Annals of Occupational Hygiene Advance Access originally published online on July 24, 2008
Annals of Occupational Hygiene 2008 52(7):623-633; doi:10.1093/annhyg/men046
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Quantitative and Qualitative Assessment of Exposure among Employees in Norwegian Smelters
altyt
Benth5,2
1 Department of Medicine, Akershus University Hospital, PO Box 75, N-1478 Lørenskog, Norway
2 Faculty of medicine, University of Oslo, Oslo, Norway
3 National Institute of Occupational Health, Oslo, Norway
4 Eurofins Norway, Oslo, Norway
5 Helse Sør-Øst Health Services Research Centre, Lørenskog, Norway
6 Department of Respiratory Medicine, Rikshospitalet Medical Centre, Oslo, Norway
* Author to whom correspondence should be addressed. Tel: +0047-92042621; fax: +0047-62590912; e-mail: helle.laier{at}dadlnet.dk
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONFUNDINGACKNOWLEDGEMENTSREFERENCES |
|---|
Objectives: To generate a job exposure matrix (JEM) for dust exposure in Norwegian smelters to be used in an epidemiologic study of respiratory diseases and to identify determinants of exposure.
Methods: The arithmetic mean and geometric mean (GM) of 2619 personal dust exposure measurements were applied in constructing the JEM, which was assigned to 2620 employees participating in a respiratory survey including yearly spirometry and a respiratory questionnaire. A qualitative exposure classification was constructed: (i) line operators were those employed full time in the production line, (ii) non-exposed employees were those who did not work in production and (iii) the remainder were classified as non-line operators.
Results: In the ferrosilicon alloy and silicon metal production group (FeSi/Si-metal), the median GM of dust exposure was 2.3 mg m–3 (0.04–5.6) (10–90% percentiles) compared with 1.6 mg m–3 (0.02–2.3) in the silicomanganese, ferromanganese and ferrochromium production group (SiMn/FeMn/FeCr). Multivariate analyses showed that dust exposure concentration levels decreased significantly with increasing age (FeSi/Si-metal), was significantly lower in females than in males and was significantly higher in current smokers than in never-smokers. Dust exposure concentration levels were also higher in employees reporting previous exposure to dust, fumes and gases than in employees without such previous exposure, though, significant only in the FeSi/Si-metal production group.
Conclusion: The dust exposure levels of the employees were higher in the FeSi/Si-metal production group than in the SiMn/FeMn/FeCr production group. Age, gender, smoking status and previous exposure were significant determinants of dust exposure and should be evaluated in future analyses of the relationship between health outcomes and dust exposure in this industry.
Keywords: job exposure matrix • longitudinal study • qualitative exposure classification • smelting industry • total dust exposure
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONFUNDINGACKNOWLEDGEMENTSREFERENCES |
|---|
In epidemiological studies, various exposure indices have been used, including duration of exposure or employment, qualitative expert-based classifications of employees, employee work histories and job exposure matrices (JEM) based on measurements of specific exposure agents (Stewart et al., 1991; Goldberg et al., 1993; Blanc et al., 2004). The advantages and limitations of different methods of exposure classification depend on the availability of data and the study design (Stewart et al., 1991). In community studies, exposure assessed by an expert panel has been found to be preferable (Rybicki et al., 1997; Benke et al., 2001; de Vocht et al., 2005). In industry-based studies, the ideal approach is quantitative measurements of exposure for each of the study subjects (Stewart et al., 1996; Benke et al., 2000; Checkoway et al., 2004). This ideal goal is hardly ever accomplished, however, and an estimation of exposure has to be made (Goldberg et al., 1993; Seixas and Checkoway, 1995).
In constructing a JEM, the first step is to identify the exposure of interest (Seixas and Checkoway, 1995; Stewart et al., 1996). In previous studies, dust has been found to be a predictor of lung function impairment (Kauffmann et al., 1982; Becklake, 1989; Viegi and Di, 2002; Trupin et al., 2003).
In 1996, the Norwegian smelting industry initiated a longitudinal respiratory study (Soyseth et al., 2007). A quantitative JEM for all job groups in this industry was not available. The objectives of the present study were to generate a JEM for dust exposure for an epidemiologic respiratory study in Norwegian smelters, to compare this with a qualitative exposure classification and to identify determinants of exposure.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONFUNDINGACKNOWLEDGEMENTSREFERENCES |
|---|
Materials
Between 1996 and 2003, a total of 4234 industrial hygiene personal dust exposure measurements were carried out in 15 smelters, all members of the Norwegian Federation for Process Industry in 1996. During the period 1997–2001, all employees in these smelters were invited to participate in a respiratory survey with annual health examinations. At each examination, the 2620 participants, aged 20–55 years at inclusion, completed a standardized respiratory questionnaire including smoking habits and current job function.
The study was approved by the Regional Committee for Medical Research Ethics, Eastern Norway.
Production
The smelters and related workplaces serving the smelters were divided into two production groups: (i) ferrosilicon alloys (FeSi) and silicon metal (Si-metal) and (ii) other ferroalloys such as silicomanganese (SiMn), ferromanganese (FeMn) and ferrochromium (FeCr).
The production of metallic alloys in these smelters uses high temperature processes, with raw materials transported into the plant to be fed into a smelting furnace. The production requires carbon (such as coke, coal and in some cases charcoal and wood chips) in a solid form to reduce the minerals to molten metals, and a direct supply of electric power to achieve the necessary high process temperature. Electrical power is supplied through three submerged carbon electrodes. Three main types of electrodes are used: Søderberg electrodes, pre-baked electrodes or electrodes combining the characteristics of the other two types of electrodes, depending on the process. The Søderberg electrodes are self-baking carbon electrodes covered with an iron or steel casing, while pre-baked electrodes are baked before they are used in the smelting process.
When tapped from the furnace, the molten Si-metal or ferroalloys are poured out to cool in moulds and then finally crushed to specified sizes. In other production processes, some end products are sized by granulation. Dust is emitted into the working atmosphere during raw material handling, smelting, tapping (condensation of tapping fumes), crushing and handling of the end products. Job tasks by job title and department in the smelters are described in Table 1.
|
FeSi alloys and Si-metal production.
The reduction materials are mixed with quartz and iron sources or other compounds depending on the end products. The raw materials are charged into the top of the cylindrical furnaces, which have a diameter of 5–13 m. The smelting temperature at the centre of the furnaces is typically between 1500 and 2000°C. The furnaces are partly open (Zulehner, 1993; Neuer and Rau,1993).
FeMn, SiMn and FeCr alloys production.
Depending on the desired end product, the reduction materials are mixed with manganese ore or chromium ore, iron sources or other compounds such as quartz. In the sinter plants of FeMn and FeCr production, fine grain raw material ores are sintered into coarser materials. The temperature at the centre of the furnace reaches up to 1600°C (Wellbeloved et al., 1990; Fichte, 1986). In contrast to FeSi and Si-metal production, where the furnaces are semi-closed or open-air furnaces, the furnaces in FeMn, SiMn and FeCr alloy production are closed.
In Norway, closed furnaces have wet scrubbers, scrubbing the furnace gas (with water) to get hold of the particulates. Semi-closed or open-air furnaces use ‘dry filter bag technology abatement’ to trap the condensed particles from the furnace gas. Thus, differences in the overall dust and gas exposure levels of the furnace house operators in the two production groups may exist.
Quantitative exposure classification and construction of JEM
More than 70% of the industrial hygiene dust exposure measurements in FeSi/Si-metal production and in the FeMn/SiMn alloy production were part of investigations performed by the National Institute of Occupational Health (NIOH) and were made available to our study. The measurements were performed randomly in accordance with the recommendations given by the Norwegian Labour Inspection Authority. The remaining dust exposure data originate from routine sampling programmes in the smelters and were analysed by three different laboratories serving the smelters.
Sampling and calculation of exposure estimates.
Dust (so-called ‘total dust’) was collected at a sampling rate of 2 l min–1 on mixed cellulose filters (AAWP, Millipore Corporation, MA, USA) with an 0.8-µm pore size, fitted in 25 or 37 mm closed-faced three-part plastic cassettes (MP cassettes). The particle mass was measured using a microbalance (Sartorius AS, Goettingen, Germany) with a detection limit of 0.06 mg.
The MP cassette used in this study has been widely used for sampling so-called total dust. It seems that the sampling efficiency of particulates (aerosols) for this cassette is closer related to the thoracic fraction than to the inhalable fraction (CEN-convention: NS-EN 481, 1993). However, for particulates with aerodynamic diameters >15 µm, this cassette overestimates the thoracic fraction (Vincent, 1995; Kenny et al., 1997).
The dust concentration measurements assembled were assessed as representative for the whole-study period as the measurements were performed randomly during the period and only minor changes in production and abatement technology were introduced. This assumption was supported by mixed-model analyses performed to assess if a time trend existed for the different job titles. A significant time trend was only found regarding two job titles in two smelters. When examined, it became clear that the time trend regarding these job titles was based on <10 measurements.
Of the 4234 personal dust exposure measurements performed, only samplings by MP cassette were used for the development of the JEM. Measurements made by MP cassettes were available in 13 smelters (NMP cassette = 2680). Of the remaining 1554 personal measurements, 1497 were performed by Institute of Occupational Medicine (IOM) samplers (IOM, Edinburgh, UK) as part of a study comparing results obtained with MP cassettes and IOM samplers. As such, the measurements performed by IOM samplers were taken from the same individuals and at the same time as the measurements performed by MP cassettes.
Samples with a dust concentration level in excess of 50 mg m–3 were excluded as they were considered invalid due to sampling errors, as assessed by the industrial hygienist (S.M.H.), who has conducted exposure measurements projects in both the FeSi/Si-metal and the SiMn/FeMn/FeCr production groups and has extensive knowledge of the Norwegian smelting industry [number of excluded samples: n = 20, range 50–1905 mg m–3, standard deviation (SD) 415 mg m–3]. The 20 measurements with concentrations in excess of 50 mg m–3 were randomly distributed between the smelters and originated from eight of the 15 smelters encompassing 12 different job codes.
The average length of a shift during the study period was 480 min. Measurements with a sampling period of <240 min were excluded (n = 41, range 0.21–94 mg m–3, SD 18 mg m–3). Of the included measurements, 86% were recorded either as ‘full shift’ measurements or had a duration of 
420 min. As such, the included industrial hygiene measurements were considered by the industrial hygienist (S.M.H.) to be the representative for the whole work shift and were not transformed into 8-h time-weighted averages.
If the measured personal dust exposure concentrations were less than the detection limit (3.5% of the measurements), the results were substituted by a concentration level equal half of the detection limit.
Finally, the data set used for construction of the JEM consisted of 2619 personal dust exposure measurements.
The arithmetic mean (AM) and geometric mean (GM) dust concentration level was assigned to the corresponding exposure group (smelter/department/job title) when five or more measurements were available. When less than five dust measurements were available for a given exposure group, the group was assigned the AM and GM dust exposure level of the respective job title in all smelters of the corresponding production group (FeSi/Si-metal or SiMn/FeMn/FeCr).
Employees in the administration department, who were regarded as non-exposed (‘office work only’ in Table 1), were assigned 1% of the AM and GM dust exposure concentration of all departments (exclusive electrode and refractory departments) of the corresponding smelter. Employees regarded as ‘partly exposed', such as administrative personnel with part time supervision in the production, were assigned 10% of the AM and GM dust exposure concentration of the smelter. As only one personal dust exposure measurement existed for the job title ‘maintenance cleaner', the dust exposure was assessed as half the exposure of the smelter (exclusive the electrode and refractory department).
In 10 out of 15 smelters, we were not able to differentiate between tappers, furnace operators and other job functions held by the operators in the furnace house section due to lacking specificity of the work histories. In these smelters, a new job title ‘furnace section worker’ was created and their dust exposure concentration estimated as the AM and GM of all the included dust measurements of the corresponding furnace house.
Exposure groups for the JEM.
An exposure group was defined by a unique combination of smelter, department and job title. We used a classification system of job titles and departments developed by the smelting industry. The matrix included 15 smelters, with each smelter divided into departments (5–10 departments per smelter) encompassing 49 different job titles (6–16 job titles per smelter). As the job titles were unique for the different departments, this resulted in 222 unique groups of exposed workers (smelter/department/job title). In Table 1, 10 of the departments are shown, the remaining four departments were departments only associated with one smelter. In the same way, 19 of the 49 different job titles are shown. Even if not shown in Table 1, the specified job titles were used in the analyses.
Allocation of exposure for the employees.
By each health examination, up to three job titles could be recorded for each of the 2620 employees participating in the respiratory survey. This resulted in 13 166 individual registrations during the 11 335 health examinations performed in the study period. The 222 unique combinations of smelter, department and job titles with a specific dust exposure concentration level were assigned to the employees as follows: where an employee had held more than one job in the 12 months prior to the health examination, the AM and GM of dust exposure for the employee were calculated weighted by the time spent in each of the jobs (with a maximum of three job titles). For time periods of no employment in the industry, i.e. employees on leave, an exposure of zero was assigned to the employees.
Qualitative exposure classification
The qualitative exposure classification of the employees was based on their job functions in the year before each health examination. This classification was used for cross-sectional and longitudinal analyses performed before the quantitative exposure classification was available (Johnsen et al., 2007; Soyseth et al., 2007; Johnsen et al., 2008). The qualitative exposure classification has been thoroughly described in a former paper (Johnsen et al., 2007). Briefly, the employees were classified into three exposure categories: (i) ‘line operators’ were those working full-time on the production line with handling and mixing of raw materials before charging the furnaces, all full-time jobs in the furnace house and crushing, screening and packing of end products; (ii) ‘non-line operators’ included employees loading and unloading raw materials and end products outside the plant, and employees working part time on the production line, such as foremen, maintenance and laboratory workers; (iii) ‘non-exposed employees’ were primarily full-time office staff. Classification of each of the reported job titles into one of the three categories was made blinded in regard of all other information obtained from the respiratory questionnaire, such as health outcomes, smoking habits and previous exposure.
Data analyses
First multivariate mixed-model analyses were carried out to evaluate the differences in dust measurements between production group, departments and job titles and to explore if a time trend existed regarding the dust exposure levels. Second, the measured dust concentration levels were used to calculate the AM and GM of dust exposure in all 222 unique exposure groups. Based on these estimates, each employee was assigned a dust exposure concentration value as outlined above. As up to three job titles could be registered for each employee per year, the dust exposure concentration values were assigned time weighted to the employees. The medians of the GMs for each job across all of the workplaces in each production group were calculated. Third, univariate analyses were performed to investigate the association between dust exposure concentration levels and individual characteristics of the employees (data from the questionnaires used in the respiratory survey) (Fitzmaurice et al., 2004). Fourth, the association between dust exposure concentration levels and individual characteristics was analysed using a multivariate linear mixed model. The strategy for model selection has been described elsewhere (Soyseth et al., 2007). The Akaike Information Criterion was used for model selection (Sherrill and Viegi, 1996).
The following interaction terms were included in the initial models: line and non-line operator x production group, line and non-line operator x current smoking and line and non-line operator x former smoking.
In order to adjust for differences between the smelters, a dummy variable for each smelter was included in the model.
The analyses were performed using the Statistical Package for the Social Sciences (SPSS Inc., Chicago, IL, USA; version 12.0.1) and SAS PROC MIXED (SAS Institute Inc., Cary, NC, USA; SAS version 9.1).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONFUNDINGACKNOWLEDGEMENTSREFERENCES |
|---|
The distributions of the AM and GM of measured dust in the two production groups by department for the period 1996–2003 are shown in Table 2. The number of measurements was highest in the FeSi/Si-metal production group. This represented the largest production group, encompassing 11 smelters (nemployees = 1697), compared with four smelters (nemployees = 933) in the SiMn/FeMn/FeCr production group. The highest total dust concentration level was found for the sinter plant of the SiMn/FeMn/FeCr production group.
|
Table 3 shows the results of the multivariate mixed-model analyses, which were performed to evaluate the difference in measured dust exposure concentration levels for the production groups and for the different job titles. Job titles that were found in more than one smelter are shown in the table. The results showed no overall significant time trend for the measured dust exposure data. A significant difference between the two production groups, FeSi/Si-metal and SiMn/FeMn/FeCr, was found. Except for logistic workers, furnace operators, laboratory department workers, electricians and wet filtering workers all the job categories listed in Table 3 had higher dust exposure levels than transport operators.
|
An exact link between the job title of the employees (smelter/department/job title), collected during the respiratory survey, and the JEM exposure group was achieved for 4454 individual registrations (33.8% of all registrations). In 3881 individual registrations (29.5%), exposure was based on dust exposure concentration of the corresponding department. Due to inability to make a direct link, exposure in 1687 registrations (12.8%) was based on the dust exposure concentration of the respective job title or department in the production group and in 25 registrations (0.2%) the dust concentration of the respective job title or department in all the smelters (both FeSi/Si-metal and SiMn/FeMn/FeCr production) was used. The remaining 3127 registrations (23.8%) represented non-exposed and partly exposed employees as well as retired employees and employees on leave.
In Table 4, the estimated dust exposure of the job titles in 10 of the departments is presented for the two production groups. Overall dust exposure was found to be highest in the FeSi/Si-metal production group. The highest dust exposure concentration levels were found among electrode and refractory workers of the FeSi/Si-metal production group and among sinter plant workers of the SiMn/FeMn/FeCr production group. Because of the low number of workers in these jobs and as such the short total time spent here during the study, their contribution to the total exposure of the work force was modest. The number of smelters represented by the measurements is shown in Table 4.
|
Table 5 presents the univariate association between the dust exposure and the individual characteristics of the employees in the two production groups, using the time-weighted exposure estimates for the employees.
|
The time-weighted median GM of dust exposure for line operators, non-line operators and non-exposed employees in the study population was 2.6 mg m–3 (1.4–6.2) (10–90% percentiles), 1.1 mg m–3 (0.19–3.5) and 0.018 mg m–3 (0.013–0.054), respectively (data not shown).
In the multivariate analyses, the interaction terms between both line operators and non-line operators and the production groups were significant in the model where both production groups were included. Thus, the multivariate analyses were performed in separate models for each of the two production groups (Table 6).
|
Age was found to be negatively associated with dust exposure in all the models, indicating that the youngest employees had the highest exposure to dust (Table 6). This association was nevertheless not significant regarding the SiMn/FeMn/FeCr production group. Females were found to be less exposed to dust than males. Employees reporting previous exposure to dust, fumes or gases had higher dust exposure than employees not reporting such previous exposure. This association was significant only in the FeSi/Si-metal production group. The multivariate analyses also showed that the apparent decline in dust exposure concentration levels of the employees during the study (time in study) was significant for SiMn/FeMn/FeCr production only. In the SiMn/FeMn/FeCr production group, former and current smokers were found to have higher dust exposure than those who had never smoked.
The mixed-model analyses showed that the dust exposure was higher in non-line operators and line operators compared with non-exposed employees in the SiMn/FeMn/FeCr production group. As the dust exposure of non-exposed employees was computed as a preset percentage of the measured values, it is not meaningful to test these differences. The variables non-exposed employee, non-line operator and line operator were nevertheless included in the multivariate analyses because of confounding. To test for the difference between non-line operator and line operator, we performed the multivariate analyses excluding non-exposed employees and non-line operators with dust exposure estimated as 10% of the dust exposure of the smelter. These analyses showed that line operators had significantly higher dust exposure levels than non-line operators in both the FeSi/Si-metal and the SiMn/FeMn/FeCr production groups (results not shown).
In the FeSi/Si-metal production group, the interaction terms between both non-line operators and line operators and current smoking were found to be positive and significant. Thus, for this production group, the multivariate analyses were performed separately for current smokers and never-smokers. For currently smoking line operators, the GM of dust exposure concentration was 3.5 mg m–3 above the GM of dust exposure concentration for non-exposed employees, whereas the GM of dust exposure concentration of line operators who had never smoked was 3.2 mg m–3 above that of non-exposed employees. For non-line operators, the results for current smokers and never smokers were 1.7 and 0.71 mg m–3, respectively.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONFUNDINGACKNOWLEDGEMENTSREFERENCES |
|---|
Personal dust exposure concentration levels were generally higher in the FeSi/Si-metal production group than in the SiMn/FeMn/FeCr production group. Gender, age, current smoking, job categories and previous exposure were found to be important individual characteristics of the dust exposure concentration levels of the employees in these industries.
Methodological considerations
Classification of exposure in epidemiologic studies can be made using different methods (Stewart et al., 1991; Checkoway et al., 2004). In our study, a qualitative exposure classification and a JEM based on personal dust exposure measurements were constructed.
A JEM based on exposure measurements may have several limitations due to variation in measurements and misclassification (Seixas and Checkoway, 1995). First, the samples may not have been randomly collected, and as such the exposure assessment may be biased (Goldberg et al., 1993; Stewart et al., 1996). Furthermore, within each exposure group a considerable random variation may exist between different workers and from one day to another (Seixas and Checkoway, 1995). Moreover, job tasks carried out by workers with identical job titles may vary not only between smelters but also between individuals at the same smelter and of different gender, or job tasks may be performed with different frequencies or under different conditions (Goldberg et al., 1993; Messing et al., 1994; Seixas and Checkoway, 1995; Benke et al., 2000). Thus, within each job category, there may be wide variation in the level of dust exposure, which may introduce systematic as well as random errors of the estimates.
In the present study, >70% of the dust exposure measurements were part of investigations performed by NIOH and were randomly collected. The remaining samples were part of the routine sampling programmes in the smelters and at least some of these samples might have been collected for compliance. Undoubtedly, an inter-person and between-person variability were present in our study as in other industry-specific studies. The job tasks carried out by workers with identical job titles varied between the smelters, leading to misclassification when the mean GM dust exposure for a given job title in one of the production groups was assigned to employees holding this job in smelters without dust measurements for the actual job title. Females were assigned the same dust exposure levels for a given job title as men, leading to a probably misclassification as females have been found to be less exposed than men within identical job titles (Messing et al., 1994). As such, also the present study suffers from misclassification which may alter the results when the JEM is used in analyses of the association between exposure and health effects. In epidemiological studies, exposure misclassification is typically thought to be non-differential because exposure assessment is made independent of the health outcome (Blair et al., 2007). As such, the present misclassification would lead to a weakening of the association between occupational dust exposure and health outcome in the epidemiological analyses (Goldberg et al., 1993). As most of the exposure assessment in the FeSi/Si-metal production group was based on measurements from a smaller proportion of the smelters in the production group, than was the case in the SiMn/FeMn/FeCr production group, one could expect misclassification to be greatest in the former group.
The JEM was constructed for the period 1996 to 2003. No time trends of the dust exposure levels were found during this period. This was in accordance with the fact that only minor changes in production and workplace dust abatement technologies were introduced during the period. As such, the decline in the dust exposure levels of the employees in the SiMn/FeMn/FeCr production group during the study period could be explained by a decrease in the number of highest exposed workplaces in favour of less exposed workplaces.
An exact link between the job title of the employees from the respiratory survey and the exposure group of the JEM (job title or department in a given smelter) was achieved for >80% of the registrations. In the remaining cases, estimates of dust exposure were constructed using the AM and GM of the dust concentration levels in the corresponding production group. This method of using broader exposure groups when the exposure measurement data become scarce resembles the method used by Seixas et al. (1991) in their estimation of exposure for the national study of coal workers pneumonconiosis in US. Even when the above-mentioned estimates in the present study were based on recommendations of a skilled industrial hygienist (S.M.H.), this approach is likely to produce misclassification as the job tasks assigned to the job titles could differ somewhat between the various smelters (Kromhout et al., 1987; Stewart et al., 1996; Tielemans et al., 1998). It is, however, difficult to know if these approximations would lead to overestimation or underestimation of dust exposure concentration levels. As the great majority of the samples were collected randomly, this misclassification should most likely be regarded as non-differential.
In the furnace house, a significant difference in dust exposure concentration levels between tappers and furnace operators was found. Unfortunately, we were not able to separate these two job functions of the employees in 10 of the 15 smelters. In these smelters, a new job title, ‘furnace section worker', was constructed. Thus, in smelters where the job classification of the employees did not differentiate between tappers and furnace operators, the tappers were underestimated regarding exposure, whereas the furnace operators were overestimated. This misclassification is likely to weaken the association between health outcome and dust exposure in the epidemiological analyses.
In several of the departments, the 25th and the 75th percentile of the measured dust exposure concentration levels were approximately half and twice the median, respectively. There is therefore a 50% probability that the true dust exposure level was less than half or more than twice the estimated value, and we must expect that a considerable proportion of the workers were misclassified regarding dust exposure. Most likely, this misclassification was non-differential, leading to an underestimation of the relative risk (Goldberg et al., 1993).
Dust exposure levels
Total dust exposure in the Norwegian smelting industry has been described to some extent in former studies (Langard, 1980; Kjuus et al., 1986; Hobbesland et al., 1997). Historically dust concentration >5 mg m–3 and upto 30 mg m–3 was not uncommon (Langard, 1980; Kjuus et al., 1986). In a study from 1997, Hobbesland et al. describe that furnace workers in FeSi and Si-metal production had a total dust exposure of 3.4 mg m–3 (95% confidence interval: 1.1, 13.8) in the period 1986–1990 (Hobbesland et al., 1997). The latter finding is comparable to our findings in the present study, indicating that only small changes in workplace exposure were observed from 1990 to 1996.
Individual characteristics of exposure
Some determinants of exposure need further comments. First, women were found to have lower dust exposure than men, and the oldest workers were found to have lower dust exposure than younger workers. As the dust exposure concentration level for a given exposure group (job title, department, smelter) was constant during the study period, the differences found in relation to gender and age during the study may originate from women holding less exposed jobs than men and the oldest employees holding less exposed jobs than the younger. Second, smokers were more exposed to dust than those who had never smoked. This finding is in accordance with the findings of others (Bakke et al., 1990). Third, previous exposure status was a determinant of current exposure in the FeSi/Si-metal production group, i.e. it appeared that subjects with previous exposure to dust, fumes and gases continued to have higher exposure than their colleagues. Information about previous exposure to dust, fumes and gases before current job were obtained from the questionnaire used at the first time examination of the employees and as such made differential reporting likely. As the kappa value of this question in a separate study was 0.61, which is regarded as good, we do believe that the question of previous exposure should be taken into account (Kongerud et al., 1989).
The reason for the positive relationship between both smoking and previous exposure and current occupational exposure is not known. It might be explained by differences in susceptibility to air pollutants and tobacco smoke: employees less susceptible to air pollutants or tobacco smoke tolerate exposed occupations better than susceptible subjects and hence continue in such jobs (Becklake and Lalloo, 1990). Though, others have found that the impact of smoking as a confounder in occupational studies is of minor importance (Blair et al., 2007). The finding that women worked in less exposed jobs than men might also be explained by a higher susceptibility to air pollutants of females (Becklake and Kauffmann, 1999).
|
CONCLUSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONFUNDINGACKNOWLEDGEMENTSREFERENCES |
|---|
The dust exposure concentration levels of the employees were generally higher in the FeSi/Si-metal production group than in the SiMn/FeMn/FeCr production group. Gender, age, current smoking, job category and previous exposure were found to be significantly related to the dust exposure concentration levels of the employees and should therefore be evaluated in future analyses in this industry.
|
FUNDING |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONFUNDINGACKNOWLEDGEMENTSREFERENCES |
|---|
Confederation of Norwegian Enterprise (CNE); Working Environment Fund (S-2374, S-2495), and the Norwegian smelting industry.
|
ACKNOWLEDGEMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONFUNDINGACKNOWLEDGEMENTSREFERENCES |
|---|
The authors wish to thank the smelting industry, both management and employees, for their considerable cooperation. We are grateful to the local occupational health services for their valuable contribution to the study and to the employees who participated. We would also like to thank the advisory council, V. Digernes, J. Efskind, E. G. Astrup and H. Kjuus, for their valuable comments on the manuscript. We would like to give special thanks to E. G. Astrup for her help with the exposure classification. The study was accomplished with administrative assistance from the Federation of Norwegian Industries.
Received March 30, 2007; in final form June 18, 2008
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONFUNDINGACKNOWLEDGEMENTSREFERENCES |
|---|
Bakke P, Gulsvik A, Eide GE, et al. Smoking habits and lifetime occupational exposure to gases or dusts, including asbestos and quartz, in a Norwegian community. Scand J Work Environ Health (1990) 16:195–202.[Web of Science][Medline]
Becklake MR. Occupational exposures: evidence for a causal association with chronic obstructive pulmonary disease. Am Rev Respir Dis (1989) 140:S85–91.[Web of Science][Medline]
Becklake MR, Kauffmann F. Gender differences in airway behaviour over the human life span. Thorax (1999) 54:1119–38.
Becklake MR, Lalloo U. The ‘healthy smoker’: a phenomenon of health selection? Respiration (1990) 57:137–44.[Web of Science][Medline]
Benke G, Sim M, Fritschi L, et al. Beyond the job exposure matrix (JEM): the task exposure matrix (TEM). Ann Occup Hyg (2000) 44:475–82.
Benke G, Sim M, Fritschi L, et al. Comparison of occupational exposure using three different methods: hygiene panel, job exposure matrix (JEM), and self reports. Appl Occup Environ Hyg (2001) 16:84–91.[CrossRef][Medline]
Blair A, Stewart P, Lubin JH, et al. Methodological issues regarding confounding and exposure misclassification in epidemiological studies of occupational exposures. Am J Ind Med (2007) 50:199–207.[CrossRef][Web of Science][Medline]
Blanc PD, Eisner MD, Trupin L, et al. The association between occupational factors and adverse health outcomes in chronic obstructive pulmonary disease. Occup Environ Med (2004) 61:661–7.
CEN-convention: NS-EN 481. Workplace atmosphere—size fraction definitions for measurement of airborne particles. (1993) Norway: CEN Brussels. ICS 13.040.30.
Checkoway H, Pearce N, Kriebel D. Research methods in occupational epidemiology (2004) New York: Oxford University Press.
de Vocht F, Zock JP, Kromhout H, et al. Comparison of self-reported occupational exposure with a job exposure matrix in an international community-based study on asthma. Am J Ind Med (2005) 47:434–42.[CrossRef][Web of Science][Medline]
Fichte RM. (1986) Production. Ullmann's encyclopedia of industrial chemistry. Chromium and Chromium Alloys. Vol. A7. 5th edn. Weinheim, Germany: VCH, Verlagsgesellschaft mbH. pp. 51–6.
Fitzmaurice GM, Laird NM, Ware JH. Applied longitudinal analysis (2004) New York: Wiley.
Goldberg M, Kromhout H, Guenel P, et al. Job exposure matrices in industry. Int J Epidemiol (1993) 22(Suppl 2):S10–5.[Abstract]
Hobbesland A, Kjuus H, Thelle DS. Mortality from nonmalignant respiratory diseases among male workers in Norwegian ferroalloy plants. Scand J Work Environ Health (1997) 23:342–50.[Web of Science][Medline]
Johnsen HL, Soyseth V, Hetland SM, et al. Production of silicon alloys is associated with respiratory symptoms among employees in Norwegian smelters. Int Arch Occup Environ Health (2007) 81:451–9.[CrossRef][Web of Science][Medline]
Johnsen HL, Kongerud J, Hetland SM, et al. Decreased lung function among employees at Norwegian smelters. Am J Ind Med (2008) 51:296–306.[CrossRef][Web of Science][Medline]
Kauffmann F, Drouet D, Lellouch J, et al. Occupational exposure and 12-year spirometric changes among Paris area workers. Br J Ind Med (1982) 39:221–32.[Web of Science][Medline]
Kenny LC, Aitken R, Chalmers C, et al. A collaborative European study of personal inhalable aerosol sampler performance. Ann Occup Hyg (1997) 41:135–53.
Kjuus H, Andersen A, Langard S, et al. Cancer incidence among workers in the Norwegian ferroalloy industry. Br J Ind Med (1986) 43:227–36.[Web of Science][Medline]
Kongerud J, Vale JR, Aalen OO. Questionnaire reliability and validity for aluminum potroom workers. Scand J Work Environ Health (1989) 15:364–70.[Web of Science][Medline]
Kromhout H, Oostendorp Y, Heederik D, et al. Agreement between qualitative exposure estimates and quantitative exposure measurements. Am J Ind Med (1987) 12:551–62.[Web of Science][Medline]
Langard S. A survey of respiratory symptoms and lung function in ferrochromium and ferrosilicon workers. Int Arch Occup Environ Health (1980) 46:1–9.[CrossRef][Web of Science][Medline]
Messing K, Dumais L, Courville J, et al. Evaluation of exposure data from men and women with the same job title. J Occup Med (1994) 36:913–7.[Web of Science][Medline]
Neuer B, Rau G. (1993) Ferrosilicon production. Ullmann's encyclopedia of industrial chemistry. Silicon. Vol. A23. 5th edn. Weinheim, Germany: VCH, Verlagsgesellschaft mbH. pp. 743–4.
Rybicki BA, Johnson CC, Peterson EL, et al. Comparability of different methods of retrospective exposure assessment of metals in manufacturing industries. Am J Ind Med (1997) 31:36–43.[CrossRef][Web of Science][Medline]
Seixas NS, Checkoway H. Exposure assessment in industry specific retrospective occupational epidemiology studies. Occup Environ Med (1995) 52:625–33.
Seixas NS, Moulton LH, Robins TG, et al. Estimation of cumulative exposures for the National Study of Coal Workers' Pneumoconioses. Appl Occup Environ Hyg (1991) 6:1032–41.
Sherrill D, Viegi G. On modeling longitudinal pulmonary function data. Am J Respir Crit Care Med (1996) 154:S217–22.[Web of Science][Medline]
Soyseth V, Johnsen HL, Benth JS, et al. Production of silicon metal and alloys is associated with accelerated decline in lung function: a 5-year prospective study among 3924 employees in Norwegian smelters. J Occup Environ Med (2007) 49:1020–6.[CrossRef][Web of Science][Medline]
Stewart PA, Blair A, Dosemeci M, et al. Collection of exposure data for retrospective occupational epidemiologic studies. Appl Occup Environ Hyg (1991) 6:280–9.
Stewart PA, Lees PSJ, Francis M. Quantification of historical exposures in occupational cohort studies. Scand J Work Environ Health (1996) 22:405–14.[Web of Science][Medline]
Tielemans E, Kupper LL, Kromhout H, et al. Individual-based and group-based occupational exposure assessment: some equations to evaluate different strategies. Ann Occup Hyg (1998) 42:115–9.
Trupin L, Earnest G, San PM, et al. The occupational burden of chronic obstructive pulmonary disease. Eur Respir J (2003) 22:462–9.
Viegi G, Di PC. Chronic obstructive lung diseases and occupational exposure. Curr Opin Allergy Clin Immunol (2002) 2:115–21.[Medline]
Vincent J. Aerosol science for industrial hygienists (1995) New York: Pergamon.
Wellbeloved DB, Craven PM, Waudby JW. (1990) Production of Silicomanganese. Production of High-carbon manganese. Production of Medium-carbon manganese. Production of low-carbon manganese. Ullmann's encyclopedia of industrial chemistry. Manganese and Manganese alloys. Vol. A16. 5th edn. Weinheim, Germany: VCH Verlagsgesellschaft mbH. pp. 91–103.
Zulehner W. (1993) Metallurgical Silicons. Ullmann's encyclopedia of industrial chemistry. Silicon. Vol. A23. 5th edn. Weinheim, Germany: VCH Verlagsgesellschaft mbH. pp. 722–3.
CiteULike 
Connotea 
Del.icio.us What's this?
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||