Evaluation of workplace exposure to respirable crystalline silica in Italy

ABSTRACT

Background:

Crystalline silica is a human carcinogen and its use is widespread among construction, mining, foundries, and other manufacturing industries.

Purpose:

To evaluate occupational exposure to crystalline silica in Italy.

Methods:

Data were collected from exposure registries and descriptive statistics were calculated for exposure-related variables. The number of potentially exposed workers was estimated in a subset of industrial sectors. Linear mixed model analysis was performed to determine factors affecting the exposure level.

Results:

We found 1387 cases of crystalline silica exposure between 1996 and 2012. Exposure was most common in construction work (AM = 0.057 mg/m3, N = 505), and among miners and quarry workers (AM = 0.048 mg/m3, N = 238). We estimated that 41 643 workers were at risk of exposure in the selected industrial sectors during the same period.

Conclusions:

This study identified high-risk sectors for occupational exposure to crystalline silica, which can help guide targeted dust control interventions and health promotion campaigns in the workplace.

METHODS

Data gathering

The SIREP system has been fully described elsewhere. In brief, Italian law requires that employers collect exposure data for select carcinogens and report to SIREP every 3 years. Reporting is mandatory for carcinogens classified as 1 and 2 by the European Union (1, substance known to be carcinogenic to humans; 2, substances that should be regarded as if carcinogenic to humans), but is voluntary for other possibly carcinogenic substances, including silica. Employers are required to report carcinogen type, anagraphical and occupational data of exposed employees, and exposure levels. The information reported by employers is standardized and includes: the firm’s economic activity sector and work force size; workers’ personal data and job type; year of measurement and level of exposure (magnitude, frequency, and duration). Employers are responsible for the measurement procedures and air sampling methods in accordance with European standards, which provide technical guidance to implement a dust monitoring strategy.

Data selection

A total of 1387 RCS measurements, representing 1341 exposures (1115 workers in 144 different firms) between 1996 and 2012 were selected from the SIREP database for analysis (313 exposures were excluded from analysis as the measurement was not specified). The airborne concentration of RCS for some exposures (N = 44) was measured at least twice between 1996 and 2012. The crystalline silica forms studied included quarz, cristobalite, tridymite, and tripoli. Measurements (N = 470) that were below the analytical limit of detection (LOD) were replaced with the LOD value divided by two (LOD/2). Air sampling measurements were collected over an eight hour work shift. Information on the type of sampling methodology (personal or stationary) was not available. In order to increase the precision of the estimates, only sectors and occupations having more than 50 measurements were included in the descriptive analysis. International standard classifications were used to code economic activity sectors of firms (NACE Rev. 1) and worker occupation (ISCO-88). Descriptive statistical analyses were used to calculate the means (arithmetic and geometric) of RCS exposure levels in addition to 95% confidence intervals (CI), standard deviations (SD), and geometric standard deviations (GSD).

Estimating exposed workers

Owing to the shortage of data available about certain industrial sectors, the number of workers potentially exposed to RCS was estimated only for the sectors for which the percentage of reported work force (exposed together with non-exposed) was more than 1% of the total sector work force (RWi/Wi>1%, where RWi is the SIREP reported work force, Wi is the total work force, and i is the ith economic sector), and with at least three firms recorded. The total sector work force was estimated using national statistics from the Italian Institute for Statistics (ISTAT). For included industrial sectors, the number of potentially RCS exposed workers was reconstructed using the percentage of exposed workers in relation to both the work force size of firms recorded in SIREP and the national statistics on work force (i.e. PEi = Wi×(Ei/RWi), where PEi is the potentially exposed workers, Wi is the ISTAT total work force, RWi is the SIREP reported work force, and Ei is the SIREP exposed workers). SIREP exposed workers (Ei) is the total number of workers with RCS exposure measurements recorded in SIREP (including those with levels below the LOD), for the ith economic sector. Results were stratified by gender.

Mixed-effects model

A mixed effects model with a random firm-specific intercept was adopted to evaluate the association between exposure variables and air RCS concentration. A non-parametric one-way analysis of variance using the Kruskal–Wallis test was applied to detect and identify variables influencing the exposure level and only variables statistically significant at the P<0.05 were included in the model. All independent variables (fixed effects) in the model were categorical: RCS form (quartz, cristobalite, tripoli), economic sector (NACE codes), firm size (<10 workers, 10–20 workers, 20–50 workers, 50–100 workers, >=100 workers), firm location (by Italian region), occupation (ISCO-88 codes), and year. Exposure measurements were natural log transformed to account for positively skewed data and approximately log-normal distributed. A restricted maximum likelihood estimation method was used in the mixed model and the Akaike Information Criterion was applied to achieve the best fitting model.

The mixed-effects model is described by the following equation:

for i = 1, 2, ..., n (firms), where Yi is the exposure level. The model assumptions are: random effect corresponding to firm (Firm effect) is approximately normally distributed with a mean of 0 and a variance of σ2B residual (Error) is approximately normally distributed with a mean of 0 and a variance of σ2W; and (Firm effect) and (Error) are statistically independent.

Partial eta squared (η2) was computed for each specific effect to measure the degree of association using F-test statistics and the following formula:

The data were analyzed with SAS statistical software version 9.3 (MIXED procedure; SAS Institute, Inc., Cary, NC, USA). This study is part of a research project with the Italian Workers’ Compensation Authority in the institutional research plan for the years 2013–2015 (code P01L02).

RESULTS

Descriptive statistics

The mean airborne concentration of RCS in workplaces was slightly higher for men (N = 1350, AM = 0.05 mg/m3, SD = 0.14, GM = 0.018 mg/m3, GSD = 4.51, 95% CI = 0.017–0.019) than for women (N = 37, AM = 0.05 mg/m3, SD = 0.14, GM = 0.009 mg/m3, GSD = 5.67, 95% CI = 0.005–0.015). Since the female group had few exposure measurements, no further analysis was performed. The greatest number of measurements occurred in the construction sector (N = 505 men, GM = 0.045 mg/m3, GSD = 1.71), and the lowest level of exposure was in the manufacturing of basic metals sector (N = 181 men, GM = 0.007 mg/m3, GSD = 2.62). The highest percentage of measurements exceeding the exposure limit proposed by ACGIH was in the construction sector (93%) and for miners and quarry workers (97%). Overall, 44% of exposed workers had an exposure measurement equal to or above the limit recommended by SCOEL (0.05 mg/m3), while 50% had a value exceeding that proposed by ACGIH (0.025 mg/m3). Distribution by industrial sector is shown in Table 1 and distribution by occupational group is provided in Table 2.

Among RCS forms, the GM concentration was the highest for the cristobalite (N = 644, GM = 0.025 mg/m3, GSD = 3.92), and similar for quartz (N = 589, GM = 0.013 mg/m3, GSD = 4.71) and tripoli (N = 96, GM = 0.012 mg/m3, GSD = 2.02).

Regarding the distribution of RCS exposure level by work force size, micro and small firms (<10–20 employees) had the lowest exposure risk (N = 184, GM = 0.007 mg/m3, 95% CI = 0.005–0.009), whereas the highest exposure (N = 553, GM = 0.032, 95% CI = 0.029–0.034) was in the largest firms (100–1000 employees). The distribution of exposure level by workforce size is shown in Fig. 1. The highest level of exposure was recorded in 2006 (N = 237, GM = 0.039, 95% CI = 0.034–0.043) and the largest number of measurements in 2004 (N = 289, GM = 0.038, 95% CI = 0.033–0.043).

Estimating exposed workers

Overall, 41 643 workers were estimated as potentially at-risk of exposure to RCS in the selected industrial sectors between 1996 and 2012. The sector with the greatest percentage of exposed workers was sawing and processing of stones and marble (NACE code: 26.70.1) with 17 380 workers potentially exposed (97% male). For women, the sector with the greatest number of exposed workers was manufacture of ceramic household and ornamental articles (NACE code: 26.21.0) with 2458 workers potentially exposed. Detailed data on the number of exposed workers by industrial sector and gender are shown in Table 3.

Mixed-effects model

One-way analysis of variance showed the variables most related to silica exposure level were: occupation (eta-square = 0.574; chi-square = 827.29, P<0.0001), economical sector (eta-square = 0.473; chi-square = 722.33, P<0.0001), measurement year (eta-square = 0.251; chi-square = 568.12, P<0.0001), and firm size (eta-square = 0.141; chi-square = 382.94; P<0.0001). In the mixed effects model, exposure variables and all interactions (crossed) were analyzed to test independent effects. Non-significant variables (P>0.05) were excluded stepwise from the model. The fixed-effects of the model explained 84% of the variance in the observed exposure data. In the final model (the lowest value of the Akaike Information Criterion), between- and within-firm variances were σ2B = 0.26 and σ2W = 1.83, respectively. The final model is presented in Table 4 and the detailed results are reported as supplementary material (http://www.maneyonline.com/doi/suppl/10.1179/2049396714Y.0000000078).
