Ann. occup. Hyg., Vol. 47, No. 2, pp. 111-122, 2003
© 2003 British Occupational Hygiene Society
Published by Oxford University Press
Excessive Exposure to Silica in the US Construction Industry
1 School of Public Health, University of North Carolina, Chapel Hill, NC 27599-7431, USA; 2 Urban Public Health Program, School of Health Sciences, Hunter College–City University of New York, 425 East 25th Street, New York, NY 10010, USA; 3 Center to Protect Workers’ Rights, 8484 Georgia Avenue, Silver Spring, MD 20910, USA; 4 Harvard School of Public Health, 665 Huntington Avenue, Boston, MA 02215, USA
Received 30 August 2002; in final form 20 November 2002
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTS AND DISCUSSIONCONCLUSIONS AND RECOMMENDATIONSREFERENCES |
|---|
Exposures to respirable dust and silica were investigated among 36 construction sites in the USA. Personal measurements (n = 151) were analyzed from 80 workers in four trades, namely bricklayers, painters (while abrasive blasting), operating engineers and laborers. Painters had the highest exposures (median values for respirable dust and silica: 13.5 and 1.28 mg/m3, respectively), followed by laborers (2.46 and 0.350 mg/m3), bricklayers (2.13 and 3.20 mg/m3) and operating engineers (0.720 and 0.075 mg/m3). Mixed models were fitted to the log-transformed air levels to estimate the means and within- and between-worker variance components of the distributions in each trade. We refer to the likelihood that a typical worker from a given trade would be exposed, on average, above the occupational exposure limit (OEL) as the probability of overexposure. Given US OELs of 0.05 mg/m3 for respirable silica and 3 mg/m3 for respirable dust, we estimated probabilities of overexposure as between 64.5 and 100% for silica and between 8.2 and 89.2% for dust; in no instance could it be inferred with certainty that this probability was <10%. This indicates that silica exposures are grossly unacceptable in the US construction industry. While engineering and administrative interventions are needed to reduce overall air levels, the heterogeneous exposures among members of each trade suggest that controls should focus, in part, upon the individual sites, activities and equipment involved. The effects of current controls and workplace characteristics upon silica exposures were investigated among operating engineers and laborers. Silica exposures were significantly reduced by wet dust suppression (~3-fold for laborers) and use of ventilated cabs (~6-fold for operating engineers) and were significantly increased indoors (about 4-fold for laborers). It is concluded that urgent action is required to reduce silica exposures in the US construction industry.
Keywords: construction industry; controls; dust; mixed models; overexposure; silica; variance components
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTS AND DISCUSSIONCONCLUSIONS AND RECOMMENDATIONSREFERENCES |
|---|
Numerous construction activities, such as abrasive blasting, cutting, drilling and grinding of concrete and masonry and road milling and sweeping, generate clouds of fine dust (Riala, 1988; Blute et al., 1999; Lumens and Spee, 2001; Akbar-Khanzadeh and Brillhart, 2002; Bakke et al., 2002; Linch, 2002; Woskie et al., 2002). Since these dusts generally contain crystalline silica, a potent lung toxin and suspected human carcinogen (IARC, 1997), it is reasonable to ask whether construction workers are overexposed relative to current occupational exposure limits (OELs). Several recent studies document air concentrations of respirable silica in the US construction industry at levels above the current OEL of 0.05 mg/m3 (NIOSH, 1975; ACGIH, 2002) during everyday operations (Blute et al., 1999; Lumens and Spee, 2001; Akbar-Khanzadeh and Brillhart, 2002; Bakke et al., 2002; Linch, 2002; Woskie et al., 2002) as well as during governmental inspections (Linch et al., 1998). However, the extent of overexposure cannot be accurately gauged because the variations in air levels within and between workers have not been reported for the construction trades.
Recognizing that silica is slowly cleared from the lung, the long-term mean exposure received by a worker over months or years is ultimately predictive of lung disease. Thus, a worker’s risk of disease exceeds that inherent in the OEL when his or her long-term mean exposure exceeds the OEL, a condition we refer to as ‘overexposure’. With knowledge of the mean exposure of each trade, as well as the corresponding within- and between-worker variance components, it is possible to test the probability of overexposure for any particular trade against an a priori acceptable value (Rappaport et al., 1995, 1999; Lyles et al., 1997; Weaver et al., 2001).
When a trade has an unacceptable probability of overexposure, some form(s) of intervention is required. If the trade mean exposure is at or greater than the OEL, then trade-level interventions, including engineering or administrative controls, are needed. Furthermore, with evidence of considerable heterogeneity of exposure among trade members, the intervention strategy should reflect the particular interplay of sites, activities and equipment involved (Rappaport, 1991). We define the heterogeneity of exposure within a trade in terms of the proportion of predicted random worker effects that are significantly different from 0 (Rappaport et al., 1999; Weaver et al., 2001).
The current study was motivated by the paucity of information about the overall levels of exposure to respirable dust and crystalline silica (hereafter simply ‘dust’ and ‘silica’), as well as the within- and between-worker variance components, for the US construction trades. Data were compiled from surveys conducted at 36 sites in four construction trades. Using mixed models (Rappaport et al., 1999; Weaver et al., 2001), we estimated the within- and between-worker variance components for each trade as well as the associated probabilities of overexposure, relative to OELs of 0.05 mg/m3 for respirable silica (NIOSH, 1975; ACGIH, 2002) and 3 mg/m3 for respirable dust (ACGIH, 2002), and the random worker effects. We then employed the trade mean exposures and proportions of significant random worker effects to optimize control strategies. Finally, we employed a mixed model to evaluate the effects of existing controls and environmental characteristics upon silica exposures in the two trades with greatest representation in our sample (operating engineers and laborers).
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTS AND DISCUSSIONCONCLUSIONS AND RECOMMENDATIONSREFERENCES |
|---|
Construction sites, trades and surveys
As shown in Table 1, data were obtained from 36 construction sites in the Eastern and Midwestern USA that had been surveyed between April 1992 and October 2000. Sites were selected in part to represent activities known to generate silica-containing dust. Multiple surveys were conducted at most sites. The median interval between surveys was 6 days (range 1–580 days) and, with three exceptions (sites 6, 26 and 27), all surveys at a given site were performed within a calendar month. A total of 80 construction workers were investigated from the following trades: painters (PA), bricklayers (BL), operating engineers (OE) and laborers (LA). PA were monitored only during days when they performed abrasive blasting with either sand or coal slag; these workers always wore respiratory protection. BL were typically employed in construction or rehabilitation of commercial buildings, using hand drills, grinders, saws and jack hammers; they occasionally wore respiratory protection. OE and LA worked primarily at outdoor sites typically involving rehabilitation of roads or bridges; they rarely wore respiratory protection. OE operated large pieces of heavy equipment, including vertical drills, backhoes, street sweepers, dowel packs, road milling equipment and crushers, while sitting in open or ventilated cabs
2–3 m above the work surface. LA worked at ground level, using hand-operated equipment, including drills, saws, jack hammers, chipping guns, rakes and brooms.
|
Data were compiled from three sources: surveys conducted by the Center to Protect Workers Rights (CPWR) at 13 sites between July 1999 and October 2000; surveys conducted by the National Institute for Occupational Safety and Health (NIOSH) at 14 sites between April 1992 and August 1995; and surveys conducted by the Mount Sinai School of Medicine and Hunter College of the City University of New York (MSSM) at nine sites between September 1997 and December 1998. Data from surveys conducted by CPWR and MSSM have not been reported previously. Data from surveys conducted by NIOSH were previously summarized for other purposes (Linch, 2002). For the current study all primary data were compiled from the original NIOSH reports. Since the construction trades were not reported in most of the NIOSH reports, trades were imputed to these workers based upon descriptions of activities.
Air sampling was performed by occupational hygienists during surveys conducted by NIOSH, MSSM and the Harvard School of Public Health (CPWR data). Surveys of some abrasive blasting and masonry operations (CPWR data) were performed by trained trade members, working in tandem with occupational hygienists. Prior to sampling, these workers attended a 4 day training session to learn air monitoring methods, silica hazards and use of task-based survey methodology (Susi et al., 2000).
Exposure measurements
Personal respirable air samples were collected on 37 mm pre-weighed PVC filters (5 µm pore size). CPWR and NIOSH samples were collected at 1.7 l/min using 10 mm nylon Dorr–Oliver cyclones as pre-separators while MSSM samples were collected at 2.2 l/min using Higgins–Dewell conductive cyclones as pre-separators. Personal monitors were clipped to the workers’ lapels regardless of whether respiratory protection was worn.
Dust was measured gravimetrically according to NIOSH Method 600 and silica was measured as 
-quartz by X-ray diffraction using NIOSH method 7500 (NIOSH, 1994). One NIOSH measurement was estimated as the mean of five area samples obtained from locations surrounding an abrasive blaster in a small building; these samples were collected at 9 l/min with 37 mm PVC filters connected to 18 mm metal cyclones.
A total of 151 measurements of dust and silica were available from 80 workers. The duration of air sampling ranged from 30 to 561 min with a median value of 315 min and an interquartile range of 227–389 min. Although most of the measurements were based upon a single full-shift sample, some were averages of two or three serial measurements.
As summarized in Table 2, between 11 and 80 measurements were obtained from 8–37 workers in each trade; 1–7 measurements were obtained from each worker and at least one worker in each trade had repeated measurements. Considering the 151 pairs of dust and silica measurements, 24 observations (six for dust and 18 for silica) were below the analytical limit of detection (LOD); these measurements were assigned a value of LOD/
2 for statistical analyses (Hornung and Reed, 1990). One additional measurement was excluded from analysis because field notes indicated that the sampler had become disconnected from the pump during the work shift.
|
Statistical models
All statistical procedures were performed with SAS software v.8.02 (SAS Institute, Cary, NC). Mixed models were used to characterize the distributions of exposure to dust and silica. Briefly, the models assumed that exposures varied by job group (defined here by the trade), between workers in a given group (i.e. trade members with different mean exposures) and within workers from survey to survey. These models have been thoroughly evaluated with measurements of exposure to welding fumes among construction trades and with exposures of other occupational groups to various contaminants (Rappaport et al., 1999; Weaver et al., 2001).
Let Xhij represent the dust or silica concentration (mg/m3) measured during the jth survey for the ith worker in the hth group and let Yhij represent the natural logarithm of Xhij, i.e. Yhij = ln(Xhij). The mean and variance of Xhij for the hth group are designated as µx,h and 
, respectively, and those of Yhij as µy,h and 
, respectively. The mixed model relates to the logged exposure levels as follows:
Yhij = ln(Xhij) = µy,h + ßhi + 
hij (1)
(for h = 1, ..., H groups; i = 1, ..., kh workers in the hth group; j = 1, ..., nhi measurements from the ith worker in the hth group). In Model (1), µy,h represents the fixed mean of logged exposures in the hth group, ßhi is the random effect of the ith worker in the hth group and hij is the random error effect of the jth measurement from the ith worker in the hth group. It is assumed that ßhi and hij are normally distributed, with zero mean and variances of 
and 
, representing the between- and within-worker variance components of the hth group. We designate the variance of Yhij as 
. Thus, Yhij is normally distributed with mean µy,h and variance 
and Xhij is log-normally distributed with mean
and variance
Note also that the log-normal distribution of exposures for the hth group can be characterized with the geometric mean and geometric standard deviation, designated as
and
respectively.
In a separate analysis, the fixed effects upon silica exposure were explored for all pertinent covariates in the two largest groups in our study (OE and LA). The mixed model used for this analysis is:
(2)
(for m = 1, 2, ..., M covariates), where all common terms are the same as for Model (1) and m,h is the regression coefficient for Cmhij, representing the observed value of the mth covariate during the jth survey for the ith worker in the hth group. Separate applications of Model (2) were performed for OE and LA, with covariates consisting of calendar time and dummy variables for the following effects: indoor work, use of water to suppress dust generation, use of a ventilated cab (OE only) and use of mechanical/dilution ventilation (LA only). Variables were retained in final models at a significance level of 0.10. The same assumptions hold as for Model (1), except that Yhmij is now assumed to be normally distributed with mean
where M represents the number of covariates in the final model) and variance 
. The mean and variance of the lognormal distribution under Model (2) for group h and covariate m are given by
and variance
while the geometric mean and geometric standard deviation are given by
and
respectively.
Estimation of parameters and model fit
Model (1) is completely specified for the hth group by mean µh,y and the variance components 
and 
. These parameters were estimated under full and reduced versions of Model (1), assuming either distinct 
and 
for each group (full model) or distinct 
and common 
for all groups (reduced model). Since reduced Model (1) is more parsimonious than full Model (1), likelihood ratio (LR) tests were performed at a significance level of 0.01, to determine whether it was reasonable to pool 
across groups (Weaver et al., 2001). Model (2) is completely specified for the hth group and M covariates by mean
and the variance components 
and 
. Since Model (2) was applied separately to OE and LA, these parameters were estimated only under the full model.
Restricted maximum likelihood (REML) estimates of the parameters were obtained under Models (1) and (2) using Proc Mixed of SAS. These estimates are designated with a ‘^’, e.g. under Model (1) 
and 
represent the estimated mean and variance of Yhij for the hth group,
and
Also, the predicted random effect 
of the ith worker in the jth group was obtained under Model (1) via Proc Mixed.
Since data had been pooled over a span of years, profile plots were examined and time trends were investigated with Model (2). No evidence of trends was observed.
Testing overexposure
Exposures to dust and silica were tested in terms of the probability that a worker’s long-term mean exposure would be greater than the OEL (Rappaport et al., 1995, 1999; Lyles et al., 1997; Weaver et al., 2001). We define overexposure as the condition where a typical worker in the hth group would have a long-term mean exposure (µx,hi) greater than the OEL. The probability of overexposure can be related to Model (1) with the following expression:
(3)
where 
{x} denotes the probability that a standard normal variate would fall below the value x. The probability of overexposure for each trade was estimated as 
by substituting the estimated parameters, obtained under either full or reduced Model (1), into equation (3). Note that 
h is undefined under equation (3) when 
B,h = 0.
The magnitude of h for each group was tested using an approach recommended by Weaver et al. (2001) based upon earlier work of Lyles et al. (1997) and Rappaport et al. (Rappaport et al., 1995, 1999). Basically, the following hypotheses were tested for the hth group:
H0,h: h 
A versus HA,h: h < A for h = 1, 2, ..., H (4)
where A is an a priori acceptable level of overexposure, assigned a value of 0.10. Overexposure was evaluated by applying a test statistic, designated 
, which was constructed in one of two ways depending upon whether 
was non-zero or zero. If 
> 0, then 
was evaluated with a Wald-type test; this was the case for seven of the eight trade/contaminant combinations. In the remaining case, involving dust exposure among BL, 
= 0 and an alternative test was used to test whether the group mean was less than the OEL (since the between-worker variance component was very small or zero in this case, the probability of overexposure was very close to the probability that µx,h > OEL). In either case, the test statistic was compared to a critical value C, at = 0.05 level of significance. If 
< C, then H0,h was rejected in favor of HA,h and exposure was considered acceptable; otherwise, when 
C, exposure was considered unacceptable. Detailed instructions for computations of 
and C are provided by Weaver et al. (2001). Because this test is sensitive to departures from normality, the predicted random effects (i.e. the 
values) divided by their standard errors were tested for normality both visually, via q–q plots, and with the Shapiro–Wilks W-test at a 0.10 level of significance. All model fits were deemed acceptable by these criteria.
Heterogeneity of exposure
When an acceptable probability of overexposure cannot be inferred for a particular group, we advocate basing the type of intervention, at least in part, upon heterogeneity of exposure in the group (Rappaport, 1991). We address heterogeneity via the proportion of random effects in the hth group (ßhi) that are significantly different from zero (Rappaport et al., 1999; Weaver et al., 2001). Groups with sizeable proportions of significant ßhi (>10%, say) can reasonably be regarded as heterogeneously exposed.
|
RESULTS AND DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTS AND DISCUSSIONCONCLUSIONS AND RECOMMENDATIONSREFERENCES |
|---|
Scatter plots
The personal measurements of dust and silica are shown in Fig. 1, with data aggregated by workers in each trade (note that the exposure scale is logarithmic). The measurements show tremendous variability of exposure within and between workers in a given trade. The ranges of exposure are consistent with those presented in other recent investigations of the same US construction trades. For example, levels observed among BL in this study (0.160–29.0 mg/m3 dust and 0.007–14.2 mg/m3 silica, n = 11) are similar to those reported for hand grinding of concrete at two US sites (0.34–81.0 mg/m3 dust and 0.02–7.10 mg/m3 silica, n = 49) (Akbar-Khanzadeh and Brillhart, 2002). Likewise, ranges for OE in this study (0.01–3.00 mg/m3 dust and 0.007–0.800 mg/m3 silica, n = 46) and LA (0.01–21.0 mg/m3 dust and 0.007–5.90 mg/m3 silica, n = 80) are comparable to those reported elsewhere during highway construction (OE, 0.15–2.39 mg/m3 dust and 0.008–0.059 mg/m3 silica, n = 22; LA, 0.18–21.8 mg/m3 dust and 0.008–1.64 mg/m3 silica, n = 36) (Blute et al., 1999).
|
The comparisons of measurements to the OELs shown in Fig. 1 are striking, particularly for silica, where 116 of 151 observations (77%) exceeded the OEL of 0.05 mg/m3. Such great exceedance of 0.05 mg/m3 is generally consistent with other recent US studies [i.e. 50% of OE measurements and 63% of LA measurements during highway construction (Blute et al., 1999), 69% of BL measurements during concrete finishing (Akbar-Khanzadeh and Brillhart, 2002) and 30.8% of LA measurements and 6.8% of OE measurements during heavy and highway construction (Woskie et al., 2002)].
Summary statistics
Median values and ranges are shown in Table 3 for dust and silica exposures in each trade. The fold ranges of daily air levels (maximum/minimum) varied from 181 to 718 within each trade for dust and from 101 to 2030 for silica. Large differences in median exposures were also apparent among trades, with PA having the highest exposures (13.5 mg/m3 dust and 1.28 mg/m3 silica, n = 14), followed by LA (2.46 mg/m3 dust and 0.350 mg/m3 silica, n = 80), BL (2.13 mg/m3 dust and 0.320 mg/m3 silica, n = 11) and OE the lowest exposures (0.720 mg/m3 dust and 0.075 mg/m3 silica, n = 46).
|
Exposure distributions
The parameters estimated under Model (1) are shown in Table 4. The estimated within-worker variance component (
) was obtained under the full model for dust (P = 0.002 for LR test) and under the reduced model for silica (P = 0.076 for LR test). The estimate of 
ranged from 0.244 to 2.76, depending upon the trade, for dust and was 0.943 for silica in all trades. Since variance components for dust and silica exposures in construction trades have not previously been reported, there are no other estimates with which to compare these results. However, these values of 
are consistent with other published estimates obtained from occupational groups performing intermittent outdoor work (Kromhout et al., 1993; Peretz et al., 1997).
|
Estimates of
ranged from 0 to 2.73 for dust and 0.979 to 3.23 for silica, depending upon the trade. These ranges of 
are substantially larger than those typically observed in intermittent outdoor operations (Kromhout et al., 1993; Peretz et al., 1997). Also, 
was greater than 
in seven of the eight combinations of trade and contaminant, the only exception being for dust among BL. This is also unusual based upon data from intermittent outdoor work, where 
tended to be greater than 
(Kromhout et al., 1993; Peretz et al., 1997). Overall, the variance components estimated during construction activities suggest large differences between workers within a particular trade.
Testing overexposure
The results of testing the probability of overexposure are summarized in Table 5. Since it was not possible to demonstrate that h < A = 0.10 in any case, all groups had unacceptable probabilities of overexposure. This outcome was expected because, in most cases, 
>> A = 0.10 and 
> OEL. Indeed, with 
ranging from 64.5 to 100% for silica, it is reasonable to conclude that these trades were grossly overexposed to a well-recognized respiratory hazard. Since the probability of overexposure for a given trade was always greater for silica than for dust, it appears that the OEL of 0.05 mg/m3 for silica (NIOSH, 1975; ACGIH, 2002) is more stringent than that of 3 mg/m3 for respirable dust (ACGIH, 2002) in evaluating exposures of construction workers.
|
Evaluating the heterogeneity of exposure
Table 6 summarizes the heterogeneity of exposure to dust and silica for the four trades, based upon the percentage of significant (non-zero) random effects ßhi. These percentages ranged from <8.3% to 39.1% for all trade/contaminant combinations and were not less than 10%, except, possibly, for PA (dust and silica) and BL (dust only). These results indicate that exposures were highly heterogeneous in the construction industry, particularly for OE and LA.
|
Taken together, the results summarized in Tables 5 and 6 indicate that exposures were unacceptably high and heterogeneous in the construction industry. Given that the estimated trade mean exposures (
) to silica, which ranged from 0.210 to 3.86 mg/m3, were all much greater than the OEL of 0.05 mg/m3, it is clear that trade level interventions (engineering and administrative controls) will be needed to reduce silica exposures in this industry. Furthermore, additional interventions will be needed at the level of the individual worker to control for differences associated with particular sites, activities and equipment in a given construction trade.
Effects of covariates upon exposure to silica
Model (2) was fitted separately to the logged silica measurements obtained from the two largest groups, namely OE (k = 46) and LA (k = 80), using independent variables consisting of calendar time and dummy variables representing indoor work, use of wet dust suppression, use of a ventilated cab (OE only) and use of mechanical/dilution ventilation (LA only). The parameters of the final models are summarized in Table 7 and the corresponding predicted exposures are shown in Fig. 2. Regarding silica exposures of OE, the final model included only the type of cab that was used. When working in ventilated cabs, OE had silica exposures that were ~6-fold lower than those in open cabs (P = 0.053). This finding supports a recent Norwegian study indicating that tunnel workers operating equipment in closed cabs had ~3-fold lower silica exposures than those working in open cabs (Bakke et al., 2002).
|
|
Regarding silica exposures of LA, the important variables were the presence of wet dust suppression and indoor work. Wet dust suppression reduced silica exposures about 3-fold (P = 0.015) and indoor work increased exposures about 4-fold (P = 0.001). The 3-fold reduction in silica exposure with wet dust suppression is in line with the results of Thorpe et al. (1999), who reported a 3- to 7-fold reduction during wet use of cut-off saws.
In comparing effects of covariates upon exposures for OE and LA, it is interesting that wet dust suppression significantly reduced silica levels for LA but not for OE. Since LA tend to work close to the point of dust generation, it is perhaps reasonable to expect the use of water to reduce dust levels more for LA than for OE, who sit in cabs 2–3 m above ground level. Yet, despite the ~3-fold reduction in silica exposures of LA using wet dust suppression, the predicted geometric mean silica levels are still quite large (0.082 mg/m3 outdoors and 0.373 mg/m3 indoors) relative to the OEL of 0.05 mg/m3. Thus, wet dust suppression would be unlikely to reduce silica exposures of LA sufficiently to provide an acceptable level of overexposure relative to an OEL of 0.05 mg/m3. On the other hand, OE working in ventilated cabs had a much lower predicted geometric mean than those in open cabs (0.010 versus 0.065 mg/m3) and approach exposure levels that would be acceptable when A = 0.10.
The variance components for exposures to silica, estimated under full Models (1) and (2), are compared in Table 8. For both OE and LA, introduction of significant covariates only moderately reduced 
(15–19% reduction). In a more extensive investigation of three occupational datasets, Peretz et al. (2002) reported reductions in 
of 35–80%, after numerous fixed covariates were added to mixed models. More work is needed to characterize the sites, activities and equipment used by construction workers so that the sources of exposure heterogeneity within trades can be more fully understood.
|
|
CONCLUSIONS AND RECOMMENDATIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTS AND DISCUSSIONCONCLUSIONS AND RECOMMENDATIONSREFERENCES |
|---|
These analyses of respirable dust and silica exposures in the construction industry provide substantial evidence that construction workers are overexposed relative to the current OELs. In fact, the simple juxtaposition of measurements on the OEL of 0.05 mg/m3 for respirable silica (Fig. 1) is sufficient to document a pervasive problem with silica exposure in the construction industry. Application of mixed models to the datasets reinforces this notion by showing that typical workers in these trades are very likely to be exposed on average to levels above the OELs (Table 5). Indeed, point estimates of the probability of overexposure for respirable silica approached unity (
= 0.645–1.00) in all trades and the trade mean exposures ( 
= 0.210–3.86 mg/m3) were much greater than the OEL. Since respirable silica is a potent toxin and suspected carcinogen of the lung, such grossly unacceptable exposures across many US construction sites portend a serious health threat requiring determined action. Indeed, all tools in the occupational hygienists’ arsenal should be brought to bear on the problem, including substitution of materials, active methods of dust suppression, ventilation, redesign of equipment and short-term use of respiratory protection. Since construction work involves diverse, predominantly outdoor sites and is largely intermittent in nature, this presents an important challenge for the industry.
Despite the clear need to reduce construction workers’ exposures overall, particular control strategies depend, in part, upon the construction trade. The painters in our study all performed abrasive blasting while wearing supplied air respirators and, thus, would have been exposed to much lower silica levels than indicated by the personal sampling. Nonetheless, the enormous silica levels observed during blasting (Fig. 1) raise the question as to whether respiratory protection can be relied upon to sufficiently protect these workers and we recommend that only respirators affording the highest levels of protection (i.e. pressure-demand or positive pressure) be used. It is also important that other workers be isolated from abrasive blasting, because even short periods of silica exposure in the observed range would lead to 8 h averages >0.05 mg/m3. In addition, the use of sand or silica-containing materials for abrasive blasting should be discontinued and research should be aimed at identifying alternative methods and materials for preparation of surfaces.
Of even greater concern are the large silica exposures observed among BL and LA, who rarely wore respirators in our study. Given that workers in these trades were routinely exposed to silica in the range 0.1–1 mg/m3 (2–20 times the OEL), substantial reductions in exposure will be required to achieve acceptable levels. In fact, the magnitudes of silica exposure among BL and LA were sufficient to preclude the use of respirators with protection factors of <10. Thus, it is unlikely that disposable and half-mask respirators (the types most often used in the construction industry) would afford sufficient protection to members of these trades while they are operating mechanized hand tools.
The between-worker variance component tended to be much greater than the within-worker variance component for silica exposures. This indicates that construction work is characterized by large differences in average exposures between members of the same trade, a conclusion reinforced by evidence that more than 10% of the random worker effects for silica exposures in three of the four trades were significantly different from 0. Thus, in addition to developing trade level interventions as indicated above, it will also be necessary to investigate many individual sites, activities and types of equipment to arrive at a truly useful intervention strategy for each trade. Our evaluation of a few workplace characteristics indicates that wet dust suppression reduced silica exposures of LA ~3-fold and that ventilated cabs reduced exposures of OE ~6-fold.
In conclusion, we recommend that until such time as the construction industry is prepared to systematically investigate silica exposures at the thousands of sites active across the land, it must foster inexpensive controls, especially use of water to suppress dust. At the same time, methods for isolating workers from dust generation, such as by use of ventilated cabs or workstations, should be actively pursued. Likewise, new methods for locally ventilating sources of dust generation should be sought since our results could not confirm any benefits from mechanical or dilution ventilation. We note with interest a recent paper showing that local exhaust ventilation reduced dust and silica levels by 70–95% under controlled field conditions (Croteau et al., 2002) and we strongly encourage additional work in this area. Finally, respiratory protection programs should be instituted at every construction site, with immediate emphasis given to indoor activities, where silica levels tend to be the greatest.
Acknowledgements—This work was supported by grant U60/CCU317202 from the National Institute for Occupational Safety and Health. The authors acknowledge the assistance of the following labor organizations: Building and Construction Trades Department (AFL-CIO), International Union of Bricklayers and Allied Craftworkers, Laborers International Union of North America, International Union of Operating Engineers and International Union of Painters and Allied Trades. The authors also appreciate the contributions of those who conducted field surveys, Joe Bugay, Nancy Clark, Ryan Knoph, Jennifer Massa, Mike McClean, Jerry Taggart, Chris Trahan, Joe Ventura, George Weymouth, Sam Zizis and Norm Zuckerman, as well as of Sherri Wilson for assistance with compiling and coding data and of Ken Linch who provided helpful information about the NIOSH surveys.
|
FOOTNOTES |
|---|
* Author to whom correspondence should be addressed. Tel: +1-919-966-5017; fax: +1-919-966-0521; e-mail: smr{at}unc.edu
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTS AND DISCUSSIONCONCLUSIONS AND RECOMMENDATIONSREFERENCES |
|---|
ACGIH. (2002) Threshold limit values and biological exposure indices for chemical substances and physical agents. Cincinnati, OH: ACGIH.
Akbar-Khanzadeh F, Brillhart RL. (2002) Respirable crystalline silica dust exposure during concrete finishing (grinding) using hand-held grinders in the construction industry. Ann Occup Hyg; 46: 341–6.
Bakke B, Stewart P, Eduard W. (2002) Determinants of dust exposure in tunnel construction work. Appl Occup Environ Hyg; 17: 783–96.[CrossRef][Medline]
Blute NA, Woskie SR, Greenspan CA. (1999) Exposure characterization for highway construction. Part 1: Cut and cover and tunnel finish stages. Appl Occup Environ Hyg; 14: 632–41.[CrossRef][Medline]
Croteau GA, Guffey SE, Flanagan ME, Seixas NS. (2002) The effect of local exhaust ventilation controls on dust exposures during concrete cutting and grinding activities. Am Ind Hyg Assoc J; 63: 458–67.
Hornung RW, Reed LD. (1990) Estimation of average concentration in the presence of nondetectable values. Appl Occup Environ Hyg; 5: 132–41.
IARC. (1997) IARC monographs on the evaluation of the carcinogenic risks to humans: silica, some silicates, coal dust and paramid fibrils. Lyon: International Agency for Research on Cancer.
Kromhout H, Symanski E, Rappaport SM. (1993) A comprehensive evaluation of within- and between-worker components of occupational exposure to chemical agents. Ann Occup Hyg; 37: 253–70.
Linch KD. (2002) Respirable concrete dust-silicosis hazard in the construction industry. Appl Occup Environ Hyg; 17: 209–21.[CrossRef][Medline]
Linch KD, Miller W, Althouse RB, Groce DW, Hale JM. (1998) Surveillance of respirable crystalline silica dust using OSHA compliance data (1979–1995). Am J Ind Med; 34: 547–58.[CrossRef][Web of Science][Medline]
Lumens MEGL, Spee T. (2001) Determinants of exposure to respirable quartz dust in the construction industry. Ann Occup Hyg; 45: 585–95.
Lyles RH, Kupper LL, Rappaport SM. (1997) A lognormal distribution-based exposure assessment method for unbalanced data. Ann Occup Hyg; 41: 63–76.
NIOSH. (1975) Criteria for a recommended standard: occupational exposure to crystalline silica. Cincinnati, OH: USGPO.
NIOSH. (1994) NIOSH manual of analytical methods, 4th edn, Publication no. 94-113. Cincinnati, OH: DHHS (NIOSH).
Peretz C, Goldberg P, Kahan E, Grady S, Goren A. (1997) The variability of exposure over time: a prospective longitudinal study. Ann Occup Hyg; 41: 485–500.
Peretz C, Goren A, Smid T, Kromhout H. (2002) Application of mixed-effects models for exposure assessment. Ann Occup Hyg; 46: 69–77.
Rappaport SM. (1991) Assessment of long-term exposures to toxic substances in air. Ann Occup Hyg; 35: 61–121.
Rappaport SM, Lyles RH, Kupper LL. (1995) An exposure-assessments strategy accounting for within- and between-worker sources of variability. Ann Occup Hyg; 39: 469–95.
Rappaport SM, Weaver M, Taylor D, Kupper L, Susi P. (1999) Application of mixed models to assess exposures monitored by construction workers during hot processes. Ann Occup Hyg; 43: 457–69.
Riala R. (1988) Dust and quartz exposure of Finnish construction site cleaners. Ann Occup Hyg; 32: 215–20.
Susi P, Goldberg M, Barnes P, Stafford E. (2000) The use of a task-based exposure assessment model (T-BEAM) for assessment of metal fume exposures during welding and thermal cutting. Appl Occup Environ Hyg; 15: 26–38.[CrossRef][Medline]
Thorpe A, Ritchie AS, Gibson MJ, Brown RC. (1999) Measurements of the effectiveness of dust control on cut-off saws used in the construction industry. Ann Occup Hyg; 43: 443–56.
Weaver MA, Kupper LL, Taylor D, Kromhout H, Susi P, Rappaport SM. (2001) Simultaneous assessment of occupational exposures from multiple worker groups. Ann Occup Hyg; 45: 525–42.
Woskie SR, Kalil A, Bello D, Virji MA. (2002) Exposures to quartz, diesel, dust, and welding fumes during heavy and highway construction. Am Ind Hyg Assoc J; 63: 447–57.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  K. Hagstrom, C. Lundholm, K. Eriksson, and I. Liljelind Variability and Determinants of Wood Dust and Resin Acid Exposure during Wood Pellet Production: Measurement Strategies and Bias in Assessing Exposure-Response Relationships Ann. Hyg., November 1, 2008; 52(8): 685 - 694. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. STEINSVAG, M. BRATVEIT, and B. E. MOEN Exposure to Oil Mist and Oil Vapour During Offshore Drilling in Norway, 1979-2004 Ann. Hyg., March 1, 2006; 50(2): 109 - 122. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. LAVOUE, C. BEAUDRY, N. GOYER, G. PERRAULT, and M. GERIN Investigation of Determinants of Past and Current Exposures to Formaldehyde in the Reconstituted Wood Panel Industry in Quebec Ann. Hyg., October 1, 2005; 49(7): 587 - 602. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. R. BEAMER, S. SHULMAN, A. MAYNARD, D. WILLIAMS, and D. WATKINS Evaluation of Misting Controls to Reduce Respirable Silica Exposure for Brick Cutting Ann. Hyg., August 1, 2005; 49(6): 503 - 510. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
G. A. CROTEAU, M. E. FLANAGAN, J. E. CAMP, and N. S. SEIXAS The Efficacy of Local Exhaust Ventilation for Controlling Dust Exposures During Concrete Surface Grinding Ann. Hyg., August 1, 2004; 48(6): 509 - 518. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. R. FLYNN A Stochastic Differential Equation for Exposure Yields a Beta Distribution Ann. Hyg., April 1, 2004; 48(5): 491 - 497. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H F van der Molen, S J Veenstra, J K Sluiter, and M H W Frings-Dresen World at work: Bricklayers and bricklayers' assistants Occup. Environ. Med., January 1, 2004; 61(1): 89 - 93. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||