Understanding the limitations and application of occupational exposure models in a REACH context

Exposure modeling plays a significant role for regulatory organizations, companies, and professionals involved in assessing and managing occupational health risks in workplaces. One context in which occupational exposure models are particularly relevant is the REACH Regulation in the European Union (Regulation (EC) No 1907/2006). This commentary describes the models for the occupational inhalation exposure assessment of chemicals within the REACH framework, their theoretical background, applications, and limitations, as well as the latest developments and priorities for model improvement. Summing up the debate, despite its relevance and importance in the context of REACH not being in question, occupational exposure modeling needs to be improved in many respects. There is a need to reach a wide consensus on several key issues (e.g., the theoretical background and the reliability of modeling tools), to consolidate and monitor model performance and regulatory acceptance, and to align practices and policies regarding exposure modeling.

Evaluation of Stoffenmanager® and ART for Estimating Occupational Inhalation Exposures to Volatile Liquids

In practice, workers often handle the same chemical(s) of interest under different control measures (e.g. local ventilation, enclosed system) during a full shift. Stoffenmanager® allows users to predict either task-based or full-shift exposures. However, most previous studies evaluated the tool by comparing task-based exposures with measured exposures. Also, limited evaluation studies of the Advanced REACH Tool (ART) with the Bayesian approach (ART+B) are available, requiring additional evaluation studies. The performance of Stoffenmanager® and ART with and without the Bayesian approach was evaluated with measured full-shift exposures to volatile liquids in terms of accuracy, precision, and conservatism. Forty-two exposure situation scenarios (including 251 exposures), developed based on job tasks and chemicals handled during tasks from workplaces, were used to generate full-shift estimates. The estimates were then compared with measured exposures using various comparison methods. Overall, Stoffenmanager® appeared to be the most accurate among the testing tools, while ART+B was the most precise. The percentage of measured exposures exceeding the tools' 90th percentile estimates (%M>T) demonstrated that Stoffenmanager® (16%M>T) and ART+B (13%M>T) were more conservative than ART (41%M>T). When the 90% upper confidence limit of the 90th percentile estimate was considered, the level of conservatism changed from low (41%M>T) to medium (17%M>T) for ART and from medium (13%M>T) to high (0.8%M>T) for ART+B. The findings of this study indicate that no single tool would work for all ESs. Thus, it is recommended that users select a tool based on the performance results of three components (i.e. accuracy, precision, and conservatism), not depending on one or two components. The strength of this study is that the required tools' input parameters were obtained during the sample collection to minimize assumptions for many input parameters. In addition, unlike other previous studies, multiple subtasks, which happen often in workplaces, were incorporated in this study. Nevertheless, the present study did not cover all activities listed in the tools and was limited to volatile liquids, suggesting further studies cover other exposure categories (e.g. solid, metal) and diverse activities.

Evaluating the Theoretical Background of STOFFENMANAGER® and the Advanced REACH Tool

STOFFENMANAGER^® and the Advanced REACH Tool (ART) are recommended tools by the European Chemical Agency for regulatory chemical safety assessment. The models are widely used and accepted within the scientific community. STOFFENMANAGER^® alone has more than 37 000 users globally and more than 310 000 risk assessment have been carried out by 2020. Regardless of their widespread use, this is the first study evaluating the theoretical backgrounds of each model. STOFFENMANAGER^® and ART are based on a modified multiplicative model where an exposure base level (mg m⁻³) is replaced with a dimensionless intrinsic emission score and the exposure modifying factors are replaced with multipliers that are mainly based on subjective categories that are selected by using exposure taxonomy. The intrinsic emission is a unit of concentration to the substance emission potential that represents the concentration generated in a standardized task without local ventilation. Further information or scientific justification for this selection is not provided. The multipliers have mainly discrete values given in natural logarithm steps (…, 0.3, 1, 3, …) that are allocated by expert judgements. The multipliers scientific reasoning or link to physical quantities is not reported. The models calculate a subjective exposure score, which is then translated to an exposure level (mg m⁻³) by using a calibration factor. The calibration factor is assigned by comparing the measured personal exposure levels with the exposure score that is calculated for the respective exposure scenarios. A mixed effect regression model was used to calculate correlation factors for four exposure group [e.g. dusts, vapors, mists (low-volatiles), and solid object/abrasion] by using ~1000 measurements for STOFFENMANAGER^® and 3000 measurements for ART. The measurement data for calibration are collected from different exposure groups. For example, for dusts the calibration data were pooled from exposure measurements sampled from pharmacies, bakeries, construction industry, and so on, which violates the empirical model basic principles. The calibration databases are not publicly available and thus their quality or subjective selections cannot be evaluated. STOFFENMANAGER^® and ART can be classified as subjective categorization tools providing qualitative values as their outputs. By definition, STOFFENMANAGER^® and ART cannot be classified as mechanistic models or empirical models. This modeling algorithm does not reflect the physical concept originally presented for the STOFFENMANAGER^® and ART. A literature review showed that the models have been validated only at the ‘operational analysis’ level that describes the model usability. This review revealed that the accuracy of STOFFENMANAGER^® is in the range of 100 000 and for ART 100. Calibration and validation studies have shown that typical log-transformed predicted exposure concentration and measured exposure levels often exhibit weak Pearson’s correlations (r is <0.6) for both STOFFENMANAGER^® and ART. Based on these limitations and performance departure from regulatory criteria for risk assessment models, it is recommended that STOFFENMANAGER^® and ART regulatory acceptance for chemical safety decision making should be explicitly qualified as to their current deficiencies.

Evaluation of Stoffenmanager and a New Exposure Model for Estimating Occupational Exposure to Styrene in the Fiberglass Reinforced Plastics Lamination Process

This study aims to evaluate occupational exposure models by comparing model estimations of Stoffenmanager, version 8.2, and exposure scores calculated using a new exposure model with personal exposure measurements for styrene used in the fiberglass-reinforced plastic (FRP) lamination processes in Korea. Using the collected exposure measurements (n = 160) with detailed contextual information about the type of process, working conditions, local exhaust ventilation, respiratory protections, and task descriptions, we developed a new model algorithm to estimate the score for occupational exposures on situation level. We assumed that the source of exposure originates from the near field only (within the breathing zone of workers). The new model is designed as a simple formula of multiplying scores for job classification, exposure potential, engineering controls, chemical hazard, and exposure probability and then dividing the score for workplace size. The final score is log-transformed, ranging from 1 to 14, and the exposure category is divided into four ratings: no exposure (1), low (2), medium (3), and high (4) exposures. Using the contextual information, all the parameters and modifying factors are similarly entered into the two models through direct translation and coding processes with expert judgement, and the exposure estimations and scores using the two models are calculated for each situation. Overall bias and precision for Stoffenmanager are −1.00 ± 2.07 (50th) and −0.32 ± 2.32 (90th) for all situations (n = 36), indicating that Stoffenmanager slightly underestimated styrene exposures. Pearson’s correlation coefficients are significantly high for Stoffenmanager (r = 0.87) and the new model (r = 0.88), and the correlation between the two models is significantly high (r = 0.93) (p < 0.01). Therefore, the model estimations using Stoffenmanager and the new model are significantly correlated with the styrene exposures in the FRP lamination process. Further studies are needed to validate and calibrate the models using a larger number of exposure measurements for various substances in the future.

Comparison between Communicated and Calculated Exposure Estimates Obtained through Three Modeling Tools

This study aims to evaluate the risk assessment approach of the REACH legislation in industrial chemical departments with a focus on the use of three models to calculate exposures, and discuss those factors that can determine a bias between the estimated exposure (and therefore the expected risk) in the extended safety data sheets (e-SDS) and the expected exposure for the actual scenario. To purse this goal, the exposure estimates and risk characterization ratios (RCRs) of registered exposure scenarios (ES; “communicated exposure” and “communicated RCR”) were compared with the exposure estimates and the corresponding RCRs calculated for the actual, observed ES, using recommended tools for the evaluation of exposure assessment and in particular the following tools: (i) the European Centre for Ecotoxicology and Toxicology of Chemicals Targeted Risk Assessment v.3.1 (ECETOC TRA), (ii) STOFFENMANAGER^® v.8.0 and (iii) the Advanced REACH Tool (ART). We evaluated 49 scenarios in three companies handling chemicals. Risk characterization ratios (RCRs) were calculated by dividing estimated exposures by derived no-effect levels (DNELs). Although the calculated exposure and RCRs generally were lower than communicated, the correlation between communicated and calculated exposures and RCRs was generally poor, indicating that the generic registered scenarios do not reflect actual working, exposure and risk conditions. Further, some observed scenarios resulted in calculated exposure values and RCR higher than those communicated through chemicals’ e-SDSs; thus ‘false safe’ scenarios (calculated RCRs > 1) were also observed. Overall, the obtained evidences contribute to doubt about whether the risk assessment should be performed using generic (communicated by suppliers) ES with insufficient detail of the specific scenario at all companies. Contrariwise, evidences suggested that it would be safer for downstream users to perform scenario-specific evaluations, by means of proper scaling approach, to achieve more representative estimates of chemical risk.

How to Obtain a Reliable Estimate of Occupational Exposure? Review and Discussion of Models' Reliability

Evaluation and validation studies of quantitative exposure models for occupational exposure assessment are still scarce and generally only consider a limited number of exposure scenarios. The aim of this review was to report the current state of knowledge of models’ reliability in terms of precision, accuracy, and robustness. A systematic review was performed through searches of major scientific databases (Web of Science, Scopus, and PubMed), concerning reliability of Tier1 (“ECETOC TRA”-European Centre for Ecotoxicology and Toxicology of Chemicals Targeted Risk Assessment, MEASE, and EMKG-Expo-Tool) and Tier2 models (STOFFENMANAGER^® and “ART”-Advanced Registration, Evaluation, Authorization and Restriction of Chemicals (REACH) Tool). Forty-five studies were identified, and we report the complete information concerning model performance in different exposure scenarios, as well as between-user reliability. Different studies describe the ECETOC TRA model as insufficient conservative to be a Tier1 model, in different exposure scenarios. Contrariwise, MEASE and EMKG-Expo-Tool seem to be conservative enough, even if these models have not been deeply evaluated. STOFFENMANAGER^® resulted the most balanced and robust model. Finally, ART was generally found to be the most accurate and precise model, with a medium level of conservatism. Overall, the results showed that no complete evaluation of the models has been conducted, suggesting the need for correct and harmonized validation of these tools.

Health risk assessment from exposure to particles during packing in working environments

Packing of raw materials in work environments is a known source of potential health impacts (respiratory, cardiovascular) due to exposure to airborne particles. This activity was selected to test different exposure and risk assessment tools, aiming to understand the effectiveness of source enclosure as a strategy to mitigate particle release. Worker exposure to particle mass and number concentrations was monitored during packing of 7 ceramic materials in 3 packing lines in different settings, with low (L), medium (M) and high (H) degrees of source enclosure. Results showed that packing lines L and M significantly increased exposure concentrations (119-609 μg m^-3 respirable, 1150-4705 μg m^-3 inhalable, 24,755-51,645 cm^-3 particle number), while non-significant increases were detected in line H. These results evidence the effectiveness of source enclosure as a mitigation strategy, in the case of packing of ceramic materials. Total deposited particle surface area during packing ranged between 5.4 and 11.8 × 10⁵ μm² min^-1, with particles depositing mainly in the alveoli (51-64%) followed by head airways (27-41%) and trachea bronchi (7-10%). The comparison between the results from different risk assessment tools (Stoffenmanager, ART, NanoSafer) and the actual measured exposure concentrations evidenced that all of the tools overestimated exposure concentrations, by factors of 1.5-8. Further research is necessary to bridge the current gap between measured and modelled health risk assessments.

Testing the performance of one and two box models as tools for risk assessment of particle exposure during packing of inorganic fertilizer

Modelling of particle exposure is a useful tool for preliminary exposure assessment in workplaces with low and high exposure concentrations. However, actual exposure measurements are needed to assess models reliability. Worker exposure was monitored during packing of an inorganic granulate fertilizer at industrial scale using small and big bags. Particle concentrations were modelled with one and two box models, where the emission source was estimated with the fertilizer's dustiness index. The exposure levels were used to calculate inhaled dose rates and test accuracy of the exposure modellings. The particle number concentrations were measured from worker area by using a mobility and optical particle sizer which were used to calculate surface area and mass concentrations. The concentrations in the worker area during pre-activity ranged 63,797-81,073 cm^-3, 4.6 × 10⁶ to 7.5 × 10⁶ μm² cm^-3, and 354 to 634 μg m^-3 (respirable mass fraction) and during packing 50,300 to 85,949 cm^-3, 4.3 × 10⁶ to 7.6 × 10⁶ μm² cm^-3, and 279 to 668 μg m^-3 (respirable mass fraction). Thus, the packing process did not significantly increase the exposure levels. Chemical exposure was also under control based on REACH standards. The particle surface area deposition rate in respiratory tract was up to 7.6 × 10⁶ μm² min^-1 during packing, with 52%-61% of deposition occurring in the alveolar region. Ratios of the modelled and measured concentrations were 0.98 ± 0.19 and 0.84 ± 0.12 for small and big bags, respectively, when using the one box model, and 0.88 ± 0.25 and 0.82 ± 0.12, when using the two box model. The modelling precision improved for both models when outdoor particle concentrations were included. This study shows that exposure concentrations in a low emission industrial scenario, e.g. during packing of a fertilizer, can be predicted with a reasonable accuracy by using the concept of dustiness and mass balance models.

The general ventilation multipliers calculated by using a standard Near-Field/Far-Field model

In conceptual exposure models, the transmission of pollutants in an imperfectly mixed room is usually described with general ventilation multipliers. This is the approach used in the Advanced REACH Tool (ART) and Stoffenmanager® exposure assessment tools. The multipliers used in these tools were reported by Cherrie (1999; http://dx.doi.org/10.1080/104732299302530 ) and Cherrie et al. (2011; http://dx.doi.org/10.1093/annhyg/mer092 ) who developed them by positing input values for a standard Near-Field/Far-Field (NF/FF) model and then calculating concentration ratios between NF and FF concentrations. This study revisited the calculations that produce the multipliers used in ART and Stoffenmanager and found that the recalculated general ventilation multipliers were up to 2.8 times (280%) higher than the values reported by Cherrie (1999) and the recalculated NF and FF multipliers for 1-hr exposure were up to 1.2 times (17%) smaller and for 8-hr exposure up to 1.7 times (41%) smaller than the values reported by Cherrie et al. (2011). Considering that Stoffenmanager and the ART are classified as higher-tier regulatory exposure assessment tools, the errors is general ventilation multipliers should not be ignored. We recommend revising the general ventilation multipliers. A better solution is to integrate the NF/FF model to Stoffenmanager and the ART.