S. Lacey, N. Esmen, J. Palmer, K. Kennedy, University of Illinois at Chicago, Chicago, IL.
Published exposure data in conjunction with site-specific industrial time studies were used to generate exposure distributions for soluble and synthetic metalworking fluids (MWF) for a milling operation in a turbine engine manufacturing facility. The estimated exposure distributions were generated by Monte Carlo simulation, and the results were validated by measured exposure levels at the facility studied. The specific average task times for milling, such as part handling, setup, machining, and idle time, were obtained from the company’s industrial engineering time study data. These were converted into fractions of time denoted as unexposed, exposed to background level, and directly exposed to the contaminant generated. The time fractions were weighting factors for the associated source intensity distributions in a 1000-sample Monte Carlo simulation used to calculate the personal exposure for a milling job as defined by a company job description. For this operation, the exposure levels to synthetic MWF were: range = 10.55-2927.43; mean = 179.28; median = 111.31; and 95% confidence interval (CI) = 25.92-770.92 μg/m3. The values for exposure to soluble MWF were: range = 68.4-1703.49; mean = 401.92; median = 346.77; and 95% CI = 138.08-962.96 μg/m3. These generated distributions compared well to the measured exposure levels, as the measured values fell within the central portions of the distributions. The results of the study suggest that when site-specific data is lacking, a Monte Carlo simulation of exposure distribution using available or estimated task time distribution and source intensity is an efficient way to reconstruct exposure levels. In addition, this method is naturally extended to interpolation or extrapolation of existing data to similar conditions without available measurement. This can be done by modifying the task time or source intensity distribution, or both, in the simulation.
R. Hancock, N. Esmen, S. Lacey, K. Kennedy, University of Illinois at Chicago, Chicago, IL; G. Marsh, J. Buchanich, University of Pittsburgh, Pittsburgh, PA; T. Hall, University of Oklahoma, Oklahoma City, OK.
Surface grinding is an important machining operation for shaping difficult-to-machine alloys. This operation is used to machine gas turbine vanes and blades. In grinding operations, localized grinding zone temperatures can reach as much as 1000ºC. This temperature can result in aerosols produced by incomplete combustion (CP) and by recondensation (RP) of the metalworking fluid (MWF). Aerosols are also generated by the interaction of the MWF with the spinning grinding wheel. Through examination of standard operating procedures and company work practice documents as well as worker and engineer interviews, a semiqualitative exposure model was constructed. The exposure estimates suggest that the historically reported mass measurements and some respirable dust samples, or theoretical apportioning the inhaled aerosols, does not indicate that the CP+RP aerosol would make a significant contribution to the mass concentration measurements. This observation is problematic, because the health effects of ultrafine aerosols and coarse aerosols differ and chemical characteristics of the ultrafine aerosol is significantly different. While we used the existing samples to validate the coarse aerosol exposure estimates, the ultrafine aerosol exposure could only be indirectly validated by anecdotally reported visual observations and brief partial combustion events. The model is based on the aerosolization of MWF by a spinning wheel and uses empirical correlations of wheel diameter, rotational speed, and liquid feed rate with aerosol mass generation reported in several articles. The combustion, quenching, and recondensation aerosol is also modeled by using semiempirical correlations from the literature. The exposure model and the calculated results presented treat the coarse/respirable mechanically generated component and the ultrafine component separately. The coarse aerosol and respirable aerosol are expressed as mass measurements and compare fairly well with the existing historic observations. The respirable fraction represents about 30% of the total mass measurements scaled by inhalability. The ultrafine exposure estimates (qualitatively confirmed) are expressed as number concentrations.
S. Viet, Westat, Steamboat Springs, CO; T. Carreón, K. Hanley, A. Ruder, NIOSH, Cincinnati, OH.
In the late 1980s, NIOSH conducted a retrospective cohort study of bladder cancer among 1749 workers employed at a rubber chemical manufacturing plant. These workers were exposed to ortho-toluidine, a suspected bladder carcinogen in humans. Thirteen cases of bladder cancer were observed vs. the 3.61 expected. In the intervening years, additional cases of bladder cancer have been diagnosed in the cohort. NIOSH is re-evaluating cancer mortality and incidence in the cohort using a revised exposure assessment characterization. This study will also include eligible workers who may not have been included in the original cohort. The new exposure assessment incorporates both department and job title into the characterization. The site was revisited, workers were interviewed, and more detailed work histories were identified and abstracted. Available exposure monitoring data for aniline, ortho-toluidine, and nitrobenzene from 1975 to the present were used to develop the new characterization. A relative exposure ranking scheme was developed so that duration of exposure by time period (era) could be incorporated into the assessment.
R. Rando, Tulane University School, New Orleans, LA.
A typical goal for exposure assessment in support of prospective epidemiologic research is the estimation, with good precision and accuracy, of measures of central tendency of exposure, most often the arithmetic or geometric mean. Because most occupational exposures are best described as log-normal distributions of exposure data, which exhibit variability that is inherently geometric in nature, very large sample sizes are necessary for estimation of mean with acceptable precision. For example, a sample size in the neighborhood of 300 may be needed to estimate the arithmetic mean to within ±20% at the 95% confidence level for a data distribution that has a geometric standard deviation of 3. Given that there may be many similar exposure groups in a population under study — each with its own individual large sample size requirement — the resulting overall exposure measurement effort may be overwhelming and impractical within the constraints of a project’s time and budget. Thus there is a need for alternative strategies to simply collecting and analyzing vast numbers of exposure samples. Several approaches deserve consideration: (1) optimizing target precision levels, (2) using composite or long-term integrated sampling, (3) balancing cost/benefit in choosing the monitoring technology, and (4) using physical-chemical and statistical modeling of exposures. Examples from epidemiologic studies conducted in various industries, including wood processing, polyurethane foam manufacture, and industrial sand production, are used to illustrate the nature of the large sample size problem and to present the considerations, procedures, and relative merits of the proposed alternative strategies for exposure assessment.
C. Peters, H. Davies, P. Demers, University of British Columbia, Vancouver, BC, Canada.
The most important occupational lung disease for sawmill workers is asthma. The purposes of this study were to use a novel case definition for asthma using administrative data (physician visits) and to assess the effect of wood-dust exposure on asthma. The cohort consisted of 4856 workers employed for at least one year in one of 14 British Columbia sawmills between 1991 and 1998. Cases had a proxy physician diagnosis of asthma (two visits assigned an (ICD) codes ICD [International Classification of Diseases]-9 code of 493 [asthma] in a sliding 365-day window), or a hospital discharge record with the same code as the primary diagnosis. Two hundred and twelve cases were identified. Wood dust exposure estimates were calculated retrospectively using regression models based on measurements and expert opinion and were externally validated. Work history information was collected from the mills. The NIOSH life table analysis system (LTAS) was used to stratify person-days at risk above an exposure of 1.0 mg/m3, followed by an internal comparison of cumulative wood dust exposure and asthma using Poisson regression. Exposure to wood dust above a level of 1.0 mg/m3 was found to increase the risk for asthma the longer one was exposed. For exposure of one year or longer, the RR was 1.19 (0.78-1.81), and for exposure between one week and one year, the rank regression was 1.58 (0.95-2.62) (when compared to exposures lasting less than one week). Caucasians had significantly lower risk of developing asthma than other races. The results are difficult to interpret fully. We expected that asthma-susceptible workers would stop working sooner at workplaces with high levels of wood dust , and we do see an apparent survivor bias, with those exposed for the longest showing less risk of asthma than those in the midrange. Further investigation into different exposure metrics is under way.
W. Chen, J. Chen, R. Chen, School of Public Health, Tongji Medical College in Huazhong University of Science and Technology, Wuhan, China; W. Wallace, E. Hnizdo, NIOSH, Morgantown, WV; F. Bochmann,
Institute for Occupational Health of HVBG, Sankt Augustin, Germany.
About 60,000 pneumoconiosis cases were reported in China in 2005. About 10,000 new pneumoconiosis cases in China were diagnosed annually in the past several years. This research was designed to analyze the relationship between silica dust exposure and the risk of silicosis. A total of 4500 workers from 20 mines andfactories with silica dust exposure were selected to construct a cohort, which was followed through 2003. Chest radiographs were taken every three to five years for dust-exposed workers even if they retired. Diagnoses of silicosis were based on Chinese pneumoconiosis roentgen diagnostic criteria and comparison with International Labor Organization criteria. Historical environmental dust monitoring began in 1950; 22,890 historical estimates were developed for 30 calendar-year periods (to 2003), using 2.3 million monitoring data points. Then a facility, job title, calendar-year exposure matrix was created. The results showed that malignant neoplasm, respiratory diseses, cerebrovascular disease, and cardiovascular disease were the main diseases that threaten silica dust-exposed workers’ health and accounted for about 60-70% of all deaths. A clear exposure-response relationship was detected for silicosis and cumulative silica dust exposure. The characteristics or type of silica dust may affect the risk of silicosis. Risk of lung cancer was increased among silicotic subjects in some mines but not in others. The efficiency of pneumoconiosis prevention through dust-control operations, which greatly reduce the concentration of dust, was also evaluated.