|
|
|
|
|
|
|
|
| NIOSH Home > Safety and Health Topics >Skin Exposures and Effects >Occupational & Environmental Exposures of Skin to Chemicals- 2005> Abstracts |
|
|
 |
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Table 1. Summary of data used in RISKOFDERM modelling |
DEO Unit |
Number of data for modelling |
|
Hands |
Body |
|
Handling contaminated objects |
195 |
No data |
Manual dispersion of products |
79 |
57 |
Dispersion with hand-held tools |
117 |
107 |
Spray application |
140 |
335 |
Immersion (mechanical) |
13 |
68 |
Treatment of solid objects |
No data |
97 |
Total |
544 |
664 |
For each of the DEO unit a separate equation is made in which more or less determinants of dermal exposure are included. The equations calculate the median exposure rate (mg/min) over two hands or the remainder of the body. By considering the variation in exposure rate that is not explained by the variation in determinants, a distribution of exposure rates with the given input parameters can be calculated. An excel spreadsheet was created in which the exposure assessor can easily calculate exposure rates and resulting exposure loadings (in mg) based on these rates and the duration of exposure. Figure 1 shows an example of the spreadsheet.
In a workshop with stakeholders it was recommended to better clarify the validity boundaries of the model and to perform a series of practical tests for real life risk assessments.
Figure 1 Example of the RISKOFDERM potential dermal exposure model spreadsheet
As a first step, the model is run for a number of scenarios previously assessed and agreed upon on in the Existing Substances Regulation. This indicates the difficulties faced in using the model with data generally available in assessment of existing substances and the advantages or disadvantages of the model, compared to presently used methods. The method generally used now is the model EASE, in addition to a small number of default values based on measurements.
Preliminary results of the analyses are available for 10 substances and 8 different exposure scenarios. As several scenarios occur for several substances, there were 19 combinations of substance and scenario studied so far.
The RISKOFDERM model requires input on a number of determinants not generally presented in the EU Risk Assessment Reports (RARs) so far. Some of these parameters are also not easily evaluated from an expert’s desk. Therefore, several estimations are done with two or more variations in input values. In this way, the total number of RISKOFDERM model estimates so far is increased to 29. For each set of input values the percentiles of the outcome are calculated in steps of 10 percent (i.e.: 10th, 20th, 30th etc.).
Because in the present assessments potential dermal exposure is only estimated for hands, this is also the focus of the comparison with RISKOFDERM model estimates. However, for each of the scenarios, also an estimate of the median potential whole body exposure was made with the RISKOFDERM model to assess the possible relevance of body exposure.
The results show that descriptions of processes and exposure situations in the RARs are generally insufficient to enable proper estimation using the RISKOFDERM model. In Table 2, the necessary inputs for the RISKOFDERM model are clustered in groups based on availability in the studied RARs.
Table 2. Availability of input on parameters for RISKOFDERM model in
present EU Risk Assessment Reports
Generally available |
Seldom or not available |
|||
DEO Unit |
Intensity of contact |
|||
Scenario |
Use rate |
|||
Physical state |
(Cumulative) duration |
|||
Volatility |
Direction of airflow |
|||
Concentration in product |
Direction of application |
|||
Reasonably available or easily deducible |
||||
Dustiness |
Length of tool handle |
Viscosity |
||
Quality of ventilation |
Indoors / outdoors |
|||
Frequency of contact |
Segregation |
|||
Aerosol formation |
Distance to source |
|||
Automation / manual |
Adequate LEV |
|||
The most critical missing parameters are use rate and cumulative duration (of the actual exposure process) as these are used in most of the equations in the model. The effect of use rate can be high (depending on the scenario). The effect of duration in the model is linear.
An important scenario in practice, bagging of powders, is not included in the scenarios in the equations for handling of contaminated objects. No measurements were done for this scenario due to practical reasons. To enable assessment of this scenario, the option “weighing and adding of solids” has to be chosen. This substantially increases the uncertainty of the estimate and may also lead to biased results. Some other relevant scenarios (e.g. sampling) are also not covered specifically in this DEO unit.
For the manufacture of MDA as a solid, it is not clear from the RAR whether the input “aerosol formation” should be chosen as “yes” or “no”. The estimates of the RISKOFDERM with the option “yes” are approximately 3 times those with the option “no”.. The quality of the ventilation was not clear from the RAR for the manufacture of formulations containing methylmethacrylate. In this case the option “poor ventilation” leads to an exposure estimate approximately 3 times as high as the option “normal or good ventilation”. Such differences may be important for the conclusion of the risk assessment.
Direct comparison of the RISKOFDERM model results with the RAR estimates (based on EASE) is very difficult, due to the lack of information on relevant inputs for the RISKOFDERM model. However, a comparison of some estimates with the inputs considered most probable is presented in Figure 2.
The results show that the 10th percentile of the most probable RISKOFDERM model estimate is generally below the estimate in the RAR, while the 90th percentile is often (very far) above the estimate in the RAR. However, as duration and use rate are presently unknown, the presented RISKOFDERM estimates may be biased and no conclusion can be drawn on the over- or underestimation by the given EASE estimates.
Potential dermal exposure to the body is generally estimated to be substantially lower than that to the hands. In 6 of 16 cases potential body exposure is more than potential hand exposure. All of these were related to mixing of liquids on a relatively large scale, spray application of paint and application of paint or adhesives with a hand-held tool. All other cases showed potential hand exposures that were more than twice the potential body exposures.
The RISKOFDERM model offers the opportunity to provide much more specific estimates than the EASE model. It takes account of the real influence of important determinants, such as use rate and duration of the task. EASE is not based on real measured data. It is logical to use the model that has a better scientific basis. However, present information available to exposure assessors does not allow estimation of highly important parameters of the RISKOFDERM model.
A compilation should be made of information on the missing parameters in a large sample of relevant exposure situations. Statistical analysis can derive typical and reasonable worst case values to allow standard dermal exposure scenarios to be based on the model for a first tier assessment. For a second tier assessment, specific information on the specific values for the situation under assessment should be gathered.
Further data gathering is necessary to improve the RISKOFDERM model, specifically for handling of contaminated objects. This highly important DEO unit consists of many possible scenarios that have not yet sufficiently been studied. For immersion and treatment of solid objects the model is based on very limited measured data for hand exposure or there is no model for this at all. These scenarios also need more attention.
Content last modified: 22 May 2005
