Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 306
Health Effects of Exposure to Radon: BEIR VI E-ANNEX 1Exposures to Miner Cohorts: Review of Estimates for the Studies Colorado Plateau Uranium Miners Introduction Uranium mining in the Colorado Plateau expanded rapidly in the post-World War II period to include more than 200 mines by 1950 (see Time Line E-1). The start of an industry and the boom times did not lead to orderly administration and record keeping, c.f. Czechoslovakia and Ontario below. Moreover, some of the miners who worked in the mines during the post-war uranium boom had previously worked the same ore bodies for radium and vanadium without any accounting of exposure to radon progeny. Most of the early mines were small and depended on natural ventilation so that ambient temperature change was the driving force for exchange of the mines' air with outside air. Until 1967, mining operations were regulated only by the states where mining was taking place, even though all ore was sold to the Atomic Energy Commission. There was no requirement in place for measurement of exposure and there was not a federal standard for exposure to radon progeny. Consequently adequate ventilation practices were not uniformly introduced from the outset and the extent of radon measurement was initially quite limited. As a result, estimates of cumulative exposures to uranium miners on the Colorado Plateau were largely based on various estimation procedures rather than direct measurements relating to a particular mine shaft or even the mine where a given worker was exposed. The history of radon exposures to the miners was described by Holaday (1969) and the approaches followed by the U.S. Public Health Service for esti-
OCR for page 307
Health Effects of Exposure to Radon: BEIR VI mating exposures of individual participants in the epidemiologic study of Colorado uranium miners are described in National Institute for Occupational Safety and Health-National Institute for Environmental Sciences Joint Monograph No. 1 (Lundin and others 1971). A 1968 report of the Federal Radiation Council addressed the accuracy of the exposure estimates. SENES Consultants Limited of Ontario, Canada, has prepared a report entitled "Preliminary feasibility study into the re-evaluation of exposure data for the Colorado Plateau uranium miner cohort study" (SENES 1995). This report provides an extensive description of the calculation of the WLM values for the epidemiologic study and gives insights into the sources of variability and error in the estimates. Estimation of WLM The following description is taken largely from the 1971 monograph authored by Lundin and colleagues. The U.S. Public Health Service began surveying for radon in uranium mines in 1949. In 1950 they were joined by the Colorado State Department of Health and in 1951 by the U.S. Bureau of Mines for mines on Indian reservations. Coverage was far from complete; 1949 "a few measurements," 1950 "relatively few mines," 1951 "but again coverage was incomplete," (Lundin and others 1971). By 1952 an effort was made to survey all operating mines and radon progeny were sampled in 157 mines. This sampling may have examined most of the larger mines, but government records
OCR for page 308
Health Effects of Exposure to Radon: BEIR VI indicate that over 450 mines shipped ore in 1951. Mining companies introduced radon surveys in 1956 and the state programs continued through 1960. Both company and state-sampling efforts were made in work areas for information purposes, not for control purposes, and "are considered to be representative of the areas in the mines in which miners were exposed" (Lundin and others 1971). This early data base is of primary importance in considering the adequacy and precision of miner's exposure estimates as utilized in epidemiology assessments of risks due to radon since a large portion of the cumulative exposure occurred in the 1950's. By 1960, exposure levels had dropped precipitously in anticipation of Colorado's adoption of a 10 WL shutdown level in 1961. However, regulatory control probably reduced the validity of the measurements in mines for epidemiologic purposes. As outlined in Joint Monograph No. 1, the most complete description of the Colorado Plateau miner data (Lundin and others 1971) "Most radon daughter measurements available from Colorado, Utah, and Wyoming after 1960 were made by mine inspectors who measured air samples primarily for control purposes." This may have led to bias in the estimated exposures. As noted by Lundin and others (Lundin and others 1971), "Proportionately more measurements were made in sections of mines having high levels which tended to yield radon-progeny values greater than would have been obtained by sampling all work areas with equal frequency." In addition more measurements were concentrated in mines having high levels of radon. The U.S.P.H.S. investigators who assembled the data base for estimating cumulative exposures chose to exclude company measurements made after 1960 on the grounds that they might have been "minimized to avoid regulatory action." The aim was clearly ''to assure a consistent direction of bias, that is, over estimation of radon daughter levels" (Lundin and others 1971). Even though the number of radon-progeny measurements increased during the 1960's, the number per mine increased only slowly from about six in 1960 to almost 12 in 1968 (Figure E Annex 1-1). Measurements of radon progeny in a particular mine were never extensive and, more importantly, were not made on even a once per year basis in the majority of mines. Only 341 miners, about 10% in the Colorado Plateau miner cohort, had their exposure assignments based on measured radon-progeny concentration. For the majority of the miners, information on measured levels was combined with estimates made using a variety of methods as described by Lundin and others (1971). Many of the uranium miners were also employed as hardrock miners or previous to 1950 some had mined the same ore bodies, where uranium was found, for radium, vanadium etc., particularly in the Urivan Mineral Belt in Colorado. In the epidemiologic study, hardrock miners were assigned an exposure level of 1 WL for mining that occurred before 1935, 0.5 WL for 1935 through 1939, and 0.3 WL for later years (Lundin and others 1971). No information is given as to the basis of these estimates but a statement is included in Joint Monograph No. 1
OCR for page 309
Health Effects of Exposure to Radon: BEIR VI FIGURE E ANNEX 1-1 Frequency of radon-progeny measurements on the Colorado Plateau in two-year intervals 1950–1969. Source: Presentation to the committee of analysis of the data tapes for the Colorado Plateau miners by Duncan Thomas and Dan Stram, September 1995. (Lundin and others 1971) which indicates these estimates were thought to have been too high and that the average exposure was less. A re-evaluation of a sample of the Colorado Plateau cohort for exposure during hardrock mining is described in Monograph 1. This reassessment indicates that the tabulation of hard rock mining duration was subject to error and that misclassification of exposure was fairly common for that portion of a cohort member's work experience. For example, for a sample of 101 cases and 202 controls, misclassification was only about 10% for cumulative exposures of less than 20 WLM but 50% or more at higher levels. Nevertheless, hard rock mining may be a relatively unimportant source of exposure compared to the mining of uranium-bearing ores for which exposure levels were often much higher than 1 WLM. Because relatively few mines were initially monitored for radon or radon progeny, exposure estimates in uranium mines that occurred before 1951 were referred to as "guesstimates" in Joint Monograph No. 1 (Lundin and others 1971). According to that report, "guesstimates" were made on the basis of knowledge concerning ore bodies, ventilation practices, emission rates from different types of ores, and such radon or radon progeny measurements as were performed in 1951 and 1952.
OCR for page 310
Health Effects of Exposure to Radon: BEIR VI For mining that occurred after 1950, three other methods were used to estimate exposure levels. By far the most common was a process called area average estimation. This consisted of using the available, albeit often sparse, measured values to estimate concentrations in a given locality to obtain an "area average." In order to reduce sampling variability for these area averages it was required that three or more mines and ten or more samples had to be available for a locality in a year, otherwise the locality was assigned the average for the district in which it was located (Lundin and others 1971). If sufficient data for a district were not available, a state average was used or, in a few cases for which state data were insufficient, data for the state of Colorado were used. The degree to which area estimates were used to obtain exposure estimates is not often appreciated. Area estimates account for most of the exposure assignments throughout the study period of the Colorado Plateau cohort (Figure E Annex 1-2) Monograph 1 implies that when an individual mine was thought to differ appreciably from others in the same locality due to its ore quality or mining practices, guesstimation was substituted for an area average. To complete gaps in the measurements in calculating individual WLM estimates, a system of extrapolation, interpolation, and expert judgment was used to estimate the exposure in mines monitored less frequently than once a year. For mines with actual measurements at least once every five years, working-level estimates were obtained by interpolation, that is, averaging the measured values FIGURE E ANNEX 1-2 Bases for the assignment of exposure estimates by calendar year 1950–1969. Source: Presentation to the committee of analysis of the data tapes for the Colorado Plateau miners by Duncan Thomas and Dan Stram, September 1995.
OCR for page 311
Health Effects of Exposure to Radon: BEIR VI (Lundin and others 1971). Approximately 20% of the exposure assignments were made using this method (Figure E Annex 1-2). An assessment of this extrapolation procedure described in Monograph No. 1 indicates that it tended to overestimate exposures in the early years of mining but became more valid in the 1960s as information from more frequent measurements became available. For mines with yearly monitoring information available, the measured concentration was used to assign a worker's cumulative exposure in a given year. Table IV-3 in the BEIR IV Report (NRC 1988) indicates that the number of measurements per year per mine surveyed was usually between 10 and 20 after 1959 so that the measured values provide a reasonably stable estimate of the average working levels in those areas monitored. However, the average number per mine was somewhat less, 8–9 as illustrated in Figure E Annex 1-1. Although nearly 43,000 measurements were obtained (Lundin and others 1971), there were about 2,500 mines and measured concentrations were not a frequent method of exposure assignment. Figure E Annex 1-2 indicates that from 1959 to 1969 only 10–20% of the exposure assignments in a given year were based on direct measurement of radon progeny concentration and that even fewer were made on such direct information prior to 1959, when exposure levels were, on the whole, much higher. Assessment of Errors in the WLM Estimates A comparison of exposure estimates in relation to calendar year is given in Figure E Annex 1-3 for each assignment method. Except for 1950, estimates based on the extrapolation procedures are in reasonable agreement with those based on direct measurement while area average estimates tend to be somewhat greater than obtained by other methods. This may in part be due to measurements having been made more frequently in large mines having more employees and because of larger capital investment in better ventilation. The degree of variation in exposures among workers in a given mine was not well characterized. Before 1960 mechanical ventilation was not commonly used and a near equilibrium between radon and progeny was probably the rule under conditions of convective ventilation as indicated by the early data described by Holaday (1969). There appears to be no information on aerosol size distribution or even the unattached fraction in early mines. Even though diesel power was not common, compressed air or electricity was used to operate equipment including ore cars; dust was plentiful from drilling and hauling operations so that it is likely that the unattached fraction was low. An extensive study of air quality in nine uranium mines was carried out by the AEC Health and Safety Laboratory (HASL), now the DoD Environmental Monitoring Laboratory, in 1967–1968. Mines were selected by the U.S. Bureau of Mines to represent a cross section of the uranium mining industry (Breslin and others 1969). This investigation was in response to the concerns expressed at the
OCR for page 312
Health Effects of Exposure to Radon: BEIR VI FIGURE E ANNEX 1-3 Comparison of mean WL estimation by various methods in two year intervals 1950–1969. Source: Presentation to the committee of analysis of the data tapes for the Colorado Plateau miners by Duncan Thomas and Dan Stram, September 1995. Joint Committee on Atomic Energy hearings in 1967 in which the validity of exposure and early risk estimates of increased lung-cancer in miners were questioned. A particular point in question was "the extreme variation of atmospheric characteristics within a mine and among mines"; the HASL study was directed at exploring this question (Breslin and others 1969). The nine mines studied ranged in size from having two to 112 workers. Ore production varied from 150 to 11,000 tons per month. Mechanical ventilation rates varied from 5,600 to 100,000 cu. ft. per minute. Given this range of conditions, atmospheric conditions were surprisingly uniform, giving some credence to the validity of the estimation methods described above. In most of the mines the variation in radon-progeny concentration at different times and locations was only occasionally as large as a factor of two and 80% of the time had a coefficient of variation of 30% or less. The average WL ratio (pCi progeny to pCi radon) averaged 0.23 with a geometric standard deviation of 1.6 and showed limited variation with the absolute level of radon progeny. Equilibrium values F were also in a narrow range: about two-thirds were between 0.20 and 0.30; mode 0.25. Polonium-214 was most often at 16% of the equilibrium value, range 0.09 to 0.49. Simultaneous measurements of radon progeny were made at various locations in stopes (mining chambers) and in drifts (tunnels). While drifts showed
OCR for page 313
Health Effects of Exposure to Radon: BEIR VI greater variation than stopes, as indicated in Figure E Annex 1-4, the report's authors indicated that sampling location was not critical within a radius of 10 to 20 feet of the miners' location and that breathing-zone sampling was unnecessary. Similarly, the HASL study indicated that differences between various mining operations, for example, drilling, mucking, etc., had little effect on the measured working level (Figure E Annex 1-5). No measurements were taken immediately after blasting but such areas would not have been occupied because of other safety considerations. While the HASL study does indicate that mine-wide averaging probably provides a useful measure of worker exposure in the mines studies, this is probably less accurate for the high exposures which occurred before the introduction of mechanical ventilation in U.S. uranium mines. The recent report from SENES Consultants Limited provides additional relevant information. Tables for several mines demonstrate substantial variation in WL values within a mine during a single visit by an inspector, typically one day. For example, U.S. Bureau of Mines data for one Utah mine in 1968 showed variation from 0.4 to 5.4 WL across the mine (Table-E Annex 1-1). The Public Health Service investigators used self-reported mining history as the basis for estimating time spent underground in specific mines. This information was collected both retrospectively and prospectively during the annual miner censuses. The possibility of error in these histories has been acknowledged. The SENES report provides a series of case descriptions documenting inconsistencies in these histories and gives a compilation of exposure estimates for 78 miners for whom exposures have been calculated for both the epidemiologic study and for other purposes. Substantial variation is evident in these FIGURE E ANNEX 1-4 Variation of radon concentration with distance in ventilated uranium mine drifts on the Colorado Plateau (Breslin and others 1969).
OCR for page 314
Health Effects of Exposure to Radon: BEIR VI FIGURE E ANNEX 1-5 Variation of radon concentration with distance in ventilated uranium mine stopes on the Colorado Plateau (Breslin and others 1969). estimates, largely reflecting various discrepancies in the alternative work histories used for the purpose of estimating the exposure. New Mexico Uranium Miners Large-scale uranium mining began in the early 1950s (see Time Line E-2) with the opening of the Jackpile mine, an open-pit mine. By the late 1950s, a number of large mines were operating at Ambrosia Lake and the Churchrock mining district became active in the late 1970s. The industry continued operating into the early 1990s, longer than in other U.S. locations, so that miners working after 1968 have individual exposure records (work location estimates and estimates of exposure) for this period of employment. These were calculated based on area measurements and work locations. For the most part, post-1968 employment was in very large industrial operations with state of the art ventilation. Mean annual exposures in 1968 were about 3.8 WLM and declined to 1.2 WLM or less by 1972 (Samet and others 1986b). Earlier exposures were not estimated as accurately, although the State Health Department and the State Mine Inspector had implemented active measurement programs by the late 1950s. The state implemented a progressively more stringent series of shut-down concentrations. As for the Colorado Plateau miners (see above), median annual exposures were considerably larger during the earlier years of the industry, about 30 WLM in the 1960's. Some members of the New Mexico cohort, who had also mined in the Colorado Plateau, had annual exposures as high as 300 WLM or more (Samet and others 1991).
OCR for page 315
Health Effects of Exposure to Radon: BEIR VI Investigators directed substantial effort at tracing employment histories for the purpose of estimating the cumulative exposures for those employed before exposure estimates were individualized (Samet and others 1991). The miners' underground employment and exposures in specific mines were traced by examining company personnel records and self-reported work histories taken at the time of periodic medical examinations. Estimated exposures for miners who had worked underground on the Colorado Plateau were supplied by the USPHS (Lundin and others 1971; Samet and others 1991). Contributions to the total mean exposure from various information sources are listed in Table E Annex 1-2 (Samet and others 1991). With the notable exception of those members of the work force employed on the Colorado Plateau, this cohort probably has maintained the most extensively documented exposure estimates. In this regard, it should be noted that the state of New Mexico had more extensive and more frequent monitoring for radon then was common elsewhere in the early 1950's when exposures were very high (Lundin and others 1971). From 1957 to 1967 exposure estimates are based on 20,086 measurements taken during 1,886 visits. Most annual exposures were relatively low during this period, mean 4-5 WLM per year, so that this cohort has a large sub-cohort of miners exposed at low rates and relatively low cumulative exposures. Beaverlodge Uranium Miners The BEIR-IV report also includes a description of the exposure estimates for this cohort (NRC 1988). Exploratory uranium mining at Beaverlodge,
OCR for page 316
Health Effects of Exposure to Radon: BEIR VI TABLE E ANNEX 1-1 U.S. Bureau of Mines February 1968 survey at North Alice Mine, Utaha No. of Men Location, Operation Estimated Average Full Shift Exposure to Radon Daughtersb (WLc) 2 men, night shift 416 NE from 360 NW; mining 0.5 2 men, day shift 236 from 325 S; mining 1.7 1 man night shift 1 man day shift 240 W incline station and hoist 1.0 1 man night shift 1 man day shift 248 SE from 225 N; mining 5.4 2 men, night shift 242 S from 190 W; mining 3.6 1 man day shift 1 man night shift 240 W incline to main incline; tramming 2.2 2 men, night shift 100 S area; mining 0.6 2 men, day shift 147 N from 130 E; mining 2.8 3 men, day shift 2 men, night shift 128 S from 145 E; mining 1.5 1 man day shift 1 man night shift main incline; trip rider 0.4 1 man day shift all areas; electrician 1.4 2 men all areas; mechanics 1.6 5 men all areas; shift bosses 1.4 3 men all areas; staff 1.0 3 men all areas; bratticemen 2.6 a This table is from a February 1968 report on a Radiation Survey prepared by U.S. Bureau of Mines, obtained from SENES 1995. b Average Levels are estimated from information gained by questioning the miners about where there time is spent and weighing the radon daughter concentrations in each place by the time spent in that place. c NIOSH database: 1967 WL is 1.3, based on 39 measurements; 1968 WL is 3.3, based on 120 measurements. Saskatchewan started in 1949 and commercial production began with a greatly expanded labor force in 1953 (see Time Line E-3). Radon monitoring was carried out in 1954 and 1956 but only sporadically until the end of 1961. A number of radon-progeny measurements were also made at this time but monitoring was mostly for radon and viewed as a check on ventilation rather than as a tool for exposure control. Nevertheless, the frequency of radon-progeny measurements increased and by 1961 exposure records were maintained for all full-time underground employees. These records listed each worker's occupancy time at each work place on a daily basis. In 1970 worker's exposure records were estimated retrospectively to 1 November 1966 and in 1971 part-time underground workers were included in the exposure assessment (SENES 1989).
OCR for page 568
OCR for page 569
OCR for page 570
OCR for page 571
OCR for page 572
OCR for page 573
OCR for page 574
OCR for page 575
OCR for page 576
OCR for page 577
OCR for page 578