EPA-Expo-Box (A Toolbox for Exposure Assessors)
Screening-Level and Refined
Depending on the purpose and scope of the assessment, an exposure assessor might choose to conduct a very simple assessment based on readily available data and conservative assumptions – a screening-level assessment. However, if a very basic assessment does not yield results that adequately support decision-making, a more detailed, or refined assessment (i.e., higher tiered assessment) might need to be conducted that will require data specific to the problem at hand and more resources to conduct the assessment. Using this iterative approach, exposure assessors can start with a basic approach and layer in complexity as appropriate.
Below are characteristics typical of screening-level and refined assessments, but they are not necessarily components of all screening-level or refined assessments. For example screening assessments sometimes employ more complex models and site-specific data, but more often these types of assessments use simple models and readily available data. Similarly, not all refined assessments use a probabilistic approach, but this approach is only sometimes used in screening assessments.
|Screening-level assessments typically use readily available data and conservative assumptions to estimate a high-end exposure (e.g., reasonable maximum exposure, reasonable worst-case exposure, maximum exposure) of the exposure to a sensitive receptor. These assessments use point estimates for input data and are usually designed around a single stressor.
In contrast, a refined assessment often uses more representative exposure assumptions and inputs. Instead of point estimates, a refined assessment might use a distribution of data to generate multiple points for input into the exposure equation. Refined assessments are often based on central tendency data instead of high end exposure. They can focus on single or multiple stressors.
|Screening-level assessments use simple models or more sophisticated models parameterized with default values. These assessments typically use a deterministic approach. A refined assessment often uses site-specific assumptions, modeling or higher-precision sampling and analysis techniques, and probabilistic methods.|
|Screening-level assessments result in a conservative, quantitative estimate of exposure – typically a point value that can be compared to an established level of concern. Screening-level assessments are relatively inexpensive and quick to carry out, but results contain a high degree of uncertainty and do not typically characterize variability. This means the results are appropriate for prioritization, comparing multiple sites of interest, or ruling out potential exposure pathways.
Uncertainty can be characterized with a sensitivity analysis to determine the relative contribution of exposure pathways and variables to the exposure estimate. A sensitivity analysis following a screening-level assessment could be used to identify the most important variables contributing to exposure. These variables can then be used to determine how best to refine the assessment (e.g., which variables should use distributions of data instead of point estimates in a probabilistic assessment).
In a refined assessment, the use of more specific measurement data and sophisticated models may result in a more realistic picture of exposure potential. Refined assessments that use a probabilistic modeling approach may characterize both uncertainty and variability better.
The following resources provide information for conducting tiered exposure assessments.
Tools for Assessing Uncertainty and Variability
The following tools are resources for addressing uncertainty and variability. Additional information can be found at the quick link on Uncertainty and Variability.