Annals of Occupational Hygiene Advance Access originally published online on March 15, 2006
Annals of Occupational Hygiene 2006 50(5):469-489; doi:10.1093/annhyg/mel012
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Default Values for Assessment of Potential Dermal Exposure of the Hands to Industrial Chemicals in the Scope of Regulatory Risk Assessments
1 TNO Chemistry, Department of Food and Chemical Risk Analysis PO Box 360, 3700 AJ Zeist, The Netherlands
2 Health and Safety Laboratory Harpur Hill, Buxton, Derbyshire, UK
3 Kuopio Regional Institute of Occupational Health Kuopio, Finland
*Author to whom correspondence should be addressed. Tel: +31-30-6944-733; fax: +31-30-6944-926; e-mail: marquart{at}chemie.tno.nl
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Dermal exposure needs to be addressed in regulatory risk assessment of chemicals. The models used so far are based on very limited data. The EU project RISKOFDERM has gathered a large number of new measurements on dermal exposure to industrial chemicals in various work situations, together with information on possible determinants of exposure. These data and information, together with some non-RISKOFDERM data were used to derive default values for potential dermal exposure of the hands for so-called ‘TGD exposure scenarios’. TGD exposure scenarios have similar values for some very important determinant(s) of dermal exposure, such as amount of substance used. They form narrower bands within the so-called ‘RISKOFDERM scenarios’, which cluster exposure situations according to the same purpose of use of the products. The RISKOFDERM scenarios in turn are narrower bands within the so-called Dermal Exposure Operation units (DEO units) that were defined in the RISKOFDERM project to cluster situations with similar exposure processes and exposure routes. Default values for both reasonable worst case situations and typical situations were derived, both for single datasets and, where possible, for combined datasets that fit the same TGD exposure scenario. The following reasonable worst case potential hand exposures were derived from combined datasets: (i) loading and filling of large containers (or mixers) with large amounts (many litres) of liquids: 11 500 mg per scenario (14 mg cm–2 per scenario with surface of the hands assumed to be 820 cm2); (ii) careful mixing of small quantities (tens of grams in <1l): 4.1 mg per scenario (0.005 mg cm–2 per scenario); (iii) spreading of (viscous) liquids with a comb on a large surface area: 130 mg per scenario (0.16 mg cm–2 per scenario); (iv) brushing and rolling of (relatively viscous) liquid products on surfaces: 6500 mg per scenario (8 mg cm–2 per scenario) and (v) spraying large amounts of liquids (paints, cleaning products) on large areas: 12 000 mg per scenario (14 mg cm–2 per scenario). These default values are considered useful for estimating exposure for similar substances in similar situations with low uncertainty. Several other default values based on single datasets can also be used, but lead to estimates with a higher uncertainty, due to their more limited basis. Sufficient analogy in all described parameters of the scenario, including duration, is needed to enable proper use of the default values. The default values lead to similar estimates as the RISKOFDERM dermal exposure model that was based on the same datasets, but uses very different parameters. Both approaches are preferred over older general models, such as EASE, that are not based on data from actual dermal exposure situations.
Keywords: default values • dermal exposure • exposure models • industrial chemicals • regulatory risk assessment
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Dermal exposure is one of the issues that need to be addressed in the assessment of exposure for regulatory risk assessment of chemicals. In risk assessment of agricultural pesticides, the skin has been recognized to be a major route of exposure for many years now. Therefore, several models and databases exist to enable dermal exposure assessment for pesticide handlers, workers and bystanders (van Hemmen, 1992; Hamey, 1995; EUROPOEM, 2002). More recently, models have been developed for dermal exposure to biocides (ECB, 2002a).
The importance of dermal exposure has not been so clear for industrial chemicals. Some models have been created and used for dermal exposure of workers to industrial chemicals. These include general models such as a model used by US EPA (IT Environmental Programs, Inc., 1991) and EASE (Estimation and Assessment of Substance Exposure; ECB, 2003a) and some very specific models (Brouwer et al., 2001; Semple et al., 2001). The general models, used in regulatory risk assessment processes, are largely based on only a few experimental data and have not been validated to any reasonable degree. The dermal part of the EASE model is considered not to be based on sufficient scientific knowledge and therefore not to be very useful (Cherrie et al., 2003). The very specific models require detailed and specific information that is generally not available to exposure assessors in the scope of regulatory risk assessment. The need for better dermal exposure models, which are based on data from practical dermal exposure situations, is therefore urgent (Marquart et al., 1994, 2001; Dost, 1995; Fenske, 2000; Boeniger, 2003; van Hemmen, 2004). This was one of the reasons for starting the RISKOFDERM project (EU Fifth Framework Program, project QLK4-CT-1999-01107). This project had two main goals (van Hemmen et al., 2003): (i) the creation of dermal exposure model(s) for use in regulatory risk assessment of substances and (ii) the creation of a toolkit for risk assessment and risk management of dermal exposure in small- and medium-sized enterprises (SMEs). The work on dermal exposure model(s) for regulatory risk assessment resulted in two separate approaches. The first approach is a model, based on statistical relations between measured data and potential exposure determinants. The second approach is the derivation of so-called ‘default values’ for relatively well described situations, based on a subset of the measured data. The latter approach is the subject of this publication. The former approach is described in a companion paper (Warren et al., 2006).
Default values are used extensively in risk assessment. It is an efficient way of linking exposure situations to exposure levels or rates, because there is no need for independent interpretation of all data in every assessment. The use of agreed default values also provides the opportunity to harmonize and standardize the interpretation of measured data by different assessors and to increase the transparency of the assessments.
The purpose of this publication is to propose such a set of default values for potential dermal exposure of the hands. Data from RISKOFDERM are also available for body exposure. However, only hand exposure default values are presented here, because only hand exposure is presently assessed for new and existing substances.
The RISKOFDERM project
Fifteen research groups from 10 member states of the European Union participated in the RISKOFDERM project. It consisted of four separate, but highly linked, work parts (van Hemmen et al., 2003):
- Qualitative assessment of dermal exposure in a large number of situations;
- Quantitative assessment of dermal exposure in a smaller number of situations;
- Creation of benchmarked dermal exposure model(s) for regulatory risk assessment;
- Creation of a toolkit for SMEs for assessment and management of dermal exposure.
Possible determinants of dermal exposure were studied by the analysis of available literature and unpublished data. This resulted in a number of parameters to be studied, together with measurements of dermal exposure, to allow modelling of dermal exposure (Marquart et al., 2003). A structured questionnaire/checklist was developed based on a previous version (Hebisch and Auffarth, 2001) to gather qualitative information on substances used, methods of use, routes of exposure, exposure control measures and other parameters needed for modelling, as well as on expected or observed contamination of the skin. The questionnaire was used by trained observers for gathering qualitative information in a large number of situations (Auffarth et al., 2003). Measurements were done by several research groups in exposure scenarios covering a large number of dermal exposure situations. Potential dermal exposure was measured (i.e. all contamination reaching the skin or the covering clothing or gloves) separately for hands and body. Methods and results of these measurements were reported in a number of publications (Delgado et al., 2004; Eriksson and Wiklund, 2004; Eriksson et al., 2004; Fransman et al., 2004; Gijsbers et al., 2004; Hughson and Aitken, 2004; Mäkinen and Linnainmaa, 2004a,b; Roff et al., 2004a,b). During these measurements, a modified version of the structured questionnaire/checklist was used to gather information on parameters needed for modelling. Next to the data reported in the above-mentioned publications, additional RISKOFDERM data, gathered in the process of benchmarking the preliminary dermal exposure models, were also used in the derivation of default values. Details from these additional datasets have not been published in the scientific journals (yet), but are available in the so-called ‘deliverables’ of RISKOFDERM (RISKOFDERM DL 40, 2003; RISKOFDERM DL 41, 2003; RISKOFDERM DL 42, 2003; RISKOFDERM DL 43, 2003). The additional datasets regarded the scenarios presented in Table 1. All RISKOFDERM data were combined with other useful measured data and information available to the researchers for statistical modelling of dermal exposure (Warren et al., 2006).
|
Apart from RISKOFDERM data, a small group of other datasets, previously used in drafting the Technical Guidance Document on Risk Assessment (for new and existing chemicals; TGD) (ECB, 2003a) and reproduced in Fehrenbacher et al. (2003) was used in drafting the default values. Details on methods used in these studies are reported in the original reports, which have not been published in a scientific journal (Lansink et al., 1996, 1998; HSE, 1999; Preller and Schipper, 1999).
DEO units, RISKOFDERM scenarios and TGD exposure scenarios
The complex realm of dermal exposure in all possible work situations cannot be described by a simple general model or a simple set of default values. However, it is also impossible to construct an infinite number of separate models or exposure ranges for all possible situations. Therefore, the conceptual model of dermal exposure published by Schneider et al. (1999) was used in RISKOFDERM as a basis for clustering all possible exposure situations in so-called Dermal Exposure Operation units (DEO units). The main purpose of the DEO units is to cluster dermal exposure situations with similar routes of exposure and similar exposure sources (Table 2) and with similar (expected) relations between exposure determinants and exposure levels (van Hemmen et al., 2003). The DEO units may theoretically lead to perfect separation of one work situation from another. However, in practice, boundaries between DEO units are not clear-cut. For example, almost all work situations will involve some kind of ‘handling of objects’. The handling of a can of paint and a spray gun cannot be fully separated from the actual spraying of paint. However, as far as possible, the filling of a spray gun by pouring paint from a container into the reservoir of the spray gun would be classified as ‘handling of objects’, while the actual spraying would be classified as ‘spray dispersion of products’, although this would involve holding the spray gun, moving about the compressed air line and perhaps handling the object to be sprayed to reach all parts of the object. Similarly, ‘immersion’ of an object in a product always includes an element of ‘handling of objects’.
|
DEO units still cover highly variable situations with regard to substance and product (e.g. liquid, powder), amount of product used and exposure control (ventilation, personal protective equipment). Furthermore, within a DEO unit there may be substantial differences in techniques and specific tasks of workers. Spreading of a product with a hand-held tool can, e.g. be done by a brush, a rake, a spatula or a roller. This may lead to substantial differences in contamination of the tool, in probability of splashing and in potential for contact with contaminated surfaces. Therefore, there may be substantial differences in exposure levels or exposure rates within a DEO unit. A further division of the DEO units was therefore made into clusters of situations with more similarities (called ‘scenarios’). This clustering was largely based on the purpose of the activity (including handling of a product or object) that leads to possible exposure. For the purpose of this publication, these clustered situations will be called ‘RISKOFDERM scenarios’. Statistical evaluations of the RISKOFDERM scenarios and the DEO units were published by Rajan-Sithamparanadarajah et al. (2004) and Kromhout et al. (2004). The term ‘scenario’ can have very different interpretations in another context, even in the scope of risk assessment. The TGD (ECB, 2003a) defines an ‘exposure scenario’ as follows: ‘An exposure scenario is the set of information and/or assumptions that tells us how the contact between the worker and the substance takes place. It is based on the most important characteristics of the substance in the view of occupational exposure, e.g. the physical state, the vapour pressure as well as on its uses, processes, tasks (description, duration, frequency of exposure) and controls’. This definition may lead to more precisely defined scenarios than the RISKOFDERM scenarios. Such more precisely defined scenarios will be called ‘TGD exposure scenarios’ in this publication. For instance, the RISKOFDERM scenario ‘sawing’ (scenario 6.4 in Rajan-Sithamparanadarajah et al., 2004) might contain several TGD exposure scenarios, including, e.g. ‘sawing of trees with a chain saw’, ‘sawing of wood with a hand saw without ventilation’ or ‘sawing of wood with a circular saw under local exhaust ventilation’. The current risk assessments of existing substances in Europe use more or less precisely defined scenarios. Some are very specific, such as ‘Decomposition of photoresist materials during the production of integrated circuits’ (ECB, 2002b), while others are very broad, often including several RISKOFDERM scenarios, generally within one DEO unit. An example of the latter is ‘Production of products containing dibutyl phthalate’ (ECB, 2003c), which largely is in DEO unit ‘Handling of objects’ and may include several of the RISKOFDERM scenarios in that DEO unit, such as ‘transferring, transporting’, ‘coupling/decoupling of transfer lines’ and ‘loading (of liquids)’. The conclusions on such scenarios may also include exposures that would be part of other DEO units, such as ‘Dispersion of products with a hand-held tool’, e.g. when spilled amounts of liquid are removed from the floor by mopping. For the purpose of this publication it is essential to keep in mind that the RISKOFDERM scenarios, the TGD exposure scenarios and the loosely defined scenarios that are often used in risk assessment reports can be very different (groups of) situations, even if the name is similar.
It was not intended by the RISKOFDERM project to define an exhaustive list of all possibly relevant RISKOFDERM scenarios within each DEO unit. Table 3 shows the list of RISKOFDERM scenarios within the DEO units distinguished in the project.
|
The RISKOFDERM scenarios should, under practical conditions and in real work situations be observable, measurable, attributable to the task within the scenario and applicable to many industries (van Hemmen et al., 2003). The activities considered to constitute the RISKOFDERM scenarios were described to ensure a reasonable coherence within the RISKOFDERM scenarios. As an example, the RISKOFDERM scenario ‘rolling’ that is part of DEO unit ‘Dispersion of products with a hand-held tool’ was considered to be constituted of the activities: (i) ‘dipping of roller into product’, and/or (ii) ‘spraying of product onto surface’ and (iii) ‘dispersion of product with roller’. By breaking down RISKOFDERM scenarios into these activities, a more or less equivalent interpretation of the terms used was reached between the international research groups doing measurements in this project. Apart from specific activities for specific RISKOFDERM scenarios, some more general activities that can occur in several RISKOFDERM scenarios (under several DEO units) were also defined. These included several variants of the general activities ‘manually transferring object(s) from one position to another position’ and ‘cleaning equipment used for scenario’ (RISKOFDERM DL 17, 2003).
Default values for exposure scenarios
The approach of deriving default values of exposure (in mass, loading (mg cm–2) or rates of these) for exposure scenarios from actual measured data has been used before, both for inhalation and for dermal exposure. They are sometimes referred to as ‘surrogate exposure values’ (van Hemmen, 1993). The general approach is first to gather independent datasets with reliable quality (Tielemans et al., 2002). Datasets are grouped according to scenario. Within a scenario, a further grouping may be done per subscenario, e.g. liquids and powders are grouped separately. The output (i.e. exposure parameters) of the datasets is converted, where necessary, to the same units (e.g. mg, mg m–3 or µg cm–2 h–1). Then the datasets are presented as a group of datasets. This can be done by compiling data points and/or parameters of distributions in graphs, as done for pesticides by van Hemmen (1992) or for filling of liquids into containers by Beijer et al. (1998). Another method is to compile relevant parameters of the exposure distribution of each dataset per scenario into a table, very similar to the presentation of exposure data for specific scenarios in the risk assessment of existing substances in Europe (see e.g. Table 4.1 in the EU Risk Assessment Report on Methyl Acetate; ECB, 2003b). The compiled evaluation is then further assessed to estimate values to be used for risk assessment. One value often estimated is the ‘typical’ or ‘normal’ or ‘average’ value. In the TGD (ECB, 2003a) the ‘Typical exposure’ is defined as: ‘an estimate of the approximate location of the median levels of exposure over the whole spectrum of likely circumstances of use for each scenario’. A value that is actually more often used in risk assessment as estimator of exposure for risk assessment is the ‘reasonable worst case’ value. The TGD (ECB, 2003a) describes the reasonable worst case as follows: ‘the reasonable worst case is regarded as the level of exposure which is exceeded in a small percentage of cases over the whole spectrum of likely circumstances of use for that particular scenario. It excludes extreme use or misuse but can include the upper end of normal use as it is recognised that control of exposure may be poor or non-existent. To decide the reasonable worst case, all measured datasets and qualifying information gathered for a particular scenario should be considered. Since the aim is the identification of highly exposed workers, the aim is to try to approximate to the 90th percentile values’.
|
Whereas in the TGD an estimator of the 90th percentile is used to represent the reasonable worst case, in other risk assessment processes another percentile may be estimated, e.g. the 75th percentile or the 95th percentile (ECB, 2002a).
The estimation of the parameters can theoretically be done by means of formal statistical testing. This requires that the datasets adhere to rather stringent criteria. Most notably, there should be sufficient information to show that the data are from one population. Furthermore, all separate values should be available. And, if several sets are to be combined, there should be a method to correct for the influence of the size of the datasets. In practice, the available datasets do not adhere to such criteria. Often literature data sources are used that present only statistical parameters of the distribution. Furthermore, information is generally lacking that would enable correction for different sizes of datasets (e.g. the percentage of situations sampled from the total population in the target group, or the total size of the population). Therefore, default values representing parameters of the exposure distribution, such as ‘typical values’ and ‘reasonable worst case values’ are generally estimated in a less formal method. The reasonable worst case may be estimated from a graph by drawing a line in the graph somewhere in the high end of the area where the 90th percentiles or maxima are situated on the graph (van Hemmen, 1992). When data are compiled in tables, the derivation of the reasonable worst case estimator is, for example, done by taking the 90th percentile of the largest dataset (if one dataset is clearly larger than the others) or by taking one of the higher values out of 90th percentiles from different datasets. In any case, a large element of ‘expert judgement’ is included in this process.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
In the RISKOFDERM project, default values for ‘typical’ and ‘reasonable worst case’ potential dermal exposure of the hands were derived based on tabulated data. A more or less ‘rounded’ value was chosen for each dataset in a way that resulted in
10% of the measured values (e.g. 3 out of 30) being above this value. This approach was chosen above statistical calculation of the 90th percentile (e.g. from the geometric mean (GM) and the geometric standard deviation (GSD)), because it results in rounded values and more precisely calculated 90th percentiles would suggest an unwarranted precision in the values. Defaults for the body are not presented here, because the practice in risk assessment of new and existing chemicals so far has been to assess only dermal exposure to the hands and because existing sets or default values for industrial chemicals are also only available for hands. All exposure levels were expressed as exposure to the total product used. This was done by dividing the values measured for a specific substance that was part of the product by the fraction of substance in the product.
The conclusions on reasonable worst case and typical potential dermal exposure levels from the non-RISKOFDERM data were used without further handling as they were derived in the creation of the table of default values presented in the TGD (ECB, 2003a) and reproduced in Fehrenbacher et al. (2003). This table of default values is accompanied by notes on the separate studies, including some information on the techniques used, products handled, amounts used and duration of measurements. Not all the values presented in these sources were used in this project. Only datasets with sufficient reported information on relevant determinants, such as amounts used and duration of sampling, and at least 10 data points were taken into account.
All possibly useful datasets were first tabulated, compiling the following type of information per separate dataset.
- A brief description of the major characteristics of the defined TGD exposure scenario.
- A rough indication of the range of durations found within the scenario. If a scenario occurred repeatedly in one work shift, the duration of one occurrence was used.
- A rough indication of the relevant range for amount of product handled (if relevant).
- A very rough indication of the relevant range for area of surface treated (if relevant).
- Other relevant information: e.g. on the presence of local exhaust ventilation, the type of spraying technique or relevant details of handling specific for the dataset.
- Any other general remarks.
- Assumed exposed skin surface area (standard assumption in RISKOFDERM: two hands = 840 cm2).
- Reasonable worst case exposure mass (in mg) over two hands.
- Reasonable worst case exposure loading (mass cm–2 exposed skin surface).
- Typical exposure mass (in mg) over two hands.
- Typical exposure loading (mass cm–2 exposed skin surface).
- Brief description of the reasons for choosing the reasonable worst case value.
The original results for solids were generally expressed in terms of mass. Some original results for liquids were expressed in terms of mass and others in terms of volume. In the latter case, they were converted into terms of mass using information on the density of the liquids. A density of 1 kg l–1 was assumed if information on the density was not available.
This step resulted in default values per separate dataset with a description of the relevant scenario.
The second step was to pull together those datasets that were considered highly similar in the set of descriptive parameters tabulated per set. These were considered to be part of a single clustered ‘TGD scenario’. If the default values per dataset within a cluster were sufficiently similar, as judged by the authors and not by using formal statistical techniques, new default values per clustered TGD scenario were chosen based on the separate default values. If the default values were very different, no default values per clustered TGD scenario were derived. If only one dataset existed for a scenario, the default values from that dataset were considered to be indicative default values for a TGD scenario. However, this was only done if the number of measurements was at least 10, because a dataset with a lower number of measurements is considered to be poor, in agreement with Tielemans et al. (2002).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Summary of measured data
The results of most measured datasets in RISKOFDERM have been published elsewhere. Two publications present an overall view of the results of these measurements (Kromhout et al., 2004; Rajan-Sithamparanadarajah et al., 2004). A total of 375 potential dermal exposure levels were available from these datasets in RISKOFDERM. By far most of them (151) were from handling of objects, which is also the DEO unit with most described RISKOFDERM scenarios. Relatively small numbers of exposure data were available for dispersion with a hand-held tool (n = 30) and for immersing (n = 29). No potential dermal exposure data were available for mechanical treatment of solid objects, largely because workers used mechanics gloves to protect them from cutting and bruising themselves at machines used for sawing, grinding, etc. and at sharp edges and rough surfaces of the objects being treated. Measurement of potential dermal exposure of the hands in these situations was not feasible. Exposure loading rates varied widely with by far the lowest levels for very careful mixing of cancer treatment drugs in hospitals. These levels were generally below 1 ng cm–2 min–1 (Fransman et al., 2004). The highest potential dermal exposure loading rates were found in large scale wiping with a geometric mean for wiping with large quantities of biocide solution of 2300 µg cm–2 min–1 (Hughson and Aitken, 2004). The variation in potential dermal exposure levels is very high if data are pooled per DEO unit or per RISKOFDERM scenario. For RISKOFDERM scenarios with larger datasets (more than 20 measurements) the GSDs ranged from 2.3 in scenario ‘rolling’ (DEO unit ‘dispersion of product with a hand-held tool’) up to 27 for scenario ‘spray painting’ (DEO unit ‘spray dispersion of products’) (Rajan-Sithamparanadarajah et al., 2004). However, the analysis of variability components shows that large differences in average dermal exposure levels are present between DEO units and between RISKODERM scenarios. Classification of situations by DEO unit or by RISKOFDERM scenario explained considerable amounts of variability in dermal exposure. Within-worker variability outweighed the between-worker component. Within-worker exposure levels on average were within a 40-fold range, while individual mean dermal exposure levels were on average within a 4-fold range. (Kromhout et al., 2004).
Results of the measurements in the additional datasets are presented in Table 4 (based on RISKOFDERM DL 40, 2003; RISKOFDERM DL 41, 2003; RISKOFDERM DL 42, 2003; RISKOFDERM DL 43, 2003). These data also showed large variability, both within and between datasets. This was specifically true for dipping (both manual and mechanical), where both very low and very high values were found.
Results from non-RISKOFDERM sources are not presented separately here (Lansink et al., 1996, 1998; HSE, 1999; Preller and Schipper, 1999).
Derived default values
The tabulated information from the separate datasets is presented in Tables 5 and 6. Table 5 presents the default values and the relative place in the full dataset of the reasonable worst case values. Table 6 presents important parameters that may determine the level of dermal exposure for these situations and the references.
|
|
For some of the RISKODERM datasets the number of measurements was considered too limited to use for deriving default values. This regards the following exposure situations: (i) spraying off graffiti remover, (ii) manual dipping into paint remover and (iii) spraying off after dipping into paint remover (all from RISKOFDERM DL 42, 2003).
Two exposure situations, described by two or three sets of measurements that apparently were sufficiently similar to combine for the derivation of default values, resulted in very different default values per dataset. These are presented in Table 7. The large differences in results could not be sufficiently explained by the level of detail in the information used to combine datasets.
|
Based on the data and information in Tables 5 and 6, a number of default values from combined datasets were derived. These values are presented in Tables 8 and 9. Table 8 presents the clustered TGD exposure scenario, the derived default values and the references and Table 9 describes the ranges of values for parameters that may determine the level of dermal exposure within which the combined default values can be considered valid.
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
The use of default values for exposure in risk assessment
Default values for exposure situations structure the analogy approach in risk assessment. This requires that the analogy in both substance and situation is evaluated. The more similar two substances and situations are, the more likely it is that exposure levels for one substance and situation are reasonable predictors of exposure levels for the other substance and situation. Because DEO units are defined to describe the process of exposure and the influence of determinants on these processes, they do not necessarily cluster the level or rate of exposure. Actually, the similarities in exposure levels between different situations within one DEO unit can be very small. This is clear from comparing the careful mixing of cancer treatment drugs with the mixing of large amounts of paint. Therefore, DEO units are not a good basis for default values. The situations clustered in the RISKOFDERM scenarios are already somewhat less variable. The default values presented in this publication are for ‘TGD exposure scenarios’ that are again less variable than the original RISKOFDERM scenarios. In the TGD exposure scenarios more determinants are taken into account than in the RISKOFDERM scenarios, such as the volume of product used, the duration of the scenario and the area treated.
However, the ‘TGD exposure scenarios’ still contain rather broad groups of substances and have similarity in just a few, potentially very important, determinants of exposure. They are therefore not highly accurate and precise estimators of typical and reasonable worst case exposure levels for all substances and situations ‘fitting’ in the analogy. Even within rather narrowly defined situations, large variations in potential dermal exposure levels are to be expected, as shown by some of the large GSDs for the measured datasets.
Three types of default values are presented: default values from combined sets of data, defaults from single sets of data for which there is no comparable set and default values from single datasets with comparable sets that result in very different exposure levels. These values can be used in the following stepwise approach.
Step 1. Use default values from combined sets of data (Table 9) if available for the situation to be assessed.
Step 2. Use default values from a single dataset with no comparable dataset (Table 5) if available for the situation to be assessed if no default values from combined sets of data are available.
Step 3. Use default values from a single dataset with comparable datasets that result in very different exposure levels (Table 7) only for the specific substance and situations or substances and situations that are highly comparable in a large number of aspects, including those that are not included in the parameters described in Tables 5–10.
|
The uncertainty in the assessments using default values cannot be quantified. However, the uncertainty in default values from combined datasets is lower than in default values from single datasets. The combination of data from different substances and/or situations leading to similar exposure levels indicates that at least some of the important determinants of exposure are taken into account in the clustering. The more separate datasets lead to similar results, the lower the uncertainty is. On the other hand, the uncertainty is much higher if default values are used based on single datasets when other single datasets, that appear to be similar in important exposure determinants, have very different exposure levels. The very different exposure levels indicate that some important determinants have not been accounted for in the clustering of these sets. For all default value approaches the uncertainty is lowest when substance and exposure situation assessed are the same as those used for deriving the default values.
The default values of liquid substances were generally based on substances with a rather low volatility, such as 2-(2-butoxyethoxy)ethanol (DEGBE) and N-methyl pyrrolidone (NMP). The highest volatile substance measured was styrene, with a vapour pressure of 0.5 kPa at 20°C. The default values cannot be used for much more volatile substances.
In general it should be stressed that default values cannot be used as separate items. They should always be interpreted in the light of the data and circumstances from which they were derived. The tables provided in this publication should therefore always be used cautiously.
For a real work situation, the use of the default values would be part of an overall approach, including:
- description of the exposure situation and tasks leading to exposure;
- choice of relevant ‘TGD exposure scenarios’ for the full exposure situation;
- establishing the default values to be used per TGD exposure scenario;
- establishing a full shift exposure value based on the relevant default values, taking into account possible influences of cleaning of the skin, possible saturation of the skin, etc.
Exposure to a dusty solid pigment in the production of paint containing 1% of the pigment would, for example, consist of the following consecutive TGD exposure scenarios and reasonable worst case default values: ‘gathering closed bags of powders’ (1050 mg), ‘loading of mixers with powders’ (3000 mg, assuming similarity of the powder with calcium carbonate), ‘disposing emptied bags of powder into a container’ (900 mg) and ‘filling of containers with liquid products’ (11 000 x 0.01 = 110 mg). A high estimate of reasonable worst case exposure on a working day would be the sum of all these values, i.e. 5060 mg. It should be considered that the adding-up of several reasonable worst cases might lead to an unreasonable worst case. An assessor could therefore decide to use a lower value (e.g. 4000 mg) to account for this effect.
Similar datasets with very different exposures and datasets not used
The fact that apparently similar datasets presented in Tables 7 and 8 do not lead to similar exposure levels can partly be explained by other parameters. For the loading of mixers with powders, several differences point in the direction of higher exposure levels for the non-RISKOFDERM data (Lansink et al., 1996) compared to the RISKOFDERM data (RISKOFDERM DL 40, 2003). Calcium carbonate is a much dustier substance than zinc oxide. Calcium carbonate was sampled using cotton sampling gloves, while zinc oxide was sampled by wiping. Gloves have been reported to overestimate exposures compared to sampling by hand washing, while wiping has been reported to underestimate in comparison with hand washing (Fenske et al., 1999). Finally, measurements for calcium carbonate were done in the paint industry and zinc oxide in the rubber industry. The rubber industry is used to handling more toxic components and may therefore have better local exhaust ventilation and other exposure controls. However, the resulting differences in measured potential hand exposure levels are very large and it is difficult to imagine that these aspects fully explain the differences.
The hand exposure levels in the car cleaning study (RISKOFDERM DL 40, 2003) are substantially lower than in the other wiping studies. This was to be expected, because of lower duration of the measurements and the measurement technique (hand washing instead of cotton glove sampling). The application rate is also slightly lower than in the study on wiping of biocides (Hughson and Aitken, 2004). On the other hand, the application rate is very low in the study on graffiti removal with higher exposure levels (RISKOFDERM DL 42, 2003).
The exposure loadings (mass per cm2 exposed), in the studies on wiping surfaces in hospitals and on graffiti removal are very high. Earlier studies on immersion of hands into liquids suggest that the maximum loading of liquids on the skin is in the order of 10–15 mg cm–2 (SAIC, 1996). This suggests that the measurement technique (cotton gloves) in these studies may have led to substantial overestimations of possible exposure loadings. Another possibility is that the more volatile components of the product, including water, evaporate quickly from the skin while the non-volatile substances stay behind. In those cases the results cannot be extrapolated properly from the substance to the product.
Some data sources previously used for default values for dermal exposure of the hands in ECB (2003a) and Fehrenbacher et al. (2003) were not used in this publication. The datasets presented by Guiver et al. (1997) and by Guiver and Foster (1999) contain only a few measurements of potential dermal exposure, because most measurements have been done underneath (protective) gloves. The datasets by Marshall et al. (1992) were also too small (n = 8 for both). The default values based on Roff (1997) were based on extrapolations using the statistical relation between determinants and potential dermal exposure of the hands presented by Roff (1997) and can therefore not be compared directly with the default values presented here. However, they are in the same range as the default values here. This suggests that the default values presented here can be used to estimate exposures for similar situations. Only 7 of 10 data points from Boeniger et al. (1992) for lay-up had enough information to calculate exposure in terms of product.
The information on the TGD exposure scenarios considered in this publication is limited. A slightly extended set of information on the scenarios can be found in the original evaluation of these TGD exposure scenarios made in the scope of the RISKOFDERM project (RISKOFDERM DL 43, 2003).
Comparison of default values with the RISKOFDERM dermal exposure model
The use of default values is only one way to use the results from RISKOFDERM. Another way is to use the models that were based on statistical analyses and calculate exposure levels based on the relations with exposure determinants (Warren et al., 2006). A direct comparison of the two methods is not easy. Where the default values cluster situations in TGD exposure scenarios, the models are based on DEO units and they take account of the influence of determinants within the DEO units. The parameters used in the models are not used as such in the default values. However, a limited indicative comparison can be made by making assumptions on these parameters based on available information. This was done for the default values for clustered TGD exposure scenarios only. The assumptions for the model were based on data from the larger qualitative study in work part 1 (RISKOFDERM DL 17, 2003) if the parameters were not expected to be very much related to specific (sub)scenarios (e.g. ventilation parameters for filling of large quantities). If the parameters were expected to be more related to the specific situations on which the defaults were based, the assumptions were based on information from the specific situations (e.g. use rates in all situations). The results are presented in Table 11.
|
The comparison of default values and model outcomes is highly dependent on the input choice for each parameter in the model. Due to the variation in these parameters within the TGD exposure scenarios this choice is rather arbitrary. However, the results presented in Table 8 do indicate that the two methods largely agree, i.e. the median of the model, using reasonable worst case input values, is generally between the typical value of the TGD exposure scenario and the reasonable worst case value of the TGD exposure scenario, generally closer to the reasonable worst case value. In one case (spreading of a liquid with a comb on a large surface area), the model outcome is higher than the reasonable worst case default value for this TGD exposure scenario. This may be due to a high input value for the application rate. Using half the application rate (0.25 l min–1) leads to a value of 79 mg, instead of 161 mg and this value is well within the range of the default values.
The fact that the two methods give rather similar results is not surprising, as they are largely based on the same sets of data. Only for spraying a substantial number of data, not used for deriving the default values, was used for the model. This provides a slightly more ‘independent’ comparison, although most of the data used in the model were also used in derivation of the default values. Highly variable datasets were used with very different methods of data analysis. This might have lead to large differences in outcome when the results of the analyses are compared, but this is not known to be the case.
An advantage of the default values is the fact that they do not require choices on input values. These may be difficult to make, even if exposure has to be assessed for a specific situation, because none of the input values is necessarily fixed for the situation to be assessed. On the other hand, use of the default values requires that the analogy between the situation to be assessed and the situation described by the default values is evaluated. This may also be rather difficult and the assessor may find that it is difficult to transparently present the argumentation why a situation is or is not sufficiently analogous. In the model, built on the data, exposure rates (mg min–1) are calculated, because the exposure appears to be more or less linearly related to exposure duration (Warren et al., 2006). Such a relation is not necessarily true outside of the range of measurement durations studied. It may also only be valid after taking into account other sources of variation, which is done by the model, but not by the default values. Expressing the default values as rates would facilitate misinterpretation and extrapolation of the data beyond their validity range. This is prevented by expressing data as mass and restricting the use of the default values to exposure situations within a defined set of conditions, including exposure duration.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
RISKOFDERM has resulted in a large number of potential hand and whole body exposure levels, together with useful information on products, techniques, equipment, exposure control and other potential determinants of exposure. Combining the RISKOFDERM datasets with a few non-RISKOFDERM datasets has resulted in five sets of (clustered) default values for hand exposure, cf. the format used in the present TGD, for use in risk assessment with relatively limited uncertainty and several more sets of default values with more uncertainty (because of their smaller basis). The default values are for a specific level of aggregation of situations, here called ‘TGD exposure scenarios’. This level of aggregation is more useful than the levels ‘DEO unit’ and ‘RISKOFDERM scenario’, because it takes account of some highly important determinants. A stepwise approach, with increasing uncertainty in the default values, is presented for use of these default values. The use of these default values with the stepwise approach should be preferred over the use of EASE (ECB, 2003a), as for dermal exposure assessment EASE is not regarded to be very suitable and has no real scientific basis (Cherrie et al., 2003).
Large parts of the realm of industrial or professional dermal exposure situations are not covered by the present sets of default values. Future datasets can be used to add new sets of default values and to improve existing ones. Where no applicable default values are available, the use of the RISKOFDERM dermal exposure model should be preferred. Even for situations with default values this RISKOFDERM dermal exposure model may be as useful as the default values, because it takes account of more determinants, while still being reasonably easy to use.
Received January 24, 2005; in final form January 20, 2006
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONSREFERENCES |
|---|
Auffarth J, van Hemmen J, Hebisch R, et al. (2003) RISKOFDERM—Europas Beschäftigte sollen nicht länger ihre Haut zu Markte tragen. Gefahrstoffe—Reinhaltung Luft 63:399–405.
Beijer MW, Marquart J, Preller EA. (1998) Assessment of inhalation exposure to vapour of liquid substances during filling operations; evaluation of the EASE 2.0 and USEPA Transfer model. (TNO, The Netherlands) (TNO Report V98.1029). Zeist.
Boeniger MF. (2003) On the significance of occupational exposures of the skin to health. Ann Occup Hyg 47:591–3.
Boeniger M, Mason R, Hetcko J, et al. (1992) Use of urine samples to assess and control exposures to 4,4'-methylene dianiline in the aerospace industry. Proceedings of the Conference on Advanced Composites (ACHIG, Inc, Cincinnati, OH, USA) pp. 87–111.
Brouwer DH, Semple S, Marquart J, et al. (2001) A dermal model for spray painters. Part I: Subjective exposure modelling of spray paint deposition. Ann Occup Hyg 45:15–23.
Cherrie JW, Tickner J, Friar J, et al. (2003) Evaluation and further development of the EASE model 2.0. HSE, Research Report 136. (HSE Books, Norwich, UK) ISBN 0 7176 2714 4.
Delgado P, Porcel J, Abril I, et al. (2004) Potential dermal exposure during the painting process in car body repair shops. Ann Occup Hyg 48:229–36.
Dost AA. (1995) A European meeting held to discuss dermal exposure monitoring and related issues, Brussels, Belgium, 21–23 June 1994. Ann Occup Hyg 39:241–55.
ECB. European Chemicals Bureau. (2002a) Technical notes for guidance. Human exposure to biocidal products—guidance on exposure estimation. Report prepared under contract B4-3040/2000/291079/MAR/E2 for the European Commission, DG Environment. Ispra: European Chemicals Bureau, IHPC, JRC, Italy. Available at: http://ecb.jrc.it/biocides/.
ECB. European Chemicals Bureau. (2002b) European Union Risk Assessment Report. ACRYLIC ACID. Risk Assessment Report Vol 28. Ispra: European Chemicals Bureau, IHPC, JRC, Italy. EUR 19836 EN. Available at: http://ecb.jrc.it/existing-chemicals/.
ECB. European Chemicals Bureau. (2003a) Technical Guidance Document on Risk Assessment in support of Commission Directive 93/67/EEC on Risk Assessment for new notified substances Commission Regulation (EC) No 1488/94 on Risk Assessment for existing substances Directive 98/8/EC of the European Parliament and of the Council concerning the placing of biocidal products on the market. Ispra: European Chemicals Bureau, IHPC, JRC, Italy. EUR 20418 EN/1. Available at: http://ecb.jrc.it/existing-chemicals/.
ECB. European Chemicals Bureau. (2003b) European Union Risk Assessment Report METHYL ACETATE. Risk Assessment Report Vol 34. Ispra: European Chemicals Bureau, IHPC, JRC, Italy. EUR 20783 EN. Available at: http://ecb.jrc.it/existing-chemicals/.
ECB. European Chemicals Bureau. (2003c) European Union Risk Assessment Report DIBUTYL PHTHALATE. Risk Assessment Report Vol 29. Ispra: European Chemicals Bureau, IHPC, JRC, Italy. EUR 19840 EN. Available at: http://ecb.jrc.it/existing-chemicals/.
Eriksson K and Wiklund L. (2004) Dermal exposure to styrene in the fibreglass reinforced plastics industry. Ann Occup Hyg 48:203–8.
Eriksson K, Wiklund L, Larsson C. (2004) Dermal exposure to terpenic resin acids in swedish carpentry workshops and sawmills. Ann Occup Hyg 48:267–75.
EUROPOEM. (2002) The development, maintenance and dissemination of generic European databases and predictive exposure models to plant protection products. A Concerted Action under area 4 of FAIR, the Fourth Framework (Agriculture and Fisheries including Agro-Industry, Food Technology, Forestry, Aquaculture and Rural Development) specific Community Research and Technological Development Programme. FAIR3 CT96-1406. Final report, December 2002. BIBRA: Carshalton, Surrey, UK.
Fehrenbacher C, Arnold F, Marquart H, et al. (2003) Approaches for occupational dermal exposure assessment and management, section 2. Hazard Recognition and evaluation. Chapter 17. In DiNardi SR (Ed.). The occupational environment: its evaluation, control, and management 2nd edition (AIHA Press, Fairfax, VA, USA) ISBN. 1 931504 43 1.
Fenske RA. (2000) Dermal exposure: a decade of real progress, invited editorial. Ann Occup Hyg 44:489–91.
Fenske RA, Simcox NJ, Camp JE, et al. (1999) Comparison of three methods for assessment of hand exposure to azinphos-methyl (Guthion) during apple thinning. Appl Occup Environ Hyg 14:618–23.[CrossRef][Medline]
Fransman W, Vermeulen R, Kromhout H. (2004) Occupational dermal exposure to cyclophosphamide in dutch hospitals: a pilot study. Ann Occup Hyg 48:237–44.
Gijsbers JHJ, Tielemans E, Brouwer DH, et al. (2004) Dermal exposure during filling, loading and brushing with products containing 2-(2-butoxyethoxy)ethanol. Ann Occup Hyg 48:219–27.
Guiver R and Foster R. (1999) An additional assessment of exposure to copper during the amateur application of antifouling paint to leisure craft (HSL report JS2000862). (Health and Safety Laboratories, Sheffield, UK).
Guiver R, Chambers H, Douglas N, et al. (1997) A sampling exercise to assess exposure to copper during the amateur application of antifouling paint to leisure craft (HSL report JS2000002). (Health and Safety Laboratories, Sheffield, UK).
Hamey PY. (1995) A Comparison of the Pesticide Handlers Exposure Database (PHED) and the European Predictive Operator Exposure Model (EUROPOEM) Database. Proceedings of a workshop held in OttawaOctober 5–8, 1993Ontario, CanadaMethods of pesticide exposure assessment (Plenum PressIn Curry PB, Iyengar S, Maloney PA, Maroni M (Eds.). , New York, USA) pp. 89–93 ISBN 0 306 45130 1.
Hebisch R and Auffarth J. (2001) Dermal exposure: how to get information. Appl Occup Environ Hyg 16:169–73.[CrossRef][Medline]
HSE. Health and Safety Executive. (1999) Dermal exposure to non-agricultural pesticides. Exposure assessment document (EH74/3). (HSE Books, London, UK) ISBN 0 7176 1718 1.
Hughson GW and Aitken RJ. (2004) Determination of dermal exposures during mixing, spraying and wiping activities. Ann Occup Hyg 48:245–55.
IT Environmental Programs, Inc. (1991) Chemical engineering branch manual for the preparation of engineering assessments. (US EPA, Office of Toxic Substances, Washington, DC, USA).
Kromhout H, Fransman W, Vermeulen R, et al. (2004) Variability of task-based dermal exposure measurements from a variety of workplaces. Ann Occup Hyg 48:187–96.
Lansink CJM, Beelen MSC, Marquart J, et al. (1996) Skin exposure to calcium carbonate in the paint industry. Preliminary modelling of skin exposure levels to powders based on field data (TNO Report V96.064). (TNO, Zeist, The Netherlands).
Lansink CJM, van Hengstum C, Brouwer DH. (1998) Dermal exposure due to airless spray painting—a semi-experimental study during spray painting of a container (TNO Report V97.1057). (TNO, Zeist, The Netherlands).
Mäkinen M and Linnainmaa M. (2004a) Dermal exposure to chromium in the grinding of stainless and acid-proof steel. Ann Occup Hyg 48:197–202.
Mäkinen M and Linnainmaa M. (2004b) Dermal exposure to chromium in electroplating. Ann Occup Hyg 48:277–83.
Marquart J, Brouwer DH, van Rooij JGM. (1994) Occupational skin exposure to chemical substances. Workshop Summary. Appl Occup Environ Hyg 91:77–9.
Marquart J, Maidment S, McClaflin JL, et al. (2001) Harmonization of future needs for dermal exposure assessment and modeling. A workshop report. Appl Occup Environ Hyg 16:218–27.[CrossRef][Medline]
Marquart J, Brouwer DH, Gijsbers J, et al. (2003) Determinants of dermal exposure relevant for exposure modelling in regulatory risk assessment. Ann Occup Hyg 47:599–607.
Marshall MC, Scott JR, Howard HK. (1992) Exposure and release estimations for filter press and tray dryer operations based on pilot plant data (EPA/600/R-92/039). (US Environmental Protection Agency, Cincinnati, OH, USA).
Preller EA and Schipper HJ. (1999) Respiratory and dermal exposure to disinfectants: a study in slaughterhouses and the meat processing industry (TNO report V98.1306). (TNO, Zeist, The Netherlands).
Rajan-Sithamparanadarajah R, Roff M, Delgado P, et al. (2004) Patterns of dermal exposure to hazardous substances in European union workplaces. Ann Occup Hyg 48:285–97.
RISKOFDERM DL 17. (2003) Report of work part 1. (TNO, Zeist, The Netherlands).
RISKOFDERM DL 29. (2002) Main study report of partner 1. (TNO, Zeist, The Netherlands).
RISKOFDERM DL 30. (2002) Main study report of partner 2. (TNO, Zeist, The Netherlands).
RISKOFDERM DL 31. (2002) Main study report of partner 3. (TNO, Zeist, The Netherlands).
RISKOFDERM DL 32. (2002) Main study report of partner 4. (TNO, Zeist, The Netherlands).
RISKOFDERM DL 33. (2003) Main study report of partner 5. (TNO, Zeist, The Netherlands).
RISKOFDERM DL 34. (2002) Main study report National Institute for Working Life (NIWL), Sweden. (TNO, Zeist, The Netherlands).
RISKOFDERM DL 34a. (2002) Main study report of partner 15. (TNO, Zeist, The Netherlands).
RISKOFDERM DL 40. (2003) Benchmark study report of partner 3. (TNO, Zeist, The Netherlands).
RISKOFDERM DL 41. (2003) Main study report of partner 2. (TNO, Zeist, The Netherlands).
RISKOFDERM DL 42. (2003) Benchmark study report of partner 3. (TNO, Zeist, The Netherlands).
RISKOFDERM DL 43. (2003) Final report of Work Part 3. (TNO, Zeist, The Netherlands).
Roff M. (1997) Dermal exposure of amateur or non-occupational users to wood-preservative fluids applied by brushing outdoors. Ann Occup Hyg 41:297–311.
Roff M, Bagon DA, Chambers H, et al. (2004a) Dermal exposure to electroplating fluids and metalworking fluids in the UK. Ann Occup Hyg 48:209–17.
Roff M, Bagon DA, Chambers H, et al. (2004b) Dermal exposure to dry powder spray paints using PXRF and the method of Dirichlet tessellations. Ann Occup Hyg 48:257–65.
SAIC. Science Applications International Corporation. (1996) Occupational dermal exposure assessment, a review of methodologies and field data. (US EPA, Office of Toxic Substances, Washington, DC, USA).
Schneider T, Vermeulen R, Brouwer DH, et al. (1999) Conceptual model for assessment of dermal exposure. Occup Environ Med 56:765–73.[Abstract]
Semple S, Brouwer DH, Dick F, et al. (2001) A dermal model for spray painters. Part II: Estimating the deposition and uptake of solvents. Ann Occup Hyg 45:25–33.
Tielemans E, Marquart H, de Cock J, et al. (2002) A proposal for evaluation of exposure data. Ann Occup Hyg 46:287–97.
van Hemmen JJ. (1992) Agricultural pesticide exposure data bases for risk assessment. Rev Environ Contam Toxicol 126:1–85.[ISI][Medline]
van Hemmen JJ. (1993) Predictive exposure modelling for pesticide registration purposes. Ann Occup Hyg 37:541–64.
van Hemmen JJ. (2004) Dermal exposure to chemicals, invited editorial. Ann Occup Hyg 48:183–5.
van Hemmen JJ, Auffarth J, Evans PG, et al. (2003) RISKOFDERM: risk assessment of occupational dermal exposure to chemicals. An introduction to a series of papers on the development of a toolkit. Ann Occup Hyg 47:595–8.
Warren ND, Marquart J, Christopher Y, et al. (2006) The development of task-based dermal exposure models for regulatory risk assessment. Ann Occup Hyg 50:491–503.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  P. Boogaard Biomonitoring as a tool in the human health risk characterization of dermal exposure Human and Experimental Toxicology, April 1, 2008; 27(4): 297 - 305. [Abstract] [PDF]
|
|
|
|
 |
|
 T. OGDEN Annals of Occupational Hygiene at Volume 50: Many Achievements, a Few Mistakes, and an Interesting Future Ann. Hyg., November 1, 2006; 50(8): 751 - 764. [Abstract] [Full Text] [PDF]
|
|
|