewhat lower relative risks of lung mesothelioma, (ones et al. observed 10% cancer. As discussed by CHAP with of deaths from mesothelioma (Ex. 84- regard to their comparison of lung 138], McDonald and Fry (Ex. 84-064) cancer risk per unit of cumulative observed 1.26% of deaths from exposure, consistent patterns of higher mesothelioma, and Hobbs et aL (Ex. 84- lung cancer risk among workers exposed 132) observed 9% of deaths from toxrocidolite and amosite did not mesothelioma. emerge. With some exceptions, The Chronic Hazard Advisory Panel comparisons of lung cancer risk among on Asbestos (CHAP) stated that occupational cohorts exposed to peritoneal mesothelioma "appears to be different fibers generally have failed to most common in workers expesed to control.for the following variables: Fiber amosite. less often to crocidolite. and concentrations, age distribution, and rarely or never to ehrysotile" (Ex. 84- length of observation. These variables 256. p. 127). However. CHAP noted the are sufficiently important determinants large variation in peritoneal mesothelioma mortality among of asbestos-induced lung cancer risk such that OSHA believes these crocidolite workers, the lack of risk data particular cross-cohort comparisons expressed in terms of unit exposure, and provided inconclusive evidence the frequent misdiagnosis of peritoneal concerning differences in carcinogenic mesothelioma. Consequently. CHAP potency of the various fibers. Other suggested that factors other than fiber important variables that may confound type, such as fiber dimension, may be cross-cohort comparisons are fiber sise important for induction of peritoneal distributions, hygienic measures mesothelioma. For pleural instituted by the companies, and mesothelioma, the differences in respirator use. mortality among cohorts were not as The evidence for a differential in mesothelioma risk, particularly peritoneal mesothelioma, by fiber type is more consistent than for lung cancer. Nonetheless, this evidence appears not to be conclusi e because of failure by most investigators to control for important variables when comparing different occupational cohorts. Mesothelioma risk ia affected by the same variables that determine the degree of lung cancer risk. Furthermore, mesothelioma has a longer latency period than asbestos-induced lung cancer and rises exponentially with increasing time since initial exposure, such that length of observation is a variable that must be controlled for in cross-cohort comparisons. Chrysotile is a lung carcinogen for humans. Even if chrysotile could be definitively shown to induce fewer mesotheliomas than the other forms of asbestos, OSHA believes that epidemiologic studies indicate that chrysotile poses a significant health risk to humans by inducing lung cancer and asbestosis. OSHA has concluded that fiber type is not an important determinant of lung cancer mortality arising from asbestos exposure. Because lung cancer is the major cause of excess mortality among asbestos workers, OSHA does not deem it appropriate to permit higher levels of exposure to chrysotile than to other asbestos fiber types on the basis of the possibility that chrysotile may induce fewer mesotheliomas than the amphibotes (amosite. crocidolite). 3. Experimental Data. OSHA believes that numerous studies of laboratory animals exposed to asbestos have found that the dimensions of fibers rather than types of fibers are major determinants of risk for lung disease and cancer (Harington. Ex. 64-131; Pott. Ex. 64-173: Stanton. Ex. 84-93, 84-195; Wagner et al.. Ex. 84-198: Wright and Kuschner, Ex. 84-210). Aspect ratios (i.e. length to width ratios) of fibers have also been suggested as factors influencing carcinogenicity (Bertrand and Pezerat, Ex. 84-114). Because fiber size distribution appears to affect diseasecausing potency, the source of the asbestos used for laboratory experiments and the methods used to produce the asbestos "clouds" should be considered when interpreting the results of laboratory experiments. Chrysotile, amoeite, end croddollta have all induced cancer in animals upon administration by inhalation. Injection and implantation (NIOSH. Ex. 84-338: NIOSH/OSHA, Ex. 84-320: Wagner et al.. Ex. 84-205; Davis et al.. Ex. 84-120). Davis et al. found that UlCC chrysotile ABS-008442 LAM 025267 concentration, age distribution, and length of follow-up need to be controlled for when comparing lung cancer risk of different cohorts. Berry and Newhousa (Ex. 84-021) and Newhouse, Berry, and Skidmore (Ex. 841H:1| studies the mortality of a cohort of workers manufacturing friction - materials. Exposure was to chrysotile exclusively, except during 1929-1933 and 1939-1944 when a small percentage of the workforce was exposed to crocidolite. Asbestos exposures were low in the plant; only 6% of the male workers accumulated as much as 100 f/ cc-years. The only cause of death in excess was mesothelioma. To examine etiulogic factors for mesothefioma. a matched case-control study waa conducted for 10 mesothelioma cases. For each case. 4 controls were selected from the plant Controls were matched on year of hire at the factory, sex, age. and duration of employment: Definite crocidolite exposure was discovered in the occupational histories for 8 of the 10 mesothelioma cases and for 3 of the 40 controls. About 90% of the cases and 25% of the controls had been exposed to levels of chrysotile exceeding 8 f/cc. When Berry and Newhouse limited their comparison to those cases and matched controls who had been exposed to over 5 f/cc of chrysotile, a statistically significant greater number of caaes had an occupational history of exposure to crocidolite as well. Matching the controls for the mesothelioma cases on year of bine controlled for the important factor of length of follow-up. OSHA considers that both crocidolite and chrysotile are implicated in the causation of mesothelioma by this study: chrysotile because 90% of the cases received high chrysotile exposures compared to 25% of the controls, and crocidolite becaase 80% of the cases compared to 7.5% of the controls had been exposed to crocidolite. The Chronic Hazard Advisory Panel on Asbestos (CHAP) compared lung cancer risk per unit of cumulative exposure (also known aa Kf the lung cancer potency factor) among the cohorts exposed to different fiber types. Cl <AP reported that studies of chrysotile workers yielded both low and high values of KL, as well as studies of crocidolite workers and amosite workers. Hence, CHAP considered that evidence for a differential in lung cancer risk was mixed and incondasive (Ex. 84-256). Pathology Studes: Wagner. Berry, and Pooley (Ex. 84-202) examined the asbestos contents at tissue samples of lungs of asbestos textile workers who died from mesothelioma and other causes. Wagner et ai. did not find either higher chrysotile or higher crocidolite contents in the lungs of the mesothelioma cases compared to the lungs of asbestos workers who died from other causes. McDonald. McDonald, and Pooley (Ex. 84-175) conducted a study of the mineral fiber contents of bng tissues for 99 auiopsied mesothelioma cases and 99 autopsied controls. Controls matched for sex and age were selected far each mesothelioma case. All controls had been diagnosed as having pulmonary metastases from a primary cancer other than lung cancer. Lung tisane specimens were examined for mineral Fiber content using an electron miscrosoope and X-ray energy-dispersive analyzer. Chrysotile. amosite, crocidolite, anthophyflite. and tremolite were detected in the lung specimens of the cases and controls. For amosite and crocidolite, about 3 times as many mesothelioma cases as controls had more than 1 nriOioa fibers per gram of dried lung tissue. For anthophylUte. twice as many cases as controls had more than 1 million f/gram. For chrysotile and tremolite. Utile difference in fiber contents was observed between cases and controls. McDonald. McDonald, and Pooley stated: "The equal quantities of chrysotile fibers found in lung tissue from cases and controls rail to support any association between this mineral type of asbestos and mesothefial tumors. However, the lung chrysotile fibre content at death must be interpreted cautiously, as these fibres probably disappear in the oourse of time" (Ex. 84175. p. 420). There ere several points to consider concerning McDonald et eL's study. First, if the fibers responsible for inducing mesotheliOBa were also responsible for inducing cancers et sites other than the lung and mesoihelium, then a true effect of e fiber oa mesothelioma genesis amid foil to be detected because of the stipulation that aK controls had to have pulmonary metastases. Gaetromtestiaal cancers sometimes metastasize to the lung, and some studlea bave observed excesses of gastrointestinal cancer among asbestos ' workers. Second, since chrysotile constitutes up to 90% of commercially used asbestos and sines asbestos exposure is ubiquitous, one would expect controls to have chrysotile fibers in their lungs. The widespread exposure to chrysotile could obscure a true relationship between chrysotile and mesothelioma. McDonald et aL suggested that chrysotile may be cleared faster iram the lungs than other asbestos fibers, so that chrysotile retention in the hangs may not be the bestaeay to estimate chrysotile exposure. Rowlands. Gibbs, and McDonald (Ex. 84-478) suggested a tendency for chrysotile to be raaeved from the longs becaeie of their observation that tbs quantity of tremolite in the lungs of Quebec miners and millers was similar to that of chrysotile. despite the miners' much 9-eater exposure to chrysotile. In general, OSHA believes dwt the results of autopsy studies of fiber content of Imtgs cannot serve as direct nearness of risk differentials by fiber type. Mesothelioma Fmdkm: As mentioned in the introduction to this section. Hie primary basia for the hypothesis that crocidolite is nan carcinogenic titan other forms of asbestos it tha oboereed variation in meaothelwma mortality among cohorts exposed to different fiber types. A summary of the percentage of ' deaths from mesetbefioem among cohorts exposed to different fiber types follows. Meurman et al. (Ex. 84-256), who observed excess lung cancer mortality among anthophylfite miners, did not find any deaths from mesothelioma in their cohort In Seidman et al.'s cohort of amosite insulation production workers, a total of 14 mesothelioma! were identified by death certificate or pathological reports (Ex. 84-087). These 14 deaths from mesothelioma constituted 2.7% of total daafhs in this cohort Cohorts of chrysotile markers had relatively low percentages -of deaths from mesothelioeMk: ABS-008443 LAM 025268 51116 Federal Register / Vol. 48, No. 215 / Friday. November 4, 1983 / Rules and Regulations produced more fibrosis in rats via inhalation than did UICC amosite. Factory samples of chrysotile, however, produced similar degree of fibrosis via inhalation compared to factory samples of amosite. Wagner et al. conducted inhalation Htudics of 5 UICC asbestos samples (2 chrysotile. 3 amphibole] in C/D Wistar nts. and found similar degrees of pulmonary fibrosis and lung tumor incidences for all exposed groups (Ex. 84-098). [The Union Internationale Centre Cancer (UICC) was the source fur the asbestos fiber samples used in the study.) Animals exposed to tha-2 chrysotile samples retained much less dust in their lungs at the end pf the exposure periods than the 3 otherexposed groups. Concerning this study, the NIOSH/OSHA Asbestos Work Group commented: "in terms of degree of response related to the quantity of dust deposited and retained in the lungs of rats, chrysotile appears to be much more fibrogenic and carcinogenic than the amphiboles" (Ex. 84-320, p. 15). Of interest were the mesotheliomas occurring after only one day of exposure to amosite and crocidolite. Intrapleural administration of various forms of asbestos produced the following incidence of mesotheliomas: crocidolite--61%. amosite--36%, anthophyllite--34%, Canadian chrysotile--30%, and Rhodesian chrysotile--19% (Wagner et al. Ex. 84197). In contrast. Stanton and Wrench (Ex. 84-338) did not observe differences in mesothelioma incidence by fiber type with intrapleural implantation. Chrysotile may act as a co-carcinogen in addition to acting as a primary carcinogen. Kung-Vosamae and Vinkmann (Ex. 84-143) observed a strong interaction between chrysotile administered intratracheally and Nnitrosodiethylamine administered orally with regard to production of lung tumors in hamsters. N-nitrosodiethylamine has been demonstrated to be carcinogenic to many different species by many different routes of administration. In 1977, Stanton, Layard, Tegeris, Miller. May. and Kent published the results of tests of 17 fibrous glasses of varying types and dimensions (Ex. 84- 093). These Fibrous glasses were implanted in the pleurae of female Osbome-Mendel rats, and surviving animals were sacrificed after 2-years. Statistical analyses Indicated that the Fibers less than or equal to 1.5 micrometers (jim) in diameter and longer than 8 pm produced the highest incidence of pleural sarcomas. Fibers less than 8 pm in length appeared to be inactivated by phagocytosis. (Phagocytosis is the engulfing of foreign particles by phagocytes, which are cells that usually digest these particles in order to protect the body from them.) Stanton et al. stated: "Since neoplastic response to a. variety of types of durable fibers, particularly asbestos Fibers, was similar, our experiments reinforce the idea that the carcinogenicity of fibers depends on dimension and durability rather than physicochemical properties and emphasize that all respirable fibers should be viewed with caution'* (Ex.' 84093. p. 587). OSHA believes that Stanton et al.'s study has important implications for the question of fiber type differentials in risk. The study suggests that any observed differences in risk by fiber type may be due to differences in the fiber size distribution within the workplaces being compared rather than inherent chemical properties of the particular fibers. Stanton et al.'s study suggests that fibrous materials besides asbestos can produce a carcinogenic response in both the pleura and peritoneum, provided that these materials possess carcinogenic dimensions. Among the fibrous materials demonstrated by Stanton et al. to produce malignant tumors following implantation are: All forms of asbestos, including amosite, anthophyllite. chrysotile. crocidolite, and tremolite; borosilicate; glass; aluminum silicate glass; mineral wool; ahimiiMun' oxide; potassium titanate; silicotreerblde; sodium aluminum carbonate; wollastonite; and attapulgite. It should be noted that Stanton et al.'s study did not rule out carcinogenicity of fibers outside of the dimensions shown to be carcinogenic. Stantoh et al. (1981) continued their implantation experiments, greatly expanding the number of durable minerals, including asbestos, tested via implantation into the pleurae of female ' Osbome-Mendel rats (Ex 84-195). Stanton et al. observed that the most carcinogenic fibers were those 0.25 pin or less in dianieter and greater thart 8 pm in length; however, high correlations with carcinogenicity were also observed for fibers 1.5 pm or less in diameter and longer than 4 pm. Wright and Kuschner (Ex. 84-210) studied the production of fibrosis by intratracheal injection of asbestos fibers in guinea pigs. They observed fibrosis only from asbestos fibers longer than 10 pm. Cibbs and Hwang (Ex 84-128) pointed out that industrial processes using asbestos may reduce the percentage of shorter fibers in the air. NIOSH (Platek, Groth, FlnnelL Stoll, and Ulrich) conducted a study of the chronic effects of inhalation of short asbestos fibers, with short fibers defined as those less than 5 micrometers (pm) in length (Ex. 84-230). Both rats and monkeys were exposed to a chrysotile asbestos aerosol for 18 months. The surviving rats were observed for 8 months following the cessation of exposure. Neither pulmonary fibrosis nor tumors were increased in the exposed rats compared to the control ruts. Monkeys are being maintained for an indefinite period following exposure to determine the chronic effects of' exposure. The NIOSH investigators encountered some difficulties in trying tq generate short fibers. The ball milling method was used to generate Fibers, resulting in asbestos balls that could not meet the desired 3:1 aspect ratio. Some problems also occurred with regard to the methods of counting asbestos fibers and determining aerodynamic diameters (Ex 84-230). Consequently, OSHA, considers that Platek et al.'s study provides suggestive but not conclusive evidence that short asbestos fibers do not induce pulmonary fibrosis or tpmors. Some chemists have posited that asbestos fibers have biochemically active sites on their surfaces that'ean be modified so as to reduce the hazardous potential of asbestos fibers (Ex 84-333). The electrical charge of chemical groups on the surface of asbestos fibers has been hypothesized to influence toxicity of fibers. For example, one chemical process removes the magnesium hydroxide groups on the surface of chrysotile and replaces them with silanol groups, producing a form of asbestos known as silanized asbestos. In vitro tests (tests conducted on cells in test-tube simulations of Bering systems) 1 have been conducted for normal asbestos fibers and chemically-treated asbestos fibers. Decreased toxicity in chemically treated asbestos fibers compared to normal asbestos Fibers has been reported (Ex 84-333). OSHA considers that the hypothesis that biochemically active sites on the surface of asbestos fibers determine the degreeof carcinogenicity of the fibers is not well supported at this time. This is because many in vivo studies, especially those conducted by Stanton et al.. have found that fiber dimensions rather than chemical properties appeared to be the primary determinant of fiber carcinogenicity. Stanton et al. found that a variety of non-asbestos fibers could induce cancer, if they were milled to specific dimensions. The in vitro studies of chemically treated asbestos are not as determinative of health risk as the In vivo studies of fiber dimension and hence do not refute the findings of Stanton et al. and other investigators ABS-008444 LAM 025269 Federal Register / Vol. 48. No. 215 / Friday, November 4. 1983 / Rules and Regulations 51117 concerning the overriding importance of their diameter. Based on experimental fiber dimension. Therefore. OSHA will findings implicating long thin fibers as continue to include chemically-treated being the most carcinogenic fibers, these asbestos longer than 5 micrometers with subsequent processes may have a a 3:1 or greater aspect ratio in the scope greater carcinogenic hazard compared of the asbestos standard, on the basis to the initial mining aod milling that chemically-treated asbestos poses a operations (Nicholson. 1981. Ex. 84-071). health risk similar to that of untreated OSHA examined the question of asbestos. . .. differentials in carcinogenic risk among In summary, OSHA regards manufacturing processes which fall experimental studies of animals as under OSHA's jurisdiction. As is the suggesting that all asbestos Tiber types case for studies of fiber type, cross possess similar carcinogenic potency. cohort comparisons of processes should Animal studies also suggest that Tiber control for the following variables: fiber dimensions rather than chemical concentrations, duration of foliow-up, properties or fiber type may be the duration of exposure, and age strongest determinant of distribution. To account for the variable carcinogenicity, in OSHA's view. of fiber concentration, OSHA compared Extrapolation of the results of the potency factors for lung cancer (KJ experimental studies in animals to the of different manufacturing studies human experience necessitates an because potency factors are based on assumption that the fiber dimensions risk per unit dose. As discussed in - used for experimental studies resemble Section V,,Quantitative Risk Analysis. Tiber dimensions in workplaces. This the potency factors for lung cancer were assumption might not be valid in all not consistently higher for any instances. particular manufacturing process. 4. Other Factors. The various types of Because potency factors for asbestos fibers may differ in the mesothelioma (K,,J could be calculated percentage of fibers available for for only 4 studies. OSHA could not inhalation and the length of time that compare Kj/s for different dust clouds remain airborne [Rowe, Ex. manufacturing processes. In summary, 84-085). Crocidolite might stay airborne there does not appear to be a consistent for a longer period of time and have a pattern of differential lung cancer risk greater number of respirable fibers by manufacturing process. Therefore. compared to amosite. which in turn may OSHA deems It appropriate to continue : .exceed anthophyilite In regard to these to apply a single PEL to all segments of two characteristics (Muir, Ex. 84-350). industry covered by the Agency. The preparation method for chrysotile Relatively thick respirable fibers tend will determine these two characteristics to lodge in the lung while thinner fibers for a laboratory sample of chrysotile or can travel to the periphery of th{ lung, for a particular workplace with and penetrate and then lodge in the chrysotile exposure (Rowe. Ex. 84-085). pleura. These fine fibers lodged in the Because thicker fibers might dfbp to the pleura might be the inducers of ground faster and crocidolite Is mesothelioma (Nicholson. 1981. Ex. 84- generally finer than the other fibers, 071). Bignon et al. (Ex. 84-105) shirked crocidolite use might lead to dustier the lungs and pleurae of shipyard work environments and hence a greater workers, and found larger fibers, health risk, even if crocidolite is not especially amphiboles. in the lungs. In more dangerous on a fiber-for-fiber the pleurae, fine end small fibers, basis (Muir. Ex. 84-350). However, most usually chrysotile. were found. respirable fibers, which are the fibers In summary. OSHA recognizes that likely to cause disease, will tend to the fiber types may differ with regard to remain airborne Tor long periods of time. their tendencies to break into fine fibers. Another point to consider is whether, Crocidotite appears to divide into finer for different fiber types, the r&tio fibers more readily than other asbestos between fibers visible under a fiber types, which may render it more microscope and ultra-fine hazardous. Manufacturing processes submicroscopic fibers varies. This could increase the carcinogenicity of would confound comparisons between asbestos by generating fibers that an cohorts apparently exposed to the same thinner in diameter relative to mining levels of different visible fibers under and milling processes. While these the light microscope, if indeed fibers not factors are interesting and contribute to counted by the present optical the formation of hypotheses regarding microscopy method pose a hazard to the mechanisms-of asbestos health (Muir. Ex. 84-380). .. . carcinogenesis. OSHA believes that OSHA believes that thicker fibers am ' these factors do not provide a definitive found in mining and milling and that ' answer to the question of possible subsequent ounufactypng processes differentials in risk by fiber type. OSHA tend to break fibers apart and reduce views the data from epidemiologic and experimental studies, which have observed excess cancer risk from exposure to all asbestos fiber types, us being the most important with regard to the question of differential health risk by fiber types. 5. Tremolite and Anthophyilite. Some but not aH commercial talc deposits contain serpentine asbestos and fibrous amphibole asbestos. Including chrysotile, tremolite. and anthophyilite (Dement and Zumwelde, Ex. 84-039). Kleinfeld et aL (1987. Ex 84-181) studied the proportionate mortality of 220 talc miners and millers in New York State who had been exposed to asbestos contained in talc. All men were employed during 1940 and had 15 or more years of exposure during 1940-. 1965. Of 91 deaths in the study cohort. 10 (11%) were from respiratory system cancer end 28 (31%) were from pneumoconiosis or complications of pneumoconiosis. Kleinfeld et al. calculated that only 2.9 (3.2%) deaths from respiratory system cancer would have been expected, resulting in a greater than 3-fold risk of respiratory cancer. A follow-up study by Kleinfeld et al. (1974. Ex. 84-141) included 280 workers exposed to asbestos in talc. Of the 108 deaths observed, 13 (12%) were from respiratorycancer compared to 4 (3.7%) expected. The high proportion of deaths from pneumoconiosis may have reduced the number of deaths from lung cancer. Kiviluoto et aL (Ex. 84-181) and Nurrainen et al. (Ex. 84-181) observed pneumoconiosis and excess lung cancer mortality among Finnish worker exposed to talcs containing fibrous anthophyilite and fibrous tremolite. No deaths from mesothelioma were reported by these investigators. Yazicioglu. llcayto. Balci, Sayli. and Yorulmaz (Ex. 84-211) reported a high prevalance of pleural calcification and thickening and a high mortality rate from mesothelioma among residents of Cermik. a town in Turkey in which there are many deposits of asbestiform minerals which are used as construction materials. The construction materials have been shown to contain fibrous tremolite. antigorite. lizardite, chlorite, and talc. Yazicioglu et al. also reported an excess of lung cancer in this population. Brown. Dement, and Wagoner of NIOSH (Ex. 84-025) conducted a historical prospective study of talc miners and millers of one company in New York State reported by die company to be mining talcs not containing asbestos minerals. MOSH (Dement and Zumwaide). however, had reported asbestos exposure In the talc ABS-008445 LAM 025270 51118- Federal Register / Vol. 48, No. 215 / Friday. November 4, 1988"/ Rules and Regulations . mine and mill. NIOSH conducted an industrial hygiene survey of the study talc mine and a neighboring talc mine known to contain asbestos fibers. Silica exposures were found to be very low. well below NIOSH's recommended PEL for silica. Radon daughter measurements made by the Mine Enforcement Safety Administration (MESA) showed only nil to trace levels. Dement and Zumwalde found that exposure characteristics between the study mine and neighboring mine were substantially similar: "In fact, the airborne dust samples from the mine and mill studied by NIOSH and maintained by the company to be asbestos free were found to contain a higher proportion of positively identified asbestiform amphiboles largely due to a higher tremolite content Ail other fiber characteristics, such a[s| median length, diameter, aspect ratio, and proportion less than 5 9m in length, were not statistically different at the 0.3 level" (Ex. 84-181. p. 10). The NIOSH study cohort (Brown et al.) consisted of all white male* employed sometime duritig 1947-1959. Vital status of the 396 cohort members was determined as of June 30.1975. Comparision was made to age, calendar period, and cause-specific mortality rates of U.S. white males. Significant increases in lung cancer mortality (9 observed deaths Versus 3.3 expected) and non-malignant respiratory disease mortality (8 observed versus 2.9 expected) were observed. One death from mesothelioma occurred. Since the individual who died from mesothelioma had previously worked In the construction industry, his death could not be definitely ascribed to his exposure to tremolite or anthophyllite. Of the 10 individuals who died from respiratory system cancer, 3 had previously worked for other New York State talc companies. NIOSH investigators also addressed the potential confounding effects of cigarette smoking. In their opinion, a cohort of heavy smokers would have no more than a 40% increase in lung cancer risk in relation to bU U.S. white males. Because they observed a greater increase in lung cancer risk, almost a 3fold risk, they judged that cigarette smoking was unlikely to account for the observed excess lung cancer risk among these talc miners and millers exposed' to asbestos. The cross-sectional morbidity survey conducted by NIOSH in 1975 found a 46% prevalence of smoking, a prevalence similar to that of U.S. males. Brown. Dement and Wagoner stated that "exposures to asbestiform tremolite and anthophyllite stand out as the prime suspected etiologic factors'* associated with the observed excess risks of lung cancer and respiratory disease. They workplace exposures of the talc workers concluded that "exposures to talcs from rather than other risk factors for lung the Couvemeur mining area are - disease is greatly increased. associated with an increased risk of Comparisons with die coal and potash bronchogenic cancer and non-malignHnt miners were stratified by age, height, diseases of thetespiratory system" (Ex. smoking status, and duration of 84-181). employment in mining. As measured by optical microscopy, Compared to coal miners and potash '* average air concentrations of fibers miners, the talc workers with no greater than 5 um in length ranged from previous occupational exposure within 1.7 f/cc to 9.8 f/ec as an 8-hour time- other talc mines and thills had weighted average for 6 different job statistically significant increases in titles in the mine. In the mill, average air pleural thickening and a' higher concentrations for such fibers ranged prevalence of pleural calcification. from 1.5 f/cc to 8.4 f/cc as an 8- hour When all talc workers were combined time-weighted average fos 18 different without regard to previous occupational job titles. (Ex. 84-181, pp. 7-10). ' exposure to talc, increased prevalences In addition to the excess mortality of cough, phlegm production, dyspnea, from lung cancer and non-malignant pleural thickening, pleural calcificatioa respiratory disease observed by and irregular opacities on.the X-rays Kleinfeld et al. and Brown et al., were observed in talc workers numerous studies of talc miners and compared to one or both mining control millers exposed to asbestos contained in groups,. tales have established that these In addition, talc workers had workers have a high prevalence of significantly decreased pulmonary pleural thickenings pleural calcification, function (FEV1 and FVC). Decreased decrements in pulmonary function, and lung function was associated with fibrosing lung disease (Dreessen. 1933: increased cumulative exposures and Dreessen and Delia Valle, 1935: Siegel et lengths of exposure. al., 1942; Scheperaand Dufkan. 1965; Talc workers had a similar prevalence Messite et al., 1950; Kleinfeld et aL. 1963, of respiratory symptoms yvhen '' ,vV,' 1964,1984.1084,1985.1985.1973; Meurman et al., 1974; Kiviluoto et al1964; Ahlman et al-1972; Porro et aL. 1942; Ex. 84-181). Many of these Studies were conducted in the same geographic compared to chrysotiW asbestos workers and a much,higher prevalence . of symptoms when compared to synthetic wool textile workers. In contrast pleural thickening was four area as the studies by Brown et aL and Kleinfeld et al. A cross-sectional morbidity study of the same company whose mortality experience was studied by Brown and colleagues was performed by NIOSH investigators (Gamble, Fellner, and DeMeo, Ex. 84-181). As discussed above, NIOSH considered that this times as common in talc workers compared to the chrysolite workers. Smoking was not found to be associated with the observed radiographic changes in talc workers. Regarding their studies of morbidity . and mortality of workers exposed to talc containing asbestos, NIOSH concluded:. company's workforce was exposed to asbestos contained within talc. NIOSH observed markers of asbestos exposure in the hings of these workers in addition to respiratory symptoms and lung function decrements. Of 156 male miners "A thorough review of the aveilable literature demorutrstedthst finding* of the >* present studies are In agreement with those- - of other studies of occupational groups exposed to (he same or similar minerals or _' -'mineral mixtures. This is especially true for. occupational exposures4o anthophyllite and millers. 121 participated in the asbestos. Theee findings make It Imperative survey. Respiratory questionnaires, that workers from the mine and mill studied, chest X-rays, and spirometric testing herein, be routinely observed usingmedical were administered to participating workers. Comparison of respiratory morbidity was made to 9347 coal miners, 1097 potash miners, chiysotile asbestos workers and synthetic wool textile workers. OSHA considers that one of the major strengths of Gamble et aL's cross-sectional morbidity study was the surveillance criteria established In the OSHA and MSHA asbestos standard. Fsrthermora. all provisions of these standards should be * followed during the production and subsequent use ofthesetalcs" (Ex. M-tSI. p, 33). Still* and Tabenhaw (Ex 84-196) of ~ Taberohaw Occupational Medicine choice of comparison populations. Because talc, coal, and potash miners are likely to be similar in many non- Associates studied all male workers employed sometime during 1946-1977 at the talc mine and miff studied by occupational factors that may affect NIOSH. A total of 708 men were eligible respiratory morbidity, the likelihood of for the study, and vital status as of observed differences in respiratory December 12,1978 was ascertained for morbidity being due to spedfie ' 872 of the men. Of the 708 men, 53 ABS-008446 LAM 025271 Federal-Register / Vbl. 48, No. 215 / Friday, November 4, 1983 / Rules and Regulations 51119 workers were excluded because of lack of information on specific variables, leaving a cohort of 055 for analysis. Comparison was made to age. calendar period, and cause-specific mortality rates of U.S. white mates. For the overall cohort, excesses in mortality that were not statistically significant were observed for lung cancer and non-malignant respiratory disease. Sflila and Tabershaw judged that the nan-significant excess of lung cancer was "consistent with a smoking effect" rather than an effect from occupational exposure (Ex. 84-190. p. 481). . Stille and Tabershaw also separately analyzed the mortality of the talc workers with a history of any work experience before employment at the study mine and mill. This prior work. experience included all previous jobs, not only jobs at other talc mines and mills. For this subcohort of 540 white males, mortality from the following causes of death was significantly elevated: all cancers (SMR 192), liver cancer (SMR-1013), respiratory system cancer (SMR = 228), lymphopoietic cancer (SMR--374), and non-malignant respiratory diseases (SMR=307). Stille and Tabershaw stated: "Since the cancers and lung diseases typically have long latencies, the possibility exists that exposures price to work at the TMX study mine and mill were responsible for at least some of these diseases" (Ex. 84-196. p. 482). When the mortality of the pubcohort with no work experience prior to employment at the study mine end mill was analyzed, no causes of death were in excess. In feet, the SMR for all causes was significantly decreased (SMR=50). As would be expected, this cohort was generally younger tha'n the cohort which had previous work experience. Stille and Tabershaw characterized these findings as follows: (1) "Workers with `exclusive' TMX [study mine and millj employment seem to be at no considerable riak of having lung cancer develop" and (2) "exposures at TMX seem to be noncarcinogenic" (Ex. 84- 190, p. 483). Brown. Beaumont and Dement of NIOSH commented on Stille and Tabershaw's study (Ex. 84-218. p. 178; Ex. 84-231), listing several problems in the analysis by Stille and Tabershaw that could account for the different conclusions of the two sets of investigators. First Stille and Tabershaw failed to analyze mortality by length of followup latency Interval). Brown et tL thought such an analysis . was particularly important given that - recently hired workers were allowed to enter the cohort aa late as one year " before the study cut-off date for vital status determination. Second, Brown et al. commented that Stille and Tabershaw's division of the study group into subcohorts with and without work experience prior to their employment at the study facility resulted in "selection biases inherent in the definition of the subcohoris" (Ex. 84218, p. 179). The selection biases which tended to lessen the statistical power of the subcohort without previous work experience included the short length of follow-up, short durations of exposure, and small size of the subcohort. Brown et el. stated: "Any mortality, analysis based on such a small cohort with generally short latency is not likely tabs very informative" (Ex. 84-218. p. 179). Third. Brown et al. criticized Stille and Tabershaw's analyses concerning ' the duration of employment of the lung cancer cases. These analyse* examined the observed lung cancer deaths without calculating the expected number of deaths for each latency category and duration of employment category. In addition. Brown et eL pointed out that Stille and Tabershaw's conclusions had not adequately acknowledged that many study facility workers had been employed by neighboring talc companies with exposures similar to those of the study facility. Brown et al. concluded that Stille and Tabershaw's report did not adequately address the question of increased lung cancer risk horn working at the study facility. In addition to die points discussed above. NIOSH also commented that death certificates should have been coded according to the rules of the 1CDA in effect at the time of death and then converted to either the 7th or 8th revision of the ICDA for analysis, rather than coding the deaths directly to the 8th revision. NIOSH also had questions concerning the 53 workers eliminated from the cohort and how workers lost to follow up were treated in the analysis. Tabershaw and Thompson responded to the NI08H critique with the following comments (Ex. 64-228); First they cited mineral scientists and laboratories who disagresd with NtOSffs method of identifying asbestos in silicate mineral mixtures and who did not find significant asbestos concentrations in the taUrpracessed at the mine end mill. Second,they stated that talc ore at the study facility was not similar to that of other talamines of upstate New York. Third. Tabershaw arid Thompson reiterated that many of the lung cancer deaths occurred in short-term workers employed for lees then one year. In their opinion, this lessened the likelihood of the lung Cancers being attributable to occupational exposure at the study plant. Tabershaw and Thompson also pointed out that only one tuic miller developed lung cancer, despite the historically higher exposures of tuic millers. In conclusion. Tabershaw and Thompson stated that application of the asbestos standard to the Vanderbilt workforce was unwarranted baaed on current evidence. OSHA calculated a statistical power of 20% to detect a 50% increase fn risk of lung cancer in Stille and Tabershaw's subcohort of workers with no previous work experience. (See Ex. 84-338 for a description of how OSHA calculated statistical power.) OSHA considers that this very low statistical power supports N10SH*s criticism of the sensitivity of Stille and Tabershaw's study design. The statistical power would have been even lower if it had been calculated for the group of workers with at least 20 years of latency. Aa discussed earlier, cross-cohort comparison may be problematic if the age distributions, durations of exposure, and lengths of follow-up differ among the cohorts being compared. OSHA believes that thia appears to be the case with TOMA's comparison of the subcohoris with and without previous work experience. Several points are notable with regard to the question of the short-term workers who developed lung cancer. First, short durations of exposure may result in high cumulative doses depending upon the intensity of the exposure. Second, the phenomenon of short-term asbestos exposure and subsequent disease has been observed in other epidemiologic studies. Third, if the majority of person-years-at-risk were contributed by short-term workers, then finding most of the lung cancer deaths among these workers would not be unusual. Also, Tabershaw and Thompson did not discuss the expected number of lung cancer deaths when they mentioned the single lung cancer death among the study facility talc millers. In 1983. Consultants in Epidemiology and Occupational Health (CEOH) prepared another analysis of the mortality of Stille and Tabershaw's study cohort (Ex. 84-257). NIOSH reviewed CEOH's analysis and stated: "Although the findings of these analyses appear to support their [CEOH's) hypothesis that the laic is non-carcinogenic. the resulting statistical analyses are based on assumptions, small numbers, and short latency. Therefore, the CEOH conclusions are baaed on analyses that are inherently deficient In being able to ABS-008447 LAM 025272 51129 Federal Register / Vol. 48. No. 215 / Friday. November 4. F963 / Rules and Regulations detect a true risk in an exposed population" (Ex. 64-375. p. 3). Smith. Hubert. SobeL and Marquet (Ex. 64-194] studied health effects of laic containing asbestos in laboratory animals. Smith et al. administered intrapleural injections of 4 different tremolitic substances into hamsters, including fibrous tremolitic talc from New York state, tremolile prepared from tremolitic talc ore of the facility studied by NIOSH. tremolite prepared from western U.S. tremolitic talc ore. and asbestiform tremolite. Periods of observation ranged from 350 days to 600 days after injection. Tumors and pleural fibrosis were observed only in animals injected with asbestiform tremolite and tremolite prepared from western U.S. tremolitic talc ore. Smith et aL found that the samples of asbestifocm tremolite had greater proportions of long thin fibers than the sample of tremolite from the facility studied by NTOSH. The sample of fibrous tremolitic talc from New York state also contained long thin fibers and Smith et aL suggested that the negative results'from this particular sample were due to its ' much lower tremolite content relative to the carcinogenic samples. Smith et aL suggested that the sample oftremolite from the facility studied by NIOSH may have failed to induce tumors because of the shorter length of its fibers, despite its high Uemolite content. Commenting on their results and on the fact that the carcinogenic samples contained material other than tremolite. Smith et al. stated: * * * we cannot be sure that their activity it due wholly, or even in port to tremolite * * consideration must be given, not merely to the amount of tremohta, but also ta other (actore. such as the morphologic charactaristics at the mineral" (Ex. 84-184. p. 338). OSHA considers thot Interpretation o4 Smith et al.'s findings is Endted by the small number of animals studied, short survival times, and short periods et observation. In summary. OSHA concludes that NIOSH studies have shown that exposure to asbestos when it is present in talc appears to have resulted in excess mortality from lung cancer and non-mallgnant respiratory disease and excess risk of pleural thickening and lung function decrements. OSHA is cognizant of the fact thut talc mining .uul milling in the absence uf asbestos exposure can result in pneumoconiosis: Furthermore. NIOSH*s findings on mortality are consistent with those of Kleinfeld et al.. Kiviluoto et al, Nurminen et al., and Yazicioglu et aL MOSH's morbidity findings are consistent with those of numerous ether studies. OSHA believes that the epidemiologic study of the take company conducted by Stille and Tabenhaw of Tabershaw Occupational Medicine Associates has serious limitations such that it does not refute the hypothesis that talc containing asbestos peses a lung cancer hazard to exposed workers. Smith et al.'s experimental study observed positive findings of tumorigenicity for asbestiform talc and non-positive findings for (he talc from the facility studied by NIOSH: however. OSHA thinks that limitationsof Smith et al.'s study render its findings inconclusive for talc from the facility studied by NIOSH. In any event the positive human findings outweigh die non-positive animal findings. Talcs containing asbestos minerals, therefore, appear to pose a significant heaUl) risk' to exposed workers, and talc workers . exposed to asbestos should receive the protection afforded by the asbestos standard. OSHA notes that the broader issue as to which mineral fibers ere to .. be included in the definition of asbestos will be pert of the rulemaking for the permanent standard. This ETS acdon does not change die definition of asbestos as found in MKJ.1Q01. 6. Summary. The ovidence presently before the Ageaeg indicates that all asbestos fiber types, except far anthophyliite, cause fang cancer, mesothelioma, and asbestosis. Anthophyliite causes fang cancer and asbestosn. Mesothelioma has not been reported in anthophyliite workers. For fang cancer, OSHA notes that Inconsistent end inconclusive findings* with regard to differential risk by fiber type have been observed by different investigators. Some studies have observed the same or higher lung cancer risk in chrysolite workers compared to workers exposed to amphibole fibers; other studies observe lower risk of King cancer in chrysotiie workers. For example, high respiratory cancer risk was observed by Dement and colleagues among chrysotiie textile workers. As discussed by CHA<P. consistent patterns of higher lung cancer risk by fiber type did not emerge. With some exceptions, investigators generally failed to control for variables affecting lung cancer risk when comparing the mortality of different occupational cohorts. Among the important detetminants of lung cancer risk, which were not consistently controlled Cor. ere fiber concentration, lengthjofobservation, and age distribution. OSHA views evidence far differential mesothelioma risk by fiber type as suggestive yet net conclusive. Occupational cohortsexposed to caBcidolite and amoaiie appear ta have had a higher proportion of deaths from mesothelioma, particularly peritoneal mesothelioma, compensate occupational cohorts exposed to chrysotiie. Nonetheless, OSHA is unable to characterize this evidence ea conclusive because most investigators failed to control for fiber concentration, length of observation, and age distribution when making cresa-cohort comparisons of mortality from mesothelioma. Controlling for length of observation is particularly important because mesothelioma usually has a latency period of 30 years or more and the risk rises rximmiatfany with increasing time since initial exposure. Lung cancer is the major cease of asbestos-induced occupational mortality. Mesothelioma end asbestesis aIso era significantcauses of excess - mortality among asbestos workers. OSHA has concluded tiwi thefre'is a grave risk for lung cancer and asbestosis among workers exposed to all asbestos fiber types and that workers need to beprotected' from these*risks by this ETS, which sets thelowest PEL that is feasible at this time. A* a practical regulatory endpoint, therefore, it is inconsequential whether the mesoftefiomariskWgi ohtei fof certain * fibertyped 9um ipraher types. Experimental stwBfes of aritmah have found that chrysotiie fa as hast as potent a carcinogen as other asbestos . fiber types. One stu^y found that chrysotiie was more carcinogenic than other asbestoe fibergwhen administered by inhalation. Many experimental studies have reported that Sber 1: dimension is a very Important determinant of carcinogenicity in laboratory animals. Fibers longer than 4 micrometers (urn] and with diameter* of 1.5 urn or less have bean implicated as the most carcinogenic fibars. A variety of different natural and synthetic substances having fibers longer than 4 . urn with widths of 14S umor leas have induced cancers vlalmplantatipo lu laboratory animals, suggesting that the chemical properties of asbestos (such as fiber type] and non-asbestos fibers may not affect carcinogenicity but that fiber dimensions may determine carcinogenicity. Thus, in OSHA's vipw.< animal studies do not support the hypothesis that there are differences in carcinogenic potency among the various fiber types. Fiber types may differ with regard to thh fineness offibess used in industrial processes. Finer fibers may b more . carcinogenic, according to animal studies, and crocidollts appears to be divided more readily into fine fibers ABS-008448 Lam 025273 Federal Register / Vol. 48. No. 215 / Friday. November 4, 1983 / Rules and Regulations 51121 than other asbestos Tiber types. Also, manufacturing processes may break fibers apart to generate Tibers thinner in diameter than those found in the mining and milling processes. NIOSH scientists (Brown, Dement, and Wagoner) observed excessive mortality from lung cancer and non- malignant respiratory disease among workers in one plant exposed to talc containing asbestos. A morbidity survey of the same plant found increased risk of pleural thickening and lung function decrements. NIOSH concluded that Increased risk of lung cancer and non- malignant respiratory disease was apparent in workers exposed to talc / containing asbestos. Kleinfeld and . colleagues also observed excess lung cancer and respiratory disease mortality among workers exposed to talc containing asbestos. and numerous morbidity surveys have documented respiratory disease excesses among talc miners and millers exposed to asbestos. Stille and Tabershaw conducted a study which they interpreted as contradicting NIOSH*s findings regarding excessive mortality among talc workers exposed to asbestos. Stille and Tabershaw's study had serious methodologic limitations such that NIOSH's conclusions still appear tq be valid. While not all talc products contain asbestos, OSHA finds that when asbestos minerals are contained in talc products, there appears to be. a health risk to exposed workers. Practical administrative considerations also are pertinent to the question of making teguhfory distinctions by fiber type. Several types of asbestos may be present in the same . workplace, as is the case for insulation, creating considerable measurement difficulties. Determining asbestos typefs) within each workplace might complicate compliance and exposure monitoring. However, OSHA is aware that the United Kingdom has been able to overcome such complications. In summary. OSHA concludes that on the basis of the epidemiologic and experimental studies reviewed above, it is not appropriate at this time to make distinctions by fiber type for regulatory purposes. Accordingly, quantitative analyses of risk should base estimates of excess risk on asbestos fibers as a whole rather than on separate fiber types. Provisons of OSHA's standard for asbestos, including the permissible exposure limit, should also be the same for all commerically used asbestos fiber types. F. Conclusions Asbestos poses a grave danger to the health of exposed workers, having caused excess mortality and disability in epidemic proportions among some groups of exposed workers. Asbestos causes non-malignant respiratory disease, which can result in complete disability and death. Asbestos also causes malignant mesothelioma, which is usually rapidly fatal, and lung cancer, which is usually fatal. The causal relationship between asbestos exposure and disease has been established by a multitude of epidemiologic studies conducted throughout the world, including the United States, United Kingdom, Canada. Australia, Italy, and Finland. Furthermore, asbestos has been shown to cause excess disease in many different occupational environments, starting with mining and milting of asbestos and continuing to cause disease in various manfacturing processes such as asbestos textile production and asbestos cement production. Asbestos also has been shown to be extremely hazardous when used as a product by insulation workers and shipbuilding workers. Numerous scientific organizations and agencies have reviewed the health data for asbestos and have concluded that asbestos exposure causes lung cancer mesothelioma, and asbestoeis.. To determine the potential of a hazardous agent to cause disease at low exposures, OSHA has examined the results of studies of workers receiving low exposures, where such studies are available and have sufficient sensitivity to detect excess risk of disease. OSHA has also attempted to predict risk at low exposures from risk observed at high exposures by using dose-extrapolation models (see section V). Both of these methods of evaluating risk at low exposures are valid. Studies are not available of workers exposed solely to asbestos levels of 2 f/ cc which have sufficient cohort size and long enough follow-up periods to permit observation of increases in cancer and lung disease mortality. Studies are available of workers with low cumulative exposures to asbestos, however. Workers exposed 50 years to the OSHA PEL of 2 f/cc would have a cumulative exposure of 100. f/cc-years. Hence, studies of workefs with cumulative exposures close to or below 100 f/cc-years provide some evidence of risk from working lifetime exposures to OSHA's PEL. Studies finding adverse pulmonary effects or excessive mortality from lung disease from exposure close to or below 100 f/cc-years were conducted by Berry at aL. McDonald et al.. Dement et aL. Berry and Newhouse, and Finkelstein. Dement et aU Finkelstein, and Henderson and Enterline also observed excess mortality from lung cancer among workers receiving relatively low cumulative exposures. Weill et al. and McDonald et al. did not observe increased lung cancer mortality among workers receiving relatively low cumulative exposures. Despite some inconsistencies in findings regarding risk from low cumulative exposures. OSHA considers that the many well-conducted studies which observed substantially increased risk of morbidity and mortality among workers receiving low cumulative exposures are evidence of significant hazard at exposures allowed by the existing PEL Section V., Quantitative Risk Analysis, will discuss predicted risk from exposure to OSHA's PEL A number of investigators have suggested that different fiber types of asbestos have differences in their ability to induce mesothelioma. Specifically, croddolite has been hypothesized to be the most hazardous type of asbestos because of suggestive evidence that indicates that the mesothelioma risk may be higher for crocidolite exposure than for other asbestos types. All commercially used fiber types have induced lung cancer, mesothelioma, and asbestosis in exposed workers, except for anthophyllite. which has induced lung cancer and asbestosis. OSHA regards evidence for differentials by fiber type in lung cancer risk and asbestosis risk as being inconclusive and inconsistent. However, evidence for fiber type differentials in risk with regard to Induction of mesothelioma, particularly peritoneal mesothelioma, appears to be more consistent based on epidemiologic studies of occupational cohorts exposed to different fiber types. Because variables that affect the sensitivity of studies for detection of risk of mesothelioma have not been controlled for by most investigators comparing different occupational cohorts. OSHA views the evidence for fiber type differentials in mesothelioma risk as inconclusive. Among the variables that influence the appearance of mesothelioma are length of observation, duration of exposure, fiber concentration, and age distribution. It is possible that mesothelioma risk may be higher from exposure to finer fibers, and that operations using crocidolite may tend to break fibers apart to make them more carcinogenic. Studies in laboratory animals have not consistently observed differential risk by fiber type with regard to induction of pulmonary fibrosis and cancers; in fact, some studies in animals have found chrysotile to be the most carcinogenic among the various asbestos fibers tested. ABS-00B449 LAM 025274 51122 Federal Register / Vol. 48. No. 215 / Friday. November 4. 1983 / Rule* and Regulations . Lung cancer is the major cause of i sbestos-sssocia led occupational mortality. The very high risk of lung i unrur from exposure to all asbestos filmr types supports choosing the snmc I'Ll. Tor all fiber types. Several epidemiologic studies, including one performed by NIOSH. have found that workers exposed to talc containing asbestos have excess mortality from lung cancer and nonmaligant respiratory disease. Many other studies have observed excess morbidity from respiratory disease, chest X-ray abnormalities, and lung function decrements among such workers. One non-po6itive study conducted by Stille and Tabershaw of talc workers has methodologic limitations such that it does not refute studies finding excess mortality among workers exposed to talc containing asbestos. - A number of studies in laboratory animals have implicated long thin asbestos fibers in the etiology of cancer and lung fibrosis. Stanton et al. found that Fibers greater than 4 pm in length and 1.5 pre or less tn diameter were the most carcinogenic; however. OSHA believes that Stanton et al.'s studies do not exclude the possibility of carcinogenicity of fibers shorter or thicker than these dimensions. At least 12 studies have observed excess mortality from gastrointestinal cancer (stomach, esophagus, colon and rectum) among asbestos exposed workers. 7 of which found statistically significant excesses. Other cancer sites have also been observed to be increased by different investigators. OSHA believes that the positive human studies finding gastrointestinal cancer excesses outweigh the non-positive and equivocal unimaLstudies of ingested asbestos, as well as the non-positive hantan studies. Therefore. OSHA concludes that excess mortality from gastrointestinal cancer should be included in quantitative risk analyses for asbestos. Cigarette smoke and asbestos have a multiplicative (synergistic) effect with regard to production of hing cancer. Asbestosis also might be increased in cigarette smoking asbestos workers relative to non-smoking asbestos workers OSHA is not aware of any evidence fur an effect of cigarette smoking on induction of mesotheliomas or gastrointestinal malignancies among asbestos workers. Hammond et al. found that both smoking and non smoking asbestos insulation workers had relative risks for long cancer of about 5. V. Quantitative RMc Analysis A. Introduction As explained above, OSHA's determination that currently exposed workers face a grave risk of asbestosrelated disease is primarily based on the results of quantitative risk assessment performed by The Agency. The process of assessing the risk from asbestos exposure includes several steps. OSHA has critically evaluated the scientific evideace concerning the health risk from asbestos exposure. OSHA, as well as other scientific groups, believes that asbestos exposure causes lung disease, respiratory cancer, mesothelioma, and gastrointestinal cancer. OSHA has also examined (and discussed in the section entitled Epidemiologic Evidence on Risk from Exposures at Current the PEL. Section IV B.) evidence that indicates that excess disease risk has been observed at cumulative exposures at or below those permitted by the existing OSHA 8-hour permissible exposure limit In addition, OSHA has made risk estimates of excess mortality from long cancer. mebotheUoma, gastrointestinal cancer and incidence of asbestosis using mathematical models that describe the data observed in epidemiologic studies oenducted in varies* industrial populations. in many cases the elevated risks seen in worker populations reflect past exposures that were higher than those permitted today. OSHA's quantitative risk assessment entails using the directly observed Ksks from these past exposures to estimate risk at lower exposure levels. OSHA believes this is a scientifically appropriate and valid procedure. In some instances. OSHA estimated risks using studies which actually observed risks at or below cumulative exposures permitted by the current standard. The range of studies covers many'different work situations and exposure levels upon which to base the extrapolations. Where possible. CSHA has quantified the ranges of uncertainties in the estimates. These numerical estimates, as well a* those risks observed at low exposures, were evaluated to determine the significance of the risk and to determine whether the standard will lead to a substantial reduction in risk. The sections below provide a brief synopsis of OSHA's quantitative risk estimates derived from matketnatica! models. A more detailed description of OSHA's calculations for asbestoerelated cancers is in the public docket and is available open request (Ex. 84392). B. Dose-Response Modelfor Long Cuncvr Dose-response information for lung cancer bused on the epidemiological studies indicates that die risk of asbestos-induced lung cancer increases linearly with dose. Therefore, OSHA believes it is appropriate lease a mathematical model which assumes that the excess risk is linearly proportional to dose to estimate the excess mortafity which is associated with asbestos exposure. This excess risk is expressed as the number of additional King cancer deaths per 1000 workers exposed for a specific time period. Excess risk equals the cumulative dose multiplied by a potency coefficient (slope of the doseresponse curve) estimated from the studies. [Note: Total dose, also referred to as cumulative exposure oroumuletive dose, is a measure of the amount of asbestos inhaled; tt is the product of the duration of exposure (in years] and the intensity of exposure (which is workplace air concentration in mppcf or fibers/ce). Under this definition of exposure, a person exposed to airborne asbestos at 2 fibers/cc for 20 years (40 fibers/cc-years) has die same total dose as a person who is exposed to asbestos at 4 fibers/cc for 10 years (40 fibers/cc-. years/]. Evideace of the linear dose-response relationship for lung cancer Is found in several well-conducted epidemiologic studies that examined lung cancer mortality in relation to the cumulative asbestos exposure in the workplace (for example. Henderson and Enterline. 1979, Ex. 84-48; Liddell eiaL 1977. Ex. 84-59; and Dement et aL 1982, Ex. 84-35).in the three studies cited above, workplace asbestos air concentrationswere available from measmements made in the worksite studied. Although the studies differ in the magnitude of the risk found (discussed later m the section), all three doasonoteate a linear relationship over threatire range of observation. Other scientists and scientific groups who have attempted to estimate risk from asbestos exposarc harassed the linear model lor lung cancer {Crump, Ex. 85-22. British Advisory Gbmmittee on Asbestos. Ex. 84-218. Acheson and Gardner. Ex. 84-943, Seiikoff. Ex. 82-2, EPA. Ex. 84-180. CHAP. Ex. 84-258). The model is generally accepted and OSHA believes use of the linear model for predicting lung cancer due to asbestos exposure is reasonable and wellsupported. Relative risk is defined as the ratio of the mortality rate of exposed persons to the mortality rate of equivalent non- ABS-008450 LAM 025275 Federal Register / Vol. 48. No. 215 / Friday, November 4, 1983 / Rules and Regulations 51123 exposed persons. Relative risk is frequently approximated by the standardized mortality ratio (SMR). which is the observed number of deaths in the exposed population divided by the number of deaths that would be expected in the exposed population. The number of expected deaths is usually derived from the specific age. sex. and calendar year mortality rates in the comparison population. Based upon the considerations listed above and on the data, the relative risk model used by OSHA in assessing the risk of developing lung cancer from asbestos exposure is described by the following equation: rl *re l' * f * dt-io)J That is, (RL/RE)-l Kl x f x dt_)0 (Eq. 2) where |l^ Is the lung cancer mortality resulting from the asbestos exposure, R Is the expecteo mortality in the aDsence of exposure, f fs the intensity of exposure in fibers/cc, o istne duration of exposure In years, t is the time from the onset of the asbestos exposure in years (minus Mi years to allow for a minimum latent period) ano *L is the proportionality constant that is a measure .of the carcinogenic potency of the asbestos exposure (slope of the dose-response curveJ. (For a more detailed rationale for the choice of this model, see Ex. 84-392). This calculation was performed for eacn five-year age interval; tne overall Vang cancer risSTs-tnen computed as the sum of the risks in each of the five-year .intervals from age 25 to age 7(3. The cooespt ofdirectproportionality of the risk of lung cancer to the dose can be demonstrated with a simple example. Using Equation 2, if does is increased by a factor of three, sadi as increasing the duration of exposure from two years to six years, the-excess relative risk ((Ri7 ' Rt) -- 1) risk increases by a factor of three. Also, reducing dose by reducing air concenhattona.suah as from 4 fiber^ cc to 1 fiber/cc. reduces the excess relative risk by a factor of four, (ft should be noted that the estimates of risk given in Tables 10 and 11 are estimates of excess risk, not excess relative risk. Therefore, direct proportionality of dose and risk as described above may not be observed. Further explanation is given in Section ] C. Dose-Response Mode!for Mesothelioma OSHA believes that the risk of mesothelioma is best estimated by an absolute risk model. Absolute risk is calculated as (observed-deeIht/person- years-et risk). Use of SMR's or relative risk is not appropriate for mesothelioma because the expected number of deaths in a cohort would be dose to zero due to the rarity of the disease. In addition to using absolnte risk rather than relative risk, this model is different from that used for lung cancer because both duration of time since initial exposure and duration of exposure are determinative risk. The magnitude of the risk Increases linearly with intensity of exposure, whereas the risk increases exponentially with duratine of exposure and time from onset of exposure. The rationale for such a model for describing mesothelioma risk has been discussed by several authors (Armitage and Doll. 1988, Ex. 84-252; Pika 1986, Ex. 84-385). Such a model was utilized by Newhoiue end Berry (1938. Ex. 84-342) in predicting masothatiema mortality among a cohort ol factory workers In England. Limited data are also available from three studies on the dose-response relationship for mesothelioma (Seidman et si., 1979, Ex. 84-67; Hobbs, et al., 1980. Ex. 132, and Jones et al., 1980, Ex. 84138). The data indicate that mortality from mesothelioma begins to increase only after a minimum of ten years following the-initial exposure and begins to decrease after 45 years from onset of exposure (Selikoff et al,, 1979, Ex. 84-90 and Nicholson ekaU 1983. Ex. 84-251). The mortality from mesothelioma can be described by the following equations (Equations 3). ABS-008451 LAM 025276 51124 Federal Register / Vol. 48, No. 215 / Friday. November 4, 1983 f Rules and Regulations ARj, f * *- f * Lit - I0J3 - (t'-IO-a)3] it-10)3 arm 0 for t210i for IO*d>t2tlO for I0>t where AR^ is the morulity from mesothelioma, f is the intensity of exposure in ftbers/ce, d is the duration of exposure in years, t is time after first exposure in years, and Kn is the proportionality constant tnat is a measure of the mesothelioma carcinogenic potency (slope of a dose-response curve) (Ex. 84-392;. O. OSHA `s Quantitative Risk Assessment OSHA's critical evaluation of all relevant animal and epidemiological studies resulted in selection of eleven epidemiological studies for calculation of the Kl for lung cancer (Selikoff et at.. 1979, Ex. 84-90; Seidman et al., 1979, Ex. 84-87; Henderson and Enterline, 1979, Ex. 84-48: Weill et al.. 1979, Ex. 84-206; Finkelstein. 1983. Ex. 84-240; Peto. I960. Ex. 84-169; Dement et al., 1982, Ex. 8435; Berry and Newhouse, 1983. Ex. 84-21; Liddell et al., 1977, Ex. 84-65: Nicholson et al., 1979. Ex. 84-72; and Rubino et al., 1979, Ex. 84-86) and four for KM for mesothelioma (Selikoff et al.. 1979, Ex. 84-90; Seidman et al.. 1979. Ex. 84-87; Finkelstein et. al.. 1983, Ex. 84-240; and Peto, 1980. Ex. 84-169). In general, studies of human cohorts in the workplace should provide a better basis for quantitative risk populations at risk and the populations from which the risk estimates are derived. In determining the potency coefficients KL and K*. Equations 1 and 3 were used to define the dose-response relationship so that cancer mortality was estimated for various exposure levels and exposure durations. A number of well-oonducted and high quality epidemiologic studies were available that contained sufficient information on which to base a quantitative risk assessment Some of these studies did not contain exposure data, but could be coupled with exposure information from other sources in order to obtain an estimate of Kv and K,,. OSHA chose not to use animal studies lo predict quantitative estimates of risk from asbestos exposure due to the many high quality human studies that exist that were conducted in actual workplace situations. Rather OSHA has supplemented the human data with results from animal studies in the evaluation of the health information and in the determination of the significance of risk because OSHA believes that the animal studies provide valuable qualitative information on asbestosrelated disease. It is not clear in all instances whether laboratory animals have been exposed to fiber size distributions similar to those found in workplaces. In addition, asbestos appears to multiply the underlying lung cancer risk of smoking and nonsmoking workers; laboratory animals generally do not have any underlying risk for developing lung cancer. However, the animal studies do show that all commercial asbestos types can cause cancer and pulmonary fibrosis. Animal studies also indicate that longer, thinner fibers may have greater carcinogenic potency than short coarse fibers. The range of estimates of risk from the eleven epidemiologic studies is rather large. The differences in results among the studies can be explained in several ways. There appears to be actual differences in risk depending upon the nature of the asbestos exposure. One potential explanation is that workplaces differ with regard to fiber size distributions (longer finer fibers appear to have greater carcinogenic potential than the coarse fibers). The observed Ki values for studies of mining and milling operations, where airborne fibers are relatively coarse, are lower than the K|, values found in studies of textile operations where fibers are fine. Differences may also be explained by the variations in study design and other factors influencing the ability to detect excess risks. One of these is the limited knowledge of past fiber exposures of those populations whose mortality was later evaluated. Prior to 1970, few measurements were made In facilities using asbestos fibers. Further, those measurements that were done usually quantified all dust present in the workplace air and not |uit fibers. Current techniques, which involve use of membrane filters and phase contrast microscopy for the counting of fibers longer than five micrometers have been utilized in Great Britain and the United States only since 1984 (Ayer et al., 1985. Ex. 84-253;) and have been standardized in the United States only since 1972 (Leidel, 1979. Ex. 84-62) and even later in Great Britain. In any case, sampling has occurred only for few worksites and then only occasionally. OSHA has evaluated these differences and have dealt with their implications on a study by*study basis, as explained in the quantitative risk assessment (Ex. 84- 392). OSHA notes that despite these apparent limitations, taken as a whole, the asbestos studies contain data of unusually high quality which has enabled OSHA to make the risk estimates with a high degree of confidence. In addition, variability in work activities and in sampling circumstances add considerable uncertainty to knowledge of dose. Some of the epidemiologic studies, including those by Dement et al. (Ex. 8435). McDonald et aL (Ex. 84-48). and Henderson and Enterline (Ex. 84-48), have measured air concentrations at the exposure site and used job histories of the study population to estimate exposure. In these cases the doseresponse curve was calculated by estimating total asbestos exposure (in mppef-years or in fiber/cc-years) according to the time that an individual spent at a job with a measured exposure value. A conversion factor for converting from mppcf to f/cc was employed on a study by study basis depending upon the data available. Other epidemiological studies, for example those by Selikoff et al. (Ex. 8490). and Seidman et al. (Ex. 84-87). had neither job histories nor direct industrial hygiene measurements for the studied worker population. For these studies, exposure estimates were derived from industrial hygiene surveys of similar work operations and processes for which industrial hygiene data were available. The study by Seidman et al.. however, contained good information regarding duration of exposure (which can often be examined as a surrogate for dose in establishing the shape of the exposure-response relationship). As discussed in Section IV, OSHA has concluded that workers exposed to ABS-008452 LAM 025277 FedandRagbtos / V-ot. 48, No. 215 / Friday. November 4, 19g?/_Rules and Regulations 51325 asbestos an likely to be at en inenaaed or in some cases, etatistical risk of gastrointestinal cancer. Though uncertainties associated with small an excess of Cl cancer has not been 'numbers. Detailed derivation of each observed consistently inevery study of range of uncertainty is discussed in Ex. asbestos workers, aiU while the ratio of 84-392. gastrointestinal cancer to lung cancer The distinct nature of mining-milling varies considerably from study to study, data (and benco, the estimates of KL there appears to be sufficient evidence from these data) have been considered to roughly estimate the excess earlier. There ie some evidence that gastrointestinal cancer risk in asbestos- risks in the aebeatos mining-milling exposed papulations. In general, the risk operations are lower than other ranges from about 5 to 20* of the excess industrial operations doe to differences lung cancer risk. (A detailed table of the in fiber size. Thus, in determining the risk from gastrointestinal cancer best overall value for KL from the eleven observed in 27 studies is given in Ex. 84- studies, the data were examined both 392). In an attempt to quantify the risk of with and without the Ki calculated from gastrointestinal cancer, OSHA considers the studies of mining-milling processes. that a simple risk model in which the The range of individual values for Ki lung cancer excess is multiplied by 0.1 coven two orders of magnitude, from (10%) is appropriate for estimating the 0.0006 to 0.006. The arithmetic mean of risk from gastrointestinal cancer. The the eleven values of KL (unit risk per f-y/ estimates of dak tern gastrointestinal cc) ta 0.0201, and 0.0287 when the Kl a 'cancer are also given in Tables 10 and 11. from mining and milling are excluded. The geometric mean of the data is 0.097; Cancers at sites other than the lung. mesothelium, andgastraintestinal tract have been shown to be elevated in some when the estimates of KL from miningmilling operations are excluded, the geometric men of the Ki* is 0.0113. The asbestos exposure studies, including laryngeal, kidney, pharyngeal and buccal cavity cancer. To OSHA. it appears that the. excess risk for "other cancers" is about the same as for gastrointestinal cancers. OSHA recognizes many uncertainties in quantifying this risk, in view of the inconsistencies in findings among different epidemiologic studies. (Some studies have found excess risk from other cancers, while other studies have not.) The sites showing excess risk have also varied among studies. Therefore. OSHA has not made numerical ' estimates of risks for these other causes at this time. The data indicating gastrointestinal cancer excesses are stronger and more consistent than the data suggesting excess at these other cancer sites. OSHA does not feel compelled to quantify this risk at this time. The high quality and wellsupported estimates of excess risk of Kl's have a median of 0.0051 with the mining-milling processes and a median of 0.0138 when the mining-nulling processes are excluded. Considering the industrial processes other than mining and milling. OSHA . believes 0.01 to be a reasonable estimate of Kl. It is die geometric mean and median of the Ki'a derived from studies of asbestos manufacturing and insulation application processes. The geometric mean has the advantage of minimizing the influence of outlying values and a KL of 0.01 is approximately within one order of magnitude of all the estimates of KL. In sum. the KL of 0.01 is a best estimate which contains appropriate recognition of studies with higher and lower values of Kl. It should ' be noted however, that the uncertainties around this estimate of KL ere such that, an appropriate estimate of KL could lie between 0.003 and 0.03. mortality from lung cancer, The estimates of KH given in Table 15, mesothelioma and asbestosis alone are derived from studies with four of the provide sufficient bases upon which to five highest Kl values. That is. there is justify this action. some bias in examining the value of KM independent of the Kl in the same E. Estimates of Cancer Mortality studies because it is likely that these K* A best estimate of Kl was calculated for each of the eleven epidemiologic would tend to be slightly higher than those derived from other studies, due to studies and an estimate of K* was the demonstrated high power of these calculated from four of these studies studies to detect risk The arithmetic (see Table S) (Ex. 84-392). For each mean of the KH's is 4.98X10-*, and the study, the best estimate for Kl and K,, is geometric mean of the KM`s is 2.91 x 10~*. indicated along with a range of To account for some of this bias when uncertainty. The ranges Hated are those estimating Ku, it is useful to examine the era the result of estimates of exposure ratio of Km to Kl. For the four studies for uncertainties (usually a factor of 2), which Kw was calculated, the range of methodological uncei taintle* that led to the ratios of K to Kl is only two-fold alternate evaluations of risk or exposure (from 0.75x10-* to 1.79X10**). Both the arithmetic and geometric means of these ratios are 1x10"*. Thus, 1X10"* is an appropriate choice as the best estimate of Kh/Kl. Using this estimate of the ratio Kh/Kl and the preferred estimate of KL (0.01). the preferred estimate of KH would be 1X10** (KM=1X10-* XlXlO"*). A range of 3X10"* to 3XJ0"* for Kh would appropriately represent most exposure situations. There is no evidence in this analysis that would suggest that a special lung cancer potency is ascribable to a specific type of fiber. Some of the highest and lowest values for KL are obtained from pure chrysotile exposures (for example. KL calculated from data of Dement et ai. is 0.042; using data from Peto et al. gives a Kl of 0.0078). Exposures involving a mixture of fibers, including amoaite and crocidollte. also span a large range of values for KL. Wide differences also occur in the results of separate epidemiological studies of similar work conditions. Some scientists have suggested that some asbestos processes, such aa asbestos textile manufacturing, may pose a greater hazard than other processes. Far example, mining and milling appears to pose a lesser carcinogenic hazard than manufacturing processes. OSHA compared the potency factors for lung cancer (KJ among different epidemiology studies of manufacturing processes because the potency factors are based on risk per unit dose. No consistent pattern of differential lung cancer risk (i.e., higher Ki's) by process emerged. One study of asbestos textile workers found a very high risk while another found a much lower risk, and the same waa true for the two studies of asbestos production workers and the two studies of asbestos cement workers. Therefore, the choice of a midpoint unit risk for all industrial processes (KL=0.01) is a reasonable and justified choice. The best estimates of KL and KM were utilized to estimate the mortality from exposures to varying concentrations of asbestos for different time periods beginning at different ages. The calculations are age. intensity and duration specific. Tables 10 and 11 show the excess asbestos-related mortality rates from lung cancer, mesothelioma, end gastrointestinal cancer (gastrointestinal cancer excess is assumed to be 10 percent of the lung cancer excess). In these calculations. Equation 1 and Equation 3 were used with values of Kt equal to 0.01 and KM equal to 1X10-- * and the 1977 U.S. male background lung cancer mortality rates. Because of age-specific increases in lung cancer rates In older men since 1977, aBS_008453 lam 025278 51126 Federal Register / Vol. 46. No. 215 / Friday. November 4, 1963 / Rules and Regulations estimates baaed on more recent background rates would be higher. Calculations were done for each.5-year age inierval. and then summed to give a. total lifetime risk. The calculations performed to give the results in Tables 10 and 11 assumed that the relative risk increased following ten years after onset of exposure and continued to rise until ten years after cessation of exposure, after which it remained constant. Table 10 gives estimates of risk for one year of exposure to asbestos at various concentrations for workers beginningexposure at ages 20, 30, 40. 50 and 60. It should be noted that employees exposed at earlier ages show higher risk of all cancers due to the long period of time during which it will be possible for disease to develop. One year of ex'posure to asbestos at 2 f/cc starting at age 20 may result in a total excess cancer risk of 345 per 100,000 workers. Table 11 gives the predicted excess lifetime risk of cancer for exposures of one year, 20 years, and 45 years assuming first exposure at. age 25. Several comments should be made regarding the results in Tables 10 and 11. Though excess relative risk is linear in dose, the excess mortality rates given in Tables 10 and 11 are not strictly linear in dose. Therefore, for example, the risk at 2 f/cc is not exactly 4 times * the risk at 0.5 f/cc, though there ia a close approximation. It should also be noted that the risks for longer periods of exposures do not appear to be a, straight-forward multiplication of the ' risks of shorter duration. lathe longer exposures categories, where exposure will affect older workers, some adjustments have been made for competing risks which are likely to affect the death rate from lung cancer. In addition, when looking at the total cancer risks, it must be remembered that these include the risk of mesothelioma, which is related to time in an exponential fashion. As can be seen from Table 11, the predicted risk, from mesothelioma is '. approximately equal to the lung cancer risk for one year of exposure and, about one-half the risk value for lung cancer in' the 20-year exposure group. The excess, risk of mesothelioma after a lifetime - exposure (45 years) to asbestos is approximately one-third the lifetime excess lung cancer risk. These predictions comport with observations in several populations, where, mortality from mesothelioma is observed to comprise approximately 50 percent of the excess mortality from lung cancer. : Using the equations given earlier, and based on the calculations in Table It, OSHA predicts a lifetime excess risk of total cancer for a lifetime exposure (45 years) to 2 f/cc as 6,411 excess deaths per 100.000 workers, or approximately 64 per 1.000. Recognizing that a 20 year exposure to asbestos may be another approximation of actual worker ' experience of interest, the toodelS predict an excess cancer mortality of ' 4.392 deaths per 100.000'wdrkers. Reduction In the PEL from 2'f/cC tO 04 f/cc reduces the risk from lifetime exposure from 64 per 1,000 to 17 per' 1.000. Similarly, for a 20 year exposure, the risk is reduced from 44 per 1.000. to 11 per 1.000. representing a. 75% reduction in risk. t The lifetime risk from one year of exposure follows a similar course. The risk reduces from 290 per 100,000 at 2 f/: cc. to 74 per 100,000 at 0.5 f/cc. As discussed above, this implies that the lifetime excess risk from a sixmonth exposure to asbestos would be i ' approximately 198 per 100406 at 2 f/cc, and 37 per 100,000 at 04 f/cc. Lastly. Table 11 contain the risks for levels higher than2 f/de, becauseOSHA believes some industrial areas (such at' construction) may be at these higher levels. This population of workers would consequently experience's touch greater"' reduction in risk byexposutes to 04 - f/cc, or lest. ,, aajjM com te-as-it ABS-008454 lam 025279 Federal Register / Vol. 48. No. 215 / Friday, November 4.1983 / Rules and Regulations Table 9 Estimates of the Slope Parameters for Lung Cancer (K^ and Mesothelioma (KM), by Process 51127 5tudy MANUFACTURING KlI/ KhxIO8 Asbestos Products- Henderson & Enterline (1979) .................. - 0.0Q47 (0.002b - 0.0066) Asbestos Cement Products Weill et al. (1979) 0.0033 (0.0016 - 0.0086) Finkeistein- (1983) 0.067 (0.033 - 0.13) 12 (4-30) Textile Products Peto (l980) Dement et ai:(l982) 0.0076 (0.0009 - 0.023) .042 (0.023 - 0.21> 0.7 (Q.3-2) Friction Products berry 1 Newhouse (1983) Insulation Products 0.0006 (0 - 0.008) - - Seidman et al. (1979) 0.068(0.0049-0.14) 5.7 (3-11) INSULATION APPLICATION Selilcoff et al. (1979) MINING-MILLING 0.020 (0.008 : 0.030) ' *" 1.5 (0.5 - 2.5) Liddell et al. (1977) Nicholson et al. (1979) 0.00065 (0.0002 - 0.0011) 0.0023 (0.001 - 0.007) Rubino et al. (1979) .. 0.0051 (0 - 0.009) . 1 Values In parentheses represent* the Tinge'of uncertainty around the estimates of Kj. and Kg. These are computed on * study by study basis and calculations for each study are discussed in the text. Kh/kJxIO6! 1.79 0.92 0.84 .75 i ABS-008455 LAM 025280 51128 Federal Ragjsier / Vol. 48, No. 215 / Friday. November 4.1983 / Rules and Regulations Table 10 ` Estimated asbestos related cancer mortality from a one year exposure to various fiber concentrations Asbestos fiber concentration (f/ml) Cancer mortality /iOO.OOO exposed Mesothe- Lung lioma Gastrointestinal1' Total Age at first exposure: 20 0.1 .7.2 9.5 .7 0.2 14.3 18.9 1.4 0.5 35.8 47.3 . 3.6 2.0 142.8 186.7 14.3 16.2 34.6 86.7 345.8 '* Age at first exposure: 30 0.1 7.3 4.9 .7 0.2 14.5 9.8 1.4 0.5 36.3 24.4 3.6 2.0 144.9 97.4 14.5 12.9 25.7 64.3 256.8 Age at first exposure: 40 0.1 7.1 2.1 .7 0.2 14.1 4.3 1.4 0.5 35.3 10.7 3.5 2.0 140.9 42.6 14.1 9.9 19.8 49.5 197.6 Age at first exposure: 50 0.1 6.1 0.7 .6 0.2 12.2 i .5 1.2 0.5 30.4 3.7 3.0 2.0 121.7 14.7 12.2 7.4 14.9 37.0 146.6 Age at first exposure : 60 0.1 4.0 0.2 .4 - 4.6 0.2 8.0 0.3 .8 9.1 0.5 19.9 0.9 2.0 22.8 2.0 79.6 3.4 8.0 91.0 1 estimated as 10k of lung cancer risk rather than calculated using dose-response information. ABS-008456 LAM 025281 Federal Register / Yol. 48, No. 215 / Friday. November 4. 1983 / Rules and Regulations Table 11 Estimated Asbestos Related Cancer Mortality per 100,000 by Nujfcer of .Years Exposed and Exposure Level 1 51129 Asbestos fiber concentration If/ml) 0.1 0.2 0.5 2.0 4.0 5.0 10.0 0.1 0.2 0.5 2.0 4.0 5.0 10.0 0.1 0.2 0.5 2.0 4.0 5.0 10.0 Cancer mortality /100.000 exposed Lung Mesothe1 mm lioma Gastrointestinal2 Total 7.2 14.4 36.1 144 288 360 715 139 278 692 2713 5278 6509 12177 231 460 1143 4416 8441 10318 18515 l year exposure 6.9 13.8 34.6 138 275 344 684 0.7 1.4 3.6 14.4 28.8 36.0 71.5 20 year exposure 73 146 362 1408 2706 -3317 6024 13.9 27.8 69.2 271.3 527.8 650.9 1217.7 ____ 45 years exposure 82 164 407 1554 2924 3547 6141 23.1 46.0 114.3 441.6 844.1 1031.8 1851.5 14.8 29.6 74.3 296.4 591.8 740.0 1470.5 225.9 451.8 1123.2 4392.3 8511.8 10476.9 13996.7 336.1 670.0 1664.3 6411.6 12209.1 14896.8 26507.5 1 Assumes exposure begins at age 25. Risks are calculated using U.S. male lung cancer background rates for 1977. 2 Estimated as 10* of lung cancer risk rather than calculated using dose-response information. MUMOCOOC 10-1*-C ABS-008457 LAM 025282 51130 Federal Register / Vol. 40. No. 215 / Friday, November 4. 1983 / Rules and Regulations F. Quantifying the Excess Risk From Asbestosis Many studies have found the nonmalignanl lung disease, asbestosis among asbestos-exposed workers. In fact, early occupational health standards governing asbestos exposure primarily sought to reduce the incidence of asbestosis. Asbestosis is a type of pulmonary fibrosis diagnosed on the liusis of a history of exposure to asbestos; it is characterized by radiologic changes to the lung, breathlessness, impaired lung function, and other clinical features of fibrosing lung disease. Asbestosis can be manifested in a range of degrees of severity and can result in disability and death. An early response by the lung to usbestos exposure is formation of plaques, which are opaque patches visible on chest X-rays. The presence of plaques might indicate increased risk of future development of asbestosis. but this is not certain. Although the disease significance of pleural plaques is not clear, the presence of the plaques is not normal. Berry et al. (1979. Ex. 84-20) defines "possible asbestosis" as a lung condition based upon signs of early asbestosis. He classifies "possible asbestosis" on the finding of basal rales or crepitations, radiological changes of varying degree, a falling gas transfer factor, and restrictive changes in lung volume or ventilatory capacity. The definition of ashestosis varies among different investigators. Berry et al.'s definition of possible asbestosis, for example, is not sufficient for the victim to qualify for British Disablement Benefit. Asbestosis can progress to a disabling condition where active work is no longer possible. Although no single definition of disability exists, various governments have adopted definitions of asbestos disability to administer theh worker's compensation laws. Finkelstein explains the system used by the Ontario Workmen's Compensation Board: Workers submitting cluims to the Compensation Board ara examined by the Advisory Committee, which recommends whether or not a Disability Pension be grouted. There are no strict criteria for the .wording of a Disability Pension for .asbestosis. but some of the factors considered include a history of "adequate" exposure, a history of dyspnea, the presence of crepitations or clubbing, radiographic signs ol pulmonary fibrosis and abnormal puimonnry funclinn results (F.x. 84-44. p. 4!)fi). Asbestosis has been known to progress or worsen after cessation of exposure, probably due to irreversible injury ;md/or the retention of asbestos fibers in the king. In addition to hmg function impairment, asbestosis contributes to Increased asbestosrelated mortality. Increased resistance created by the lung obstruction can lead to heart failure. Because of the many possible definitions of asbestosis given by different groups, the quantification of a single risk associated with asbestosis is difficult, it is dear that materiel , impairment from asbestosis occurs prior to the onset of its disabling stage. Quantitative studies exist, primarily for the disabling forms of the disease; specifically, two separate studies provide information to develop a doseresponse relationship between asbestos exposure and incidence of asbestosis (Ex. 84-20 and 84-44). Two definitions will be helpful in interpreting the data concerning asbestosis. Incidence is the rate at which new cases of asbestosis develop in a given period of time. It is a direct measure of the risk of developing the ' disease. On the other hand prevalence measures the number of cases alive in a population at a given period of time, Numerically, it eqoals the sum of ell the incidence cases in the pastminim al! the deaths that have occurred in people who had developed die disease. Prevalence can be reflective of existing risk for asbestosis in a population: prevalence however, can be high for other reasons, such as increased survival. Moreover, incidence establishes a time sequence whereas prevalence looks at both cause and effect simultaneously. The best estimates of risk of asbestosis have been calculated from incidence date of two studies. Berry and Lewtnsohn (Ex. 84-254) and Finkelstein (Ex. 84-44). OSHA has also examined prevalence rates to support a quantitative assessment of risk for asbestos, and these data wrU be discussed first. Berry et al. (1979. Ex. 84-20) studied a group of 379 men who worked at an asbestos textile factory for at least 10 years. Dust measurements were available and were correlated to ends job performed for each year under study. Health effects were correlated to cumulative exposure. Table 12 shows the observed prevalence of crepitations, "possible asbestosis". and certified asbestos for data from Berry et al., as taken from the fitted curve in Figure 4 of their paper. Possible asbestosis was diagnosed by the factory medical officer if he thought that a man was developing signs or symptoms of early asbestosis' 50% of the men diagnosed with possible asbestosis received certification within the 8.5'years following this diagnosis. The results in Table 12 represent a group of workers employed after 1950, who had a relatively short duration of follow up (maximum interval bom first exposure was 23 years). A higher percentage of asbestosis would most likely have been observed if the follow up period in the stlidy had been extended. The observations of Berry et al. presented in Table 12 are probably underestimates of risk. First, the risk shown in Table 12 does not give the probability of developing disease after exposure has ended, but rather, reports the disease found at one point in time. Second, the data includes some workers who may not have been followed long enough for asbestosis to appear. TABLE 12 PREVALENCE Of CREPITATIONS, POSSIBLE ASBESTOSIS AN0 CERTIFIED ASBESTOSIS FOR CUMULATIVE EXPOSURES TO ASBESTOS FROM BERRY et al. (1979) Percent with condition 1 5 10 IS Cumulative exposure, fiber/cc - years Crepitations Possible Asbestosis 37 46 65 84 86 118 100 134 Certified Asbestosis 63 100 130 148 ABS-008458 LAM 025283 Federal Register / Vol. 4B, No. 215 / Friday, November 4. 1983 / Rules and Regululions 51131 These data of Berry et al. demonstrate several features of the nature of (he asbestosis risk. The data show a continuumf ctinical response over the range of doses: that is, these clinical observations support the existence of a dose-response relationship. Second, these observations also support the existence of a low. or possibly no. threshold for asbestosis. since there is increased risk at cumulative exposures as low as 37 fiber/cc-years. Berry and Lewinsohn (1979, Ex. 84254) have reported the incidence of asbestosis in this sameadbestos textile factory. These data are presented in Table 13. A dose-response relationship is apparent for the incidence data, but it is not quite as consistent for the prevalence data. Table 13 Incidence of Possible Asbestosis by Cumulai.ive Exposure Berry and Lewinsohn Cumulative Exposure (fiber/cc-years) X Incidence First employment Before 1951 * After 1950 0-49 50-99 100-149 150-199 200-249 300-349 - 3 2-6 3.9 > 6.2 4-6 0.4 1 2 Finkelstein 11962. Ex. 84-44) looked al the development of compensable {certified) asbestosis among 201 workers at an ajbestes-cement-factory in Ontario. A dose-response relationship was developed iieing estimated cumulative exposures based on plant dust measurements and using medical information from <he Ontario Workmen's Compensation Board. Table 14 shows the incidence of certified asbestosis cases ee a function of cumulative exposure from the Finkelstein study. TABLE 14 LkOSEMCE OF CEKTlfifO ASBESTOSIS AS A FUNCTION OF CUMULATIVE 'EXPOSURE FINKELSTEIN <1982) Cumulative exposure fiber-years/cc X Incidence 0-49 50-99 100-149 150-199 200-249 0.5 J.4 6.5 7.9 14.3 Finkelstein's (1982) observations may overstate the incidence of asbestosis because at-autopsy. there was histologic evidence of silicosis as well as asbestosis in many men. Finkelstein States that "we have, nevertheless, chosen to call their disease 'asbestosis' as we believe that is the pathologic process of most significance. Most of the parenchymal -radiographic abnormalities were small imgdlar opacities and (he mortality pattern among the men was consistent with fhe'toxic'Vffects of asbestos." On the other hand. Finkelstem's-study may have understated asbestosis risk by examining only certified disability from asbestosis. which is an advanced stage of asbestosis. OSHA's estimates of risk were derived isem a.simple linear regression of the incidence on (be midpoints of the cumulative exposure data of Berry and Lewinsohn (Table 13) and Finkelstein (Table 14). A linear relationship was assumed, at least to the point estimation of 0.5 fibers/cc fbr-45 years (or 22.5 fiber years/cc). This assumption is consistent with the fact that early stages of diseases are observed at low exposures. A similar conclusion is drawn in the report of the British Advisory Committee on Asbestos (Ex. 54-218, volume 2. p. 38). "The present authors come down in favor of a dose response relationship without a `threshold for chrysotile within the range experienced in industry." ABS-008459 lam 025284 51132 federal Register / Vol. 48, No. 215 / Friday, November 4. 1983 / Rules and Regulations. TABLE 15 ESTIMATES OF LIFETIME AS6EST0SIS INCIDENCES Exposure level. f loer/cc 0.5 1 2 5 10 Slope R2 i Incidence Finxelstein 1.24 2.49 4.97 12.43 24.86 0.055 0.975 - Berry: first employed; Before 1951 After IS 0.45 0.35 0.89 0.69 1.79 1.38 4.46 1.38 8.93 6.93 0.020 .901 0.015 .994 The results of each of the regression analyses and predictions of incidence for several dose concentrations are given in Table 15. At this time OSHA makes no attempt to extrapolate the data using this-model below the 0.5 fiber/cc level or above 10 fiber/cc level. The estimates from the 3 cohorts differ by an approximate factor of 3. This may be indicative of some of the methodological issues raised earlier. It is possible that the estimates made from Berry and Lewinsohn's data may be underestimates. The maximum duration of follow-up in that study was 23 years, with an average follow-up of 16 years. Observations from Finkelstein's data (his Table 1) demonstrate that only 41% (23/56 cases) of total incidence was experienced in the first 24 years since first exposure. That is, 59% of the asbestosis incidence was not expressed until at least 25 years from first exposure. Thus, it is likely that the low incidence rates in the Berry and Lewinsohn study (and therefore, the low estimate predicted by OSHA) are reflective of the short follow-up period for this group of workers. OSHA believes that the best estimates of the incidence of asbestosis are those derived from the Finkelstein data. They indicate that among workers exposed for a lifetime exposure to 2 f/cc of asbestos, approximately 5% will develop asbestosis. Reducing this level to 0.5 f/cc would reduce incidence to 1.24%. It should be noted that these risk estimates represent incidence of disabling asbestosis. First signs of adverse pulmonary effects are reported to occur at lower levels. C. Other Quantitative Risk Assessments Since 1979. several scientists and scientific committees have estimated risk associated with asbestos exposure. (F.XS. 84-1, 84-2. 84-216, 84-256.) The risk assessments are in approximate agreement. They all use epidemiological studies conducted in the occupational environment to generate quantitative estimates. Animal studies are used to only support and justify methodological procedures and assumptions qualitatively. For lung cancer, scientists generally accept the linear model relating increased asbestos exposure to risk' and generally accept the lack of a threshold. As stated by the British Advisory Committee on Asbestos: "For lung cancer, the available data in man, all of which are derived from industry, show an increase in risk with increasing dose of dust and we find no evidence within the range of dust levels studied for a threshold of dose below which there is no evidence pf risk" (Ex. 84-216, p. 55). In general,, all the risk assessments use cumulative exposure as the measure of exposure for all cancer risk estimates (cumulative exposure equaling intensity times duration of exposure). Given the uncertainties inherent in quantitative risk assessment as well as the inevitable variations in findings among the many epidemiologic studies, OSHA believes that the different quantitative risk estimates agree relatively well. The variations that do exist can be explained by the assumptions made or by simple methodological differences. Some variations may be due to differences in the work environment used in the assessments. For example. Crump's assessment, given as testimony to the Ontario Royal Commission on Asbestos in August 1981, contains quantitative estimates of risk for several studies, including smoking-specific risk estimates. Crump does not give a best estimate of risk or an overall risk estimate. His estimates of risk are based upon an assumption that worker exposure that would result from a standard set at a 2 fiber/cc limit is actually much lower than 2 fibers/cc. That is, his risk estimate under a 2.0 fibers/cc standard assumes that average worker exposure would be only 1.0 fibers/cc (p. 49). He reduces this level by a factor of two to account'for differences between personal and static sampling (p. 50). Thus. Crump's estimates for a 2 fiber/cc standard will be four times lower than those estimates made by OSHA for average worker exposures of 2 fiber/cc. Such differences in assumptions should be kept in mind when comparing risk assessments by different authors (Crump's estimate at 2 f/cc would be compared to OSHA's estimate at 0.5 f/cc). OSHA has presented the risk as if the working population were exposed to an average concentration of 2,1.0.5 and 0.1 fibers/ cc reflecting OSHA's belief that a standard of 2 f/cc does not preclude worker exposure at that level; in fact exposures may even exceed 2 fibers/cc for short periods and still produce an 8hour TWA below 2 f/cc - The report of the British Advisory Committee on Asbestos contains one of the first quantitative risk assessments performed for asbestos. This was updated in 1983 by Acheson and Gardner (Ex. 84-243). The report describes risks for lung cancer, mesothelioma, other asbestos related cancers and asbestosis and itcontainsa rather thorough description of the health hazards associated with asbestos exposure. The British report's risk estimates for lung cancer do not differ in a major way from OSHA's estimates. For example, with regard to exposure to chrysotile. the Acheson and Gardner update states that "We concluded that, for example, an excess mortality from asbestos-related disease of 2 percent might be associated with any point in a range of from 5 fibers/ml to 0.4 fibers/ml and that bearing in the .mind the very considerable uncertainties a figure towards the lower end of the array [0.4 f/ml] might represent an appropriate compromise." (Ex. 84-243, p. 14). As a comparison from Table 11, OSHA's best estimate of risk is that an exposure to 0.5 fibers/cc (ml) would result in a 1.2 percent increase in deaths from lung cancer (or 1143 excess deaths per 100,000). VI. Technical and Economic Feasibility Based on an evaluation of evidence contained in the record. OSHA finds that the provisions required by the ETS are technically and economically feasible. OSHA has examined the various industries and work operations impacted by the standard and their ability to comply with the provisions of the ETS. Because the ETS requires prompt reduction of risk. OSHA ABS-008460 LAM 025285 Federal Register / Val. 48. No. 215 / Friday. November 4. 1983 / Rules and Regulations 51133 assessed the industry's ability to concentrations, furthermore, for implement the required controls purposes of worst-case analysis, OSHA immediately. assumed (hat none of these industries The ETS allows considerable flexibility In achieving the PEL. As a has'any respirator program except for the shipbuilding and construction result, three options are available to segments. OSHA snakes this assumption lower the asbestos fiber concentrations because OSHA estimates thatsnod to which workers are exposed: (1) , workers in industries other than Engineering-control* auch as automatic shipbuilding and construction are bag opening devices, specialized exposed to eight-hour time-weighted vacuum equipment and increased averages less than 2 f?cc, and the OSHA ventilation; (2) work practices, such as standard issued in 1972 only requires a wet treatment of the asbestos material respirator program when engineering and increased clean-up of the work controls and work practices cannot place, and (3) use oT approved bring exposures-to Zl/cc. To the extent respirators. Due to the emergency nature that some firms do have an existing of this action which requires immediate respirator program, the costs are response to reduce worker exposure, overestimated. OSHA assumes that respirators will be the initial method used to comply with A. Technical Feasibility the ETS. A full discussion of the Theiellowing table presents the technological and economlcieasibittiy assumptions OSHA made regarding the of the alternative methods for each respirator program Elements required by industry and for the variousRBLs under each industry to obtain a PEL of 0.5 f/cc. consideration for revising the permanent The types of respirators needed for each standard will accompany the proposal industry sector were determined using which will be published separately. '. the respirator selection table in the ET8 OSHA believes that consistent wifh with reference to the estimated current the estimates of current exposure levels, exposure conditions. OSHA assumed engineering controls are cuwerrtlyta that the least costly approved respirator place and work practices, in operation would be selected. For example, where which, if applied conscieafiaaaty, Mould industries ihnare snporases Jess than ten immediately result in concentrations A times the PEL OSHA anticipated that least as low as 0.5 in many industries' disposable respirators would be (Ex. 84-262; Ex. 84-283; Ex. 84-9 and Ex. purchased, because of their lower short 84-295). For purposes of assessing the term costs. When exposures exceeded technological and^economic feasibilrty, ten times theTOL, OSHA assumed that however. OSHA assumed a worst-case some plants woukd eithense air line scenario in which each industry segment respirators, or Ml facepiece respirators, would have to implement a respirator depending upon the operation. To the program in order to achieve immediate extent that firms choose a higher-cost reduction in worker exposure levels respirator to increase the protection below the estimated current factor or durability, respirator costs may be understated. Furthermore. OSIIA hits not included in the cost analysis a consideration for lost worker productivity due to wearing respirators. Costs may be understated by whatever amount productivity is reduced. Other anticipated respirator program elements required to determine costs Tor the ETS are listed below. These elements are derived from the existing provisions found in the Asbestos Standard. 29 CFR 1910.1001 and the standards for respirators. 29 CFR 1910.134. All of the required respirators and filters are readily available and can be purchased through local distributors. Since the program relies mostly on disposable respirators. OSMA considers that there will be no supply constraints. As the worst-case (or high) estimate. 'OSHA assumes that approximately 50.000 workers will wear respirators because of the ETS who did not previously wear respirators. OSHA has concluded that the use ofrespirators will be effective in providing improved worker protection during the period of the ETS. la addition to encouraging generally more widespread use of respiratory protective measures, the ETS will stimulate a heightened understanding of the health hazards from asbestos exposure and will result in more effective use programs and .strategies. Issues involving the appropriateness of respirator use as a long-term nohitioa to controlling asbestos exposures are raised in (he section 9(b) rulemaking proceeding (see Ex. 84-345. 84-346. 84-347, 64-348). SHUNQ COOS 4S10-M-K ABS-008461 LAM 025286 51134 Federal Register/- .Vol.48, No. 213 / Friday. November 4.1983 / Rules and Regulations TABLE 16 Anticipated Respirator Program Elements Required to Meet PEL Industry Segment - Primary Manufacturing A/C Pipe & Sheet Friction Materials Asbestos Paper Gaskets Floor Ti le Paints, Coatings and Sealants Textiles Wet Process * Dry Process Secondary Fabricators Cement Sheets, Paper Products, Packing and Gaskets Anticipated Type of Respirators Used to Meet the PEL ' 1 disposable/workday/worker 2 air-line/plant (for high concentration situations)4 Other Program Elements L- Full admin, costs Fit Test*Training" 1 disposable/worker/d^y for portions exposed at 0.75 50%~admfn. costs r Fit Test Training ' ^" ' : (Nothing for PEL of 0.5) i disposable/worker/day for 75% of workers , 75% admin, coiti ' Fiti.Test Training . v; 1 dispoSabTe/worker/day 1 disposable/worker/day 1 disposable/worker/day 2 air-line/plant for high concentration situations Full-admin, costs Fit Test . ' Training fol1 admin, costs Fit Test Training . ~ . 1% admin, costs < Fit test Training 1 ;,b > /. i Textiles 1 disposable/worker/day ` * Automotive Aftermarket 1 disposable/worker/day Rebuilding and Kefacing 1 air-1, tne/plant -- 1% admin, costs Fit Test Training . 1% a<tain. Costs Fit Test Training ABS-00B462 LAM 025287 Federal Register / Vol. 48. No. 218 / Friday. November 4.1983 / Rules and Regulations Respirator Program Requirements (Cont.) 51135 Industry Segment Anticipated Type of Respirators Used to Meet the PEL Other Program Elements Brake Repair Gasoline Stations Shipbuilding/Repair Shipbuilding Ship Repair Nothing ' Nothing Nothing 1 disposal)te/worker/day for 20* of total workforce 201 of admin, costs Fit Test Training Construction Installation A/C Pipe A/C Sheet 1 disposable/worker/day 101 admin, costs Fit Test Training Roofing Felts OemoTItion/Renovatlon Nothing 1 HEPA full face for 2SX of the workforce Repair 4 Maintenance ta thing 25* admin. Costs Fit Test -Training *OSHA assumes -that certain jobs such as cleaning of vacuum equipment wlU produce exposure levels greater than 10 times the PEL. Consequently, OSHA assumes 2 air line respirators will be required per plant except for small operations which will require only one air-line respirator per plant. ^Administrative costs represent the salary of one full time professional at (30,000 per annum and one fall time clerk at (IS,000 per annum. OSHA has adjusted administrative costs -to represent expected reasonable time spent administering a respirator program in each industry. In the secondary fabrication and automotive aftermarket sectors, for example, OSHA anticipates that supervisors will spend IS minutes a day.distritouting and monito. ing the use of disposable respirators. These sectors - have: far less of a management burden than large -firms* and thus, administrative oosts are calculated at 1 percent of the full administrative amount. OSHA believes that administrative costs have been overstated in most segments in order to present a worst case scenerio. c0SHA assumes that every employee must be fit tested for respirator use at a cost of (21 per employee. d0ShA has allowed 3 hours for respirator and asbestos training. OSHA considers that this is very liberal, especially in industry segments characterizeo by small plants and by the use of disposable respirators. Costs for this element are therefore overstateo to present the worst-case scenario. HUMO COM 4C1S4K ABS-008463 lam 025288 51138 Federal Register / Vol, 48, No. 215 / Friday. November -4. 1983 / Rules and fteguiufana OSIIA also examined the feasibility of analytical methods for measuring asbestos air concentrations at the 0.5 f/ cc limit established by the ETS. OSHA's existing standard requires that all measurements of airborne concentrations of asbestos fibers be made by the membrane filter-method at 400-450 x (magnification) (4 millimeter objective) with phase contrast illumination (29 CFR 1910.1001(e)). After reviewing the relevant evidence made available since the 1975 proposal, OSHA finds that phase contrast microscopy is a feasible and effective method for measuring airborne asbestos fibers to determine compliance with the permissible exposure levels set by the Emergency Temporary Standard. The most important issue raised by communis to the 1975 notice Is whether phase contrast microscopy analysis is capable of reliably measuring airborne concentrations of 0.5 fibers/cc and less. OSHA acknowledges that asbestos airborne measurement procedures using phase contrast microscopy, as with any industrial hygiene analytical procedure, inherently contains several error sources. These errors have been statistically evaluated by Leidel et al. (Ex. 84-02), and. in 1979, by the Air Monitoring Committee of the Asbestos Information Association/North America, referred to herein as AIA (Ex. 86-002) both using round-robin sample exchange data. Chatfield also examined this question (Ex. 84-319). In the Leidel et al. and the AIA evaluations, the error, measured as a coefficient of variation (CV), was found to be related to the number of particles counted from the filter. For 100 fibers counted Leidel, et al. found a CV of 0.12. whereas the AIA report found a CV of 0.351 or errors associated with interlaboratoryintrafilter variability. Based upon these studies, taken at their face value, it appears that the phase contrast microscope analysis is capable of reasonably Tellable measurements at 04 fibers/cc. As stated in the AIA report (Ex. 86-002, p. AB-2), "The calculated results indicate that the 1 IV AIA report alto reports * separate CV Tor `Kt-rull v.iruJiilify in (he sample evaluation step plus *rsndom vanebility ia the sample collection sl'p." This "mnduni variability of sample collection'' as described in this study may be a r tonsure of the Irve difference in air concentrations from one location to the next and may not be a nuNtsnn* of random variability. A more carefully tlf'ignmi sampling strategy with precise control uf instrumentation placement, air currents, and dust generation is necessary to eliminate differences in airborne dust concentrations from one location lo .mother 95% confidence limits on a measured 8- B. Economic Feasibility hour TWA can be relatively constant with a wide. buLuaable. range down lo concentrations approaching 0.5 fibers/cc [less than) 5M." The industry costs based on the program elements described in Table 16 are presented in Table 17. For the ETS. the costs of asbestos training are The AIA report shows a higher error included with respnaforframing.Costs' than does the Leidel et al. report One _ for waminajigns are not included. possible reason for this difference may ~ since, for the purposes of tfie ETS, these be that the AIA report assessed the variability in measurements as they are being made today by the many laboratories who are making the measurements. For example, only 27 of the 46 laboratories participated in the PAT program and no counting guidelines were given, whereas, the Leidel et al. report included only a small number of laboratories operated by fohns Mansville Company, that probably used very similar procedures and conducted similar training. In late 1982, Chatfield prepared a report entitled "Measurement of Asbestos Fibre Concentrations in Workplace Atmospheres" for the Royal Commission on Matters of Health and Safety Arising from the Use of Asbestos in Ontario (Ex. 84-319). Chatfield signs could be hand-madeA? veryJew costs. OSHA did not analyzecoeta . associated with the alternate benefits scenario (See Table 4) since that scenario was constructed to show a \ . lower range of benefits but does not represent an estimate of current industry practice. ' The costs are overstated to the extent that some firms already have a current respirator program and to the extent that careful application of existing engineering controls and workpractical would reduce concentrattous to the PEL in some firms and thereby sake respirator use unnecessary.- ' Furthermore, the costs assume that dust masks and filters will have to be replaced every 8-hour day. Some of this ' cost (i.e., the disposable respirators, filters and administrative overhead) can analyzed intra-and inter-laboratory variability and arrived at conclusions be in fact spread onerfba period dreiqg which the ETS is in effect. somewhat similar to those of AIA and Leidel. Chatfield also recommended Finally, the costs assigned to the Shipbuilding/Repair and Construction- methods by which the accuraey and----- ~ Deftdlition/Renovation segments precision of phase contrast microscopic represent cost which refleef Increased analytical techniques could be compliance with current obligations (e.g. improved. Significantly, he noted "in increased respirator use to meet the view of the number and frequency of' - current standard as a regal* of the ETS measurements required, there is training requirements) and are not currently no fully developed alternative , directly attributable lo the ETS. method [to phase contrast microscopy| which could be immediately `` implemented." OSHA notes that the authorities cited Not*.--OSHA anticipates that Ike ETS will spur many employs* who previously were not in compliance to expend the necessary resources in order to come into compliance above believe that it may be possible to with the BTS., reduce phase contrast microscopy errors if improved and standardized procedures are followed, perhaps by adding requirements to the standard. It does not appear, however, that improvements of this nature can be . quickly made in the immediate format of The total cost la eatimated at $35,565,402 for 6 months. This translates into an average coal par employee of $708. Average 8 month soils per worker presented on an industry basis in Table 17 range from $251 in the automotive aftermarket segment to $873 in the this ETS. Therefore, based on the construction segment These costs are evidence before it at the time of not largo portion of indaolry shipments issuance of this emergency standard. as presented in Table X6. Moreover, OSHA beEeves'that it is generally not firms in these Industries will be able to possible to measure asbestos pass the costs forward because asbestos- concentrations below 0.5 f/cc reliably substitutes in most industries are not and reproducibly using phese contrast immediately available. Fof aH these microscopy under current laboratory reasons, OSHA finds that the ETS is practices. OSHA. finds that the phase economically feasible. contrast microscopy method can be ' feasibly used to measure asbesfos air concentrations down !o 0.5 f/cc. mujno coot -!*-* ABS-008464 lam 025289 Federal Register / Vol. 48, No. 215 / Friday, November 4,1983 / Rules and Regulations Table 17 Estimated Costs of Respirator Program for ETS with PEL of 0.5 Industry Segment Total Costs (S) V - year 6 - months 6-joo. average Cost/employee Primary Hanufacturing 6.404.965 4,143,242 461 Secondary Fabricators Automotive Aftermarket 13,547,224 1,620,197 9,435,532 1,027,864 54o 251 Shipbuilding/repair Constructs 626,266 36.329.740 324.526 19,634,236 533 973 Total 58,528,412 34,565.4(12 708 51137 Table 16 ETS Compliance Costs Compared ' to Sales by Industry Segment Industry Segment Cost/Sales i Primary hanufacturing .003 A/C Pipe t Sheet .002 Frictin Materials .006 Asoestos Paper .003 Paints, Coatings ana Sealants .005 Gaskets, Seals and Packings .001 Textiles .014 Secondary Fabricators ' Ceeent Sheets Paper Products Packings and Gaskets Textiles .011 .072 .011 .010 .014 Automotive Aftermarket Rebuilding and Refacing . .0004 Shipbulldtng/Repair Construction .0001 - .0004 Note: Sales values are for 1977 and 1978. -Values for Primary Manufacturtng and Secondary Fabricators were taken from 84-003; values' for the Automotive Aftermarket were taken from U.S. Industrial Outlook. 1983; and values for Shipbuilding/Repair and Construction were taken from the 1982/83 Statistical Abstract. The value for Construction was adjusted by substracting the value for Highway A Streets and Residential and adding the value for Oemolition/Urecking. mujnq coot mw-ja-c ; ABS-008465 lam 025290 51138 Federal Register / VoL 48. No. 215 / Friday. November 4. 1983 / Rules and Regulations VII. Environmental Impact, Requirements of Executive Order 12291, and the Regulatory Flexibility Act The National Environmental Policy Act (NF.PA| of 1909 (U.S.C 4321 et seq.). ns implemented by the guidelines (40 CFR Part 1500) of the Council on Environmental Quality (CEQ). requires that federal agencies assess their regulatory actions to determine if there is a potential for a significant impact on the quality of the human environment and. if necessary, to prepare an' environmental impact statement. In accordance with these requirements and UOL NEPA regulations (29 CFR Part 11 (Subpart B, n.in (a) (4)1) OSHA had determined thai because of the emergency nature of this standard, no environmental impact statement will be perpared for (he ETS. The courts have held that NEPA does not require advance preparation of an environmental impact statement for an OSHA ETS {Dry Color Manufacturing Assncation v. U.S. Department of Labor 486 K.2d 98. 107 (3rd Cir. 1973)). OSHA will assess, however, the environmental effects of the proposed permanent regulation of asbestos. The results of this study will be available for review and comment prior to the hearings on the proposed permanent standard and will be an appropriate issue for discussion at thc~public hearing. In the interim. OSHA welcomes any comments on any environmental effects that might occur as a result of promulgation, of a rule on asbetos. Pursuant to the authority of Section 8(a)(1) of Executive Order 12291 OSHA has not prepared a Regulatory Impact Analysis of this ETS. Preparation of such an analysis was not practicable for OSf (A to perform in time to issue this ETS to respond to the grave dangers faced by asbestos exposed workers. OSi IA. however, is completing a Preliminary Regulatory Impact Analysis of the proposal to revise the permanent standard which will be made public at the time the proposal is published. The Regulatory Flexibility Act requires an agency to prepare a Regulatory Flexibility Analysis only for those rules for which a notice of proposed rulemaking is published. OSHA's issuance of an ETS therefore is not covered by the Regulatory Flexibility Act because the OSHA Act provides that ETS's be issued without regard to notice, public comment and oilier requirements in the Administrative Procedure Act. The proposal to revise the permanent asbestos standard, however is subject to the requirements of the Regulatory Flexibility Act and OSHA will evaluate the proposal to ascertain whether employers must be able to comply with analysis under the Regulatory Flexibility the standard in a short period of lime. Act is required. . OSHA has set a level which is likely to VIII. Summary and Explanation of the ETS The requirements of this emergency temporary .standard are set out in a new paragraph, j I910.1001(k). They are limited to additional provisions to the existing asbestos standard which OSHA considers essential and feasible to protect employees from the grave danger resulting from asbestos exposure until OSHA can promulgate more comprehensive revisions in accordance with section 6(b) of the Act. The major changes in the standard made by the ETS are: (1) The new permissible exposure level: (2) methods of compliance permitted to achieve the new level; and (3) a requirement for the establishment qf'training programs within 30 days. The following section discusses the major provisions of the ETS. the necessity for including these provisions in the ETS, and some additional provisions to trigger certain requirements at the reduced permissible level of 0.5 f/cc, 1. Paragraph (k)(l) Scope. As part of the ETS. OSHA is adding a paragraph on the scope of the standard. The paragraph is intended to make clear that the emergency standard applies to all workplaces where employees may be exposed to asbestos in all industries covered by the current asbestos standard ie. general industry, construction and maritime. 2. Paragraph (k)(2) Permissible level of exposure. As part of the ETS, OSHA is adding paragraph (k)(2) which sets 9 new PEL. effective immediately, of 0.5 f/ cc on an 8 hour time weighted average basis. This reduced level may be achieved by any feasible combination of engineering controls, work practices and respiratory protection in order to allow employers to institute effective measures tp reduce employee exposures immediately. OSHA chose 0.5 f/cc as the be achieved immediately, using equipment and control techniques that are currently available Second. OSHA believes, based on the data generated by OSHA's contractor. Research Triangle Institute, that some workplaces, especially in the manufacturing sector, may be dose to achieving a 0.5 f/cc level through the more rigorous use of engineering controls, work practices and housekeeping methods which are now in placer OSHA encourages employers to continue their efforts to implement these methods in order to assure that, for the long term, the most comprehensive and effective program of protection from asbestos exposure wifi be provided. Third. OSHA believes that reliability of the currently required asbestos measurement methods to measure asbestos exposures less than 0.5 f/cc ` should be open for discussion during a rulemaking hearing, rather than imposed through an ETS. OSHA has therefore not adopted the 0.1 f/cc level petitioned by the unions but instead is considering adding provisions to the asbestos standard that may improve the reliability of both sampling and analysis in the 6(b) rulemaking for the permanent ^standard and thus allow lower levels to be reliably measured. 3. Paragraph (k)(3) Methods of compliance. The ETS adds a new paragraph (k}(3). which permits 'employers to reduce the permissible exposure limit from the current 2 f/cc limit to the 0.5 f/cc limit by any feasible combination of engineering controls. (work practices and respiratory protection. The current requirement in paragraphs (c)(1) and (c)(2) to first utilize engineering controls and work practices to reduce exposure levels to 2.0 f/cc remains unaffected by this ETS. Flexibility in choosing compliance strategies for the period of an ETS has been incorporated in most other permissible exposure level primarily previously issued emergency standards. because it believes that reducing The policy reflects OSHA's employee exposures to 0 f/cc in all determination to craft emergency industries will greatly reduce the risk.of standards that are truly responsive to developing asbestos induced cancers, `emergency conditions and which afford primarily lung cancer, mesothelioma and immediately available protection to gastrointestinal cancer. As set out in the affected workers. discussion on "grave danger" and 4. Paragraph (k)(4) Employee "benefits", the number of lives OSHA information and training. The ETS adds believes may be saved through an a paragraph to the asbestos standard immediate reduction of exposure to 0.5 requiring the employer to provide a f/cc is substantial. training program within 30 days of OSHA also believes that the 0.5 f/cc publication of the emergency standard level is appropriate for several cither for all employees whose exposures are reasons. First, because an emergency reasonably expected to exceed the PEL standard must be feasible, and of 0.5 f/cc. without regard to the use of ABS-008466 lam 025291 Federal Register / Vol. 48, No. 215 / Friday. November 4, 1983 / Ruieg and Regulations 51139 respirators, and to assure that all such employees participate in the program and are informed of specified categories of information. OSHA considers this provision to be "necessary" within the meaning of section 6(c) of the Act. to reduce the grave danger faced by asbestos exposed employees. Hie absence of a training program requirement in the asbestos standard has been pointed out as one of the serious deficiencies of the current standard. OSHA believes that participation in an adequate training program is essential for the protection of employees because most protective provisions enlist the employee as an active participant. For example, many employees handling asbestos depend on effective woric practices. Without training in applying these work practices, employee protection would be inadequate. Where the employee plays a more passive role in his protection such as where engineering controls are relied on, training is also essential, because the employee must know the sources of workplace asbestos contamination, and the health hazards of asbestos exposure, in order to assess his own exposure situation and to help recognize when engineering controls are not operating -property. Certainly where housekeeping plays an important role in control, instruction about housekeeping methods, for example, frequent vacuuming, is essential. Perhaps most importanly. where employee protection depends upon respirator use, OSHA's experience shows that training employees in the use, fitting and limitations of respirators is critical to the effectiveness of respirator protection. Accordingly/his requirement applies where airborne concentrations are reasonably expected to exceed 0.5 f/ cc. even if employees use respirators to reduce breathing zone concentrations and thereby comply with the ETS. As set forth in paragraph (k}(4) the employer must inform the employee of the health effects of asbestos exposure; the relationship between asbestos and smoking in producing lung cancer; the operations exposing employees to asbestos fibers and necessary protective steps to minimize exposure; the purpose, proper use. fitting instructions and limitations of respirators, and the contents of ail the provisions of the Asbestos Standard at 1910.1001. 5. Paragraph (k)(5) Respiratory protection during the ETS. The ETS adds a new paragraph (k)(5) which includes a table which ties respirator selection to the 0.5 f/cc PEL. Under the ETS. the concentration cut-offs for various kinds of respirators are multiples of the reduced PEL of 0.5 f/cc. rather than multiples of the previous 2 U cc permissible limit. For example, approved air purifying ,'espirators may be used only where asbestos concentrations are not expected to exceed 5 f/cc (10 x the PEL). Before the ETS. because the PEL was 2 f/cc. such respirators could be used where asbestos concentrations would not have exceeded 20 f/cc (10 x the PEL of 2 f/cc). Similarly, powered air purifying respirators may be used where asbestos concentrations do not exceed 100 times the PEL which at the new level of 0.5 f/ cc is 50 f/cc. Previously, employers could have used such respirators at concentrations up to 200 f/cc. It is tikely that the mala Impact of the reduced PEL on respirator choice will be ip operations and industries where exposure levels are between 5 f/cc and 20 f/cc. Formerly, employees exposed in this range could use half-mask air purifying respirators; now they must be protected by a powered air purifying respirator or a full facepiece respirator, or they may ase a supplied air respirator. 8. Paragraph (k)(8). Warning signs during the ETS. The ETS requires that legible signs warning of the health hazards of asbestos be displayed at locations where airborne concentrations of asbestos Fibers exceed the reduced exposure limit of 0.5 f/cc. No specific legend is required signs for newly posted during the ETS. OSHA wishes to make as practicable as possible the rapid posting of signs, especially in workplaces where there has been previous non-compliance and in areas where asbestos concentrations were formerly below the 2.0 f/cc PEL XI. Public Participation Interested persons are invited to submit written data, views and arguments with respect to the revisions to the asbestos standard made by the ETS. These comments must be postmarked on or before January 3.1984 and sent to the Docket Officer, Docket No. H-033C, Occupational Safety A Health Administration, U.S. Department jof Labor. 200 Constitution Avenue. NW., Room S-6212. Washington. D.C, 20210. The data, views and arguments that are submitted will be available for public inspection and copying at the above address. All timely written submissions will be made a part of the record of the proceeding. List of Subjects in 29 CFR Part 1910 Occupational safety and health. Asbestos. Health. Emergency temporary standard. Cancer. Authority and Signature This document was prepured under the direction of Thome C. Auchter. Assistant Secretary of Labor for Occupational Safety and Health. U.S. Department of Labor. 200 Constitution ' Avenue. NW. Washington. DC 20210. Pursuant to Sections 8(b). 6(c). 8(c) and 8(g) of the Act. 29 CFR 1910;1001 is amended by adding a new paragraph (k) as set forth below. (Sec*. 8(bj, 8(c). 8(cJ and 8(g). Pub. L 91-599. 84 Stat. 1593.1568.1599.1800; 29 U.S.C. 855. 857; Sec 107. Pub. L 91-54. 83 Slat. 98 (40 U.S.C. 333); 29 CFR Part 1911; Sccretury of Libor's Order No. 9-83 (48 FR 35738J) Signed at Washington. D.C.. this 2nd day of November 1983. Thome C. Auditor, Assistant Secretary ofLabor. PART 1910--{AMENDED] Section 1910.1001 of Part 1910 of Title 29 of the Code of Federal Regulations is hereby amended by adding a new paragraph (k) reading as follows; $1910.1001 Astasia*. *9 (k) Emergency temporary standard effective November 4.1983.--(1) Scape. This emergency temporary standard is issued pursuant to section 6(c) of the Act and applies to all workplaces where employees may be exposed to asbestos in all industries covered by the Act. including, general industry, construction and maritime. Except to the extent modified by this emergency temporary standard all provisions of $ 1910.1001 remain in effect. (2) Permissible levels of exposure. The 8-hour time-weighted average airborne concentration of asbestos Fibers to which any employee may be exposed shall not exceed one-half (0.5) Fiber, longer than 5 micrometers, per cubic centimeter of air, as determined by the method prescribed in paragraph (c) of this section. (3) Methods of compliance with the emergency temporary standard. Notwithstanding any other requirements of this section, compliance with the reduced exposure limit of 0.5 f/cc shall be achieved by any feasible combination of engineering controls, work practices, and personal protective equipment and devices. (4) Employee information and training.--{i) As soon as possible, but not later than thirty (30) days from the effective dale of this emergency temporary standard, the employer shall institute a training program for all employees exposed to airborne concentrations of asbestos in excess of ABS-008467 LAM 025292 51140 Federal Register / Vol. 48. No. 215 / Friday, Novtjgiber 4, 1983 / Rules and Regulations 0.5 f/cc. without regard t<rthe use of respirators and shall assure their participation in the program during the effective period of this emergency temporary standard. (ii) The employer shall assure that each such employee is informed of the following: (A) The health effects associated with asbestos exposure; (B) The relationship between asbestos and smoking in producing lung cancer (C) The nature of operations which could result in exposure to asbestos and necessary protective steps to minimize exposure including, as applicable, engineering controls, work practices, respirators, housekeeping and protective clothing: () The purpose, proper use, fitting instructions and limitations of respirators permitted by the standard; and (E) A review of all the provisions contained in 1910.1001. . (5) Respiratory protection during the ETS Notwithstanding any other requirement of this section, where respirators are used to achieve the permissible exposure limit of 0.5 f/cc they shall be selected according to Table 1. ' (6) Warning signs during the ETS. In uddition to the requirements of paragraph (g)(1) of this section, legible signs warning of the heolth hazards of asbestos shall bp provided and' displayed at each location where airborne conoentntions.pf asbestos fibers may exceed 04//be. ' TASLE 1 . Respiratory Protection For Airborne Concentrations of Asbestos , Airborne Concentration of Asbestos (TWA) Not in excess of 5 f/cc (10 X PEL) Not in excess of SO f/cc vlOO X PEL) Greater than SO f/cc Required Respirator1 Reusable or singla*use air purifying respirator Full facepiece air purifying respirator, or a . powered air purifying'res pirator A type *C" continuous flow or pressure deaand, supplied air respirator. 1 Respirators specified for higN concentrations nay be used at lower concentrations of asbestos. (Secs. e(b). 6(c). 8(c) and 8(g). Pub. L. 81-596. 84 Slat. 1593.1596.1599.1600; 29 U.S.C. 655. 657; Sec 107,Tub. L. 91-64.83 Stat 96 (40 U.S.C 333); 29 CFR PartWIT, Secretary of > Labor's Order No. 9-83 (46 FR 35736)) ` ` |fr Ooc. n-ioon rtkd BfcLJNQ hj p*| ABS-008468 LAM 025293