Quartz and Respirable Dust in the Dutch Construction Industry: A Baseline Exposure Assessment as Part of a Multidimensional Intervention Approach

ABSTRACT

Quartz exposure can cause several respiratory health effects. Although quartz exposure has been described in several observational workplace studies, well-designed intervention studies that investigate the effect of control strategies are lacking. This article describes a baseline exposure study that is part of a multidimensional intervention program aiming to reduce quartz exposure among construction workers. In this study, personal respirable dust and quartz exposure was assessed among 116 construction workers (bricklayers, carpenters, concrete drillers, demolishers, and tuck pointers). Possible determinants of exposure, like job, tasks, and work practices, use of control measures, and organizational and psychosocial factors, were explored using exposure models for respirable dust and quartz separately. Stratified analyses by job title were performed to evaluate the effect of control measures on exposure and to explore the association between control measures and psychosocial factors. Overall, 62% of all measurements exceeded the Dutch occupational exposure limit for quartz and 11% for respirable dust. Concrete drillers and tuck pointers had the highest exposures for quartz and respirable dust (0.20 and 3.43mg m−3, respectively). Significant predictors of elevated quartz exposure were abrasive tasks and type of material worked on. Surprisingly, in a univariate model, an increased knowledge level was associated with an increase in exposure. Although control measures were used infrequently, if used they resulted in approximately 40% reduction in quartz exposure among concrete drillers and tuck pointers. Only among concrete drillers, the use of control measures was associated with a higher score for social influence (factor 1.6); knowledge showed an inverse association with use of control measures for concrete drillers, demolishers, and tuck pointers. In conclusion, the detailed information on determinants of exposure, use of control measures, and constraints to use these control measures can be used for the determination and systematic prioritization of intervention measures used to design and implement our intervention strategy. This study underlines the need for multidisciplinary workplace exposure control strategies although larger study populations are necessary to determine a possible causal association between organizational and psychosocial factors and psychosocial factors and control measures.

METHODS

Study population

The study population consisted of construction workers from eight construction companies, recruited via branch organizations. Companies were relatively specialized and had a limited number of job titles present. All companies and individual workers gave oral and written informed consent to participate. Relevant job categories were selected based on exposure levels identified in previous studies (Flanagan et al., 2003; Lumens et al., 2009), and in a pilot study, in which we visited a small selection of building sites and performed some exploratory measurements. Besides potential exposure levels, the size of the population was considered, resulting in the selection of the following jobs: bricklayers, carpenters, concrete drillers, demolishers, and tuck pointers. Although carpenters were expected to have lower quartz exposure, this category was included because it is assumed that they may be affected by bystander exposure. A more detailed task description per job category is provided in Table 1.

Management questionnaire focused on organizational factors

Employers were asked to fill in a questionnaire to get insight into the occupational safety and health policy of the company. This questionnaire was adapted from the Occupational Health and Safety Assessment Series (OHSAS) norm for health and safety management by selecting all questions focusing on hazardous substances (Anonymous, 2007). OHSAS is an audit evaluation system to assess whether organizations have a safety management system, documented and working in practice.

Among others, the questionnaire was focused on dust-reducing practices and included topics such as (i) presence and compliance of work procedures and workplace instructions; (ii) training of employees; (iii) management support of proactive health and safety culture (e.g. toolbox meetings on a fixed and regular schedule to stimulate discussion between employers and employees); and (iv) communication and feedback with equipment contractors to improve services.

Personal exposure assessment and assessment of engineering factors

Full-shift personal air samples (mean sampling time 7.3h) were taken from 116 construction workers between November 2011 and February 2012, of whom 22% were sampled repeatedly to enable the evaluation of the within worker variance. Respirable dust samples were collected using Dewell–Higgins cyclones mounted with a PVC filter (Millipore, pore size: 5.0 µm, diameter: 25mm), connected to a calibrated Gillian GilAir pump with an airflow of 2 l min−1. The amount of dust on filters was determined gravimetrically by pre- and post-weighing of filters on an analytical balance (Mettler) in a conditioned room. Filters were acclimatized at least 24h prior to weighing (Health and Safety Executive (HSE), 2000). Quartz was determined by infrared spectroscopy and X-ray diffraction (Health and Safety Executive (HSE), 2005). None of the samples collected had values below the limit of detection (LOD) for respirable dust of 0.15mg, assessed as the average weight difference of the blank filters plus thrice the standard deviation. Seven samples (5%) had values below the analytical LOD for quartz, which was 0.01mg.

Throughout their shift, workers were observed using a structured walk-through survey to obtain detailed information on work activities (including task time), workspace, work practices, the type of tools used, type of material worked on, use of dust-reducing techniques (i.e. control measures), and respiratory protective equipment (RPE).

Worker questionnaire focused on psychosocial factors

To explore the possible role of psychosocial factors, employees were asked to fill in a questionnaire on psychosocial factors potentially related to respirable dust and quartz exposure. Psychosocial factors covered topics like knowledge and beliefs regarding effectiveness of controls, risk perception, social influence, and motivation. A detailed description of these factors is given in Table 2. Knowledge of workers was assessed using a questionnaire with specific quartz dust–related questions. Adapted formats of existing standardized scales were applied to assess motivation, beliefs regarding effectiveness of controls, risk perception and social influence (Ajzen 1991; Geer et al., 2007), and risk propensity (Meertens and Lion, 2008). All psychosocial variables that were considered had a Cronbach value ≥ 0.7, ‘acceptable’ based on the rules of thumb provided by George and Mallery (2010). Cronbach’s alpha is a measure of internal consistency for different items that form a composite variable (Gliem and Gliem, 2003). A few psychosocial factors were included in the assessment (e.g. self-efficacy) but excluded in this analysis because of unreliable scales, i.e. low Cronbach value. All psychosocial variables were measured on a ratio scale with a range, as indicated in Table 2. Two separate questions addressed factors that may be (perceived as) a constraint or facilitator for a worker to perform dust-reducing work practices. In order to gain insight into the feasibility and reliability of the questionnaire, it was tested in a pilot study among 25 workers not involved in the next study phases.

Statistical analyses

All statistical analyses were performed using SAS v9.3 (SAS Institute Inc.). Samples below the analytical LOD of quartz were assigned a value of two thirds of the detection limit. All exposure data showed a skewed distribution, requiring the exposure data to be log transformed prior to statistical analysis. Descriptive statistics were calculated per job category. Spearman correlations between respirable dust and quartz concentrations were calculated. Determinants of exposure to respirable dust and quartz were explored using mixed effects models (PROC MIXED) in order to correct for possible correlation between repeated measurements (Rappaport et al., 1999). Worker was considered as random effect, whereas job category, task, work characteristics (e.g. worker-source orientation and worker-source distance), and organizational and psychosocial factors were introduced as fixed effects. Fixed effects were included as dichotomized variables, except psychosocial factors that were included as continuous variables. Variances were estimated as between-worker and within-worker variance components.

Model building comprised two steps. In the first step, univariate analyses were used to explore which determinants were associated with exposure to respirable dust and/or quartz. In the second step, determinants that were significantly associated with exposure were introduced stepwise into the model. For model building purposes, the level of significance for inclusion in the model was set at P < 0.10.

Separate analyses were conducted to evaluate the effect of control measures, stratified by job category, because the use of specific control measures varied by job category. In addition, associations between psychosocial factors and the use of control measures were explored by job category using linear regression models (PROC REG).

RESULTS

Population characteristics

The study population consisted of male construction workers only (Table 3). Concrete drillers and tuck pointers were generally younger and had less work experience than workers in the other job categories. Carpenters had received vocational training most frequently, whereas concrete drillers were least often vocationally trained (86 versus 13%, respectively).

Organizational factors

Participating companies in this study employed mainly permanent employees. Five out of eight companies stated that compliance with work procedures and workplace instructions regarding dust-reducing practices was supervised by their management. Two companies actively provided training regarding dust-reducing work practices to their personnel. Within three companies, management-supported proactive health and safety culture focused on dust-reducing practices. Five companies consulted with their equipment contractors on improving their services regarding dust-reducing practices.

Personal exposure levels

In total, 149 full-shift personal samples were collected from 116 workers. Table 3 shows the average exposure levels for each job category. Exposure to respirable dust was highest among tuck pointers, whereas concrete drillers were exposed to the highest quartz levels. The overall geometric mean (GM) was 0.88mg m−3 for respirable dust [geometric standard deviation (GSD) 4.23] and 0.10mg m−3 for quartz (GSD 3.84). The overall correlation coefficient (r) between respirable dust and quartz was 0.76 (range 0.24–0.84 for different job categories), with the largest association for concrete drillers. The full-shift exposure measurements showed quartz concentrations exceeding the Dutch OEL for quartz (0.075mg m−3) in 62% of the measurements, whereas the Dutch OEL for respirable dust (5.00mg m−3) was exceeded in 11% of the measurements.

Psychosocial factors

The questionnaire regarding psychosocial factors was administered to all 116 workers. However, eighteen workers did not fill in the questionnaire because of language barriers (n = 2) or lack of interest. Among all workers who completed the questionnaire, two main reasons were mentioned to use control measures and/or to perform dust-reducing work practices if possible: ‘It is better for my own health’ (89%) and ‘It is less inconvenient for my eyes or my airways’ (67%). Ergonomically poorly designed tools that adversely affected their productivity were mentioned as the most important constraint to apply dust-reducing work practices.

Overall, carpenters had the highest score for knowledge regarding dust exposure and its possible health effects (0.8 on a scale of 0 to 1). Tuck pointers had the lowest scores for risk perception, beliefs regarding effectiveness of controls, social influence, and motivation compared with other job categories, whereas concrete drillers and demolishers had the highest scores. Psychosocial factor scores are presented in Table 3. Because we hypothesized that the way a company deals with their occupational safety and health policy may affect psychosocial aspects of the workers, we investigated the association between organizational and psychosocial factors. Concrete drillers were present in three companies. All of these three companies offered training on dust reduction practices to their employees, probably resulting in a similar score (0.7 on a scale of 0 to 1) for knowledge of their employees. One company employing concrete drillers did not supervise on compliance of work procedures and workplace instructions, which may be indicative for the lower score on social influence (3.5 on a scale of 1–5) compared with concrete drillers with supervision in two other companies (4.2 and 4.3 on a scale of 1–5). A similar indicative, but weak association was found when comparing tuck pointers among three companies with regard to social influence. None of these associations were statistically significant (data not shown).

Determinants of exposure

Table 4 shows the final mixed effects models for respirable dust and quartz exposure. Concrete drillers had on average a 40 times higher exposure to quartz than the reference group, ‘bricklayers’. Working indoors resulted in approximately 4.5 times higher exposure to respirable dust and quartz. The activities sanding and drilling were significantly associated with elevated respirable dust and quartz exposure (factors ranging from 1.5 to 4.5). Exposure to respirable dust increased when working at shoulder level or near-field worker-source distance. Integrated water suppression and the spraying of water resulted in almost 1.5 times lower exposure to quartz, whereas integrated local exhaust ventilation (LEV) reduced exposure to respirable dust and quartz by a factor 2. Use of a stationary extraction unit was not significantly associated with exposure. All determinants [e.g. task, product, work practices (i.e. worker-source orientation and worker-source distance), and control measure] by job category as well as their association with exposure in the univariate models are presented in the Supplementary data, available at Annals of Occupational Hygiene online.

Organizational factors did not show any relation with exposure, in both the univariate and multivariate models. The psychosocial factors also were not associated with exposure in the multivariate models. However, the univariate models did show some significant associations. Because we were explicitly interested in psychosocial factors, the results for the univariate associations are shown in Table 5. Only for one of the psychosocial factors, i.e. knowledge, there were indications for an association with respirable dust (P = 0.03) and quartz (P = 0.07) exposure in a univariate model, with increasing exposure levels when the knowledge level increased. Larger social influence seemed associated with increased quartz exposure levels only (P < 0.01). The final models explained 72 and 83% of the between-worker variance, and 69 and 22% of the day-to-day variability for respirable dust and quartz, respectively.

Control measures

Because the use of control measures was correlated with job category, we separately analysed the effect of control measure on exposure by job category. Control measures were particularly used during abrasive tasks among certain job categories, i.e. concrete driller, demolisher, and tuck pointer. The effect of control measures by job category is shown in Table 6. Examples of the control measures that were available during the field study are shown in Fig. 1. Concrete drillers used control measures more frequently (83%) than demolishers (38%) and tuck pointers (28%). Tool-integrated water suppression, used by concrete drillers during drilling and sawing and characterized by a hose connection on the tool, resulted in a nonsignificant 40% reduction in quartz exposure (P > 0.10). For tuck pointers, the use of industrial vacuum cleaners fitted to a centralized extraction ventilation system showed a borderline significant exposure reduction of 60% for respirable dust and 45% for quartz (P = 0.11). Demolishers using a stationary extraction unit had a 9-fold higher exposure to respirable dust compared with demolishers not using this type of control measure. Spraying water as control measure demonstrated a borderline significant reduction in exposure to respirable dust (P = 0.10) among demolishers.

Psychosocial aspects and use of control measures

We explored whether the use of control measures was associated with psychosocial factors by job category because this association might explain why psychosocial factors were not significantly associated with exposure in the final multivariate model. Surprisingly, concrete drillers, demolishers, and tuck pointers with a higher knowledge level less often used control measures (factor 0.3–0.8; P > 0.10). Only among concrete drillers, a higher score for social influence from colleagues or supervisors seemed to result in a significant 1.6 increase (P < 0.001) in the use of control measures (data not shown).
