7. Risk Characterization

7.1.1.1 Option – Alternatives to Default Assumptions

Appendix D presents a table summarizing some of the common default assumptions in a risk assessment and alternate assumptions that may reduce uncertainties but still protect health at a particular site. Site-specific information should be well documented so the project team clearly understands how this information was derived and how it applies to the site.

7.1.1.2 Option – Evaluate Central Tendency Exposure

The RME scenario based on the default assumptions does not fully characterize the range of possible exposures and conditions at a site. Evaluation of a central tendency exposure scenarioA set of facts, data, assumptions, and professional judgment about how an exposure occurs or does not occur. An exposure scenario addresses the (1) chemicals in environmental media and their sources; (2) exposed populations (or receptors); (3) migration of chemicals in environmental media from sources to receptors; and (4) routes of exposure (ingestion, dermal contact, inhalation). may provide risk managers with an understanding of the degree of protectiveness of the RME scenario, which can aid in the decision-making process.

7.1.2.1 Option – Considerations for Summation of Risk Results

Where relevant, decisions regarding whether remedial action is warranted at a site should be based on site-related cumulative cancer risk and the noncancer hazard index for RME under current and reasonably expected future land use (USEPA 1991e). Estimating these RME risks is the goal of the baseline risk assessment. While in some instances combining risks across exposure pathways and media may be appropriate to represent the RME, in other instances this practice is not appropriate. Most, if not all, of the default exposure scenarios and toxicity valuesDerived values (for example, reference doses and slope factors) that can be used to estimate the incidence or potential for adverse human health effects in receptor (USEPA 2015h). represent the RME for an individual pathway and media. These values are estimated assuming only one exposure mediumEnvironmental medium containing concentrations of a chemical that may be contacted by a receptor. and are based on default residential or worker exposures for RME.

At sites with more than one exposure medium or potential for exposureContact of a receptor with a chemical. Exposure is quantified as the amount of the chemical available at the exchange boundaries of the organism (for example, skin, lungs, gut) and available for absorption (USEPA 1989a). from multiple exposure pathways from the same exposure medium (for example, soil contact and inhalation of vapors from the soil), the risks may be summed if those risks may represent a one RME due to consistency in both time and location of exposure. Clearly presenting the RME risks by each exposure pathwayThe course a chemical takes from a source to a receptor. An exposure pathway describes a unique mechanism by which an individual or population is exposed to chemicals at or originating from a site. Each exposure pathway includes a source or release from a source, an exposure point, and an exposure route. If the exposure point differs from the source, a transport/exposure medium (for example, air) or media (in cases of intermedia transfer) also is included (USEPA 1989a). and medium most accurately describes potential exposure risks at the site, while also allowing flexibility to add RME risk estimates to satisfy other requirements. As USEPA RAGS Part F (USEPA 2009a) states:

Risk assessors should first identify reasonable exposure pathway combinations. Then, risk assessors should examine whether it is likely that the same individuals would consistently face the reasonable maximum exposureThe highest exposure that is reasonably expected to occur at a site (USEPA 1989a). by more than one pathway.

A related issue is presentation of risk results when multiple chemicals may have a significant contribution to site risk. An assumption of additivity of the chemical-specific toxic effects is commonly invoked when calculating risk for chemical mixtures. This assumption and other options are described in Section 5.1.3.

7.1.3.1 Option – Considerations for Probabilistic Risk Assessments

Both deterministic and probabilistic methods can support risk management decisions at sites. The methods are similar in concept and approach, but differ in the values supplied for variables in the risk equations. Deterministic methods use single point values for the equation variables, while probabilistic methods use probability distributions for these variables. The probabilistic approach quantitatively characterizes variabilityA population’s natural heterogeneity or diversity, particularly that which contributes to differences in exposure levels or in susceptibility to the effects of chemical exposures (Commission 1997a). For example, workers may perform different functions that may affect time, frequency, and duration of contact with an environmental medium). Variability cannot be reduced by collection of additional data. and uncertaintyThe lack of perfect knowledge of values or parameters used in a risk assessment. Uncertainty may be reduced by collection of additional data. in the risk estimates (USEPA 2001c).

A probabilistic risk assessment can characterize the probability distributionA distribution describes the probability or likelihood of any potential value. of the risk assessment results and provide information on the sensitivity of the input variables to help the risk assessor refine the risk assessment (collect additional site-specific information for an especially sensitive parameter) or to help in making risk management decisions. Chapter 7 of USEPA’s guidance (USEPA 2001c) describes the interpretation and use of probabilistic risk estimates in remedy decision making. Generally, a specific upper-bound level can be selected from the range of percentiles of risk estimates to inform the ultimate decision of whether a potential exposure risk meets the applicable risk level. A probabilistic risk assessment, however, is sometimes limited by the quality and applicability of data available to develop probability distributions for site input parameters and requires more time, resources, and expertise than a deterministic risk assessment.

7.2.1.1 Option – Organized and Systematic Presentation of Risk Results

The information and results presented in the risk characterization (the final step of the risk assessment) should address the issues and questions that were developed in planning the risk assessment (see Chapter 3). As USEPA (1989a) recommends, an important use of a risk assessment’s results is to highlight potential sources of unacceptable risk at a site in order to allow them to be addressed effectively (for example, through remediation or development of institutional or engineering controlsEngineered and constructed physical barriers to contain, prevent, or mitigate exposure to chemicals in an environmental medium. Examples of engineering controls include engineered caps and subslab depressurization systems, mitigation barriers, and fences. Similar to activity and land use restrictions, engineering controls also typically require a specific mechanism for noticing the presence of engineering control and related restrictions, as well as long-term maintenance and management of the control. The timing of a decision to use an engineering control, and the specific mechanism to be used, may be based on criteria outlined in statute, regulation, policy, or guidance.).

The risk characterization should communicate in text and tables the key results, assumptions, and uncertainties of the risk assessment in a transparent, clear, and consistent manner (USEPA 2000b; USEPA 1995c). The risk results should identify the major factors contributing to the risks:

• What exposure scenarios would result in unacceptable risk?
• What are the chemicals contributing to the unacceptable risk?
• What exposure pathways are resulting in the unacceptable risk (for example, groundwater ingestion, dermal contact with soil, inhalation of particulates, or inhalation of vapors)?
• Is the unacceptable risk due to current exposures or based upon reasonably anticipated future exposures that have not yet occurred?
• Does the uncertainty analysis identify the nature and magnitude of key uncertainties and bias in the risk results?

If the risk characterization clearly addresses these questions, then the risk assessment will serve as an effective tool for supporting risk management decisions. A clear, well-organized risk assessment is also effective in supporting the risk communicationRisk communication is the formal and informal process of communication among and between regulatory agencies and organizations responsible for site assessment and management, and the various parties who are potentially at risk from or are otherwise interested in the site. needs of the project. While the format for presenting these results can vary based on the needs of the project, the presentation of the risk characterization results should include concise tables and clear presentations of necessary information. An example table is presented in Appendix E.

At the time that risk management decisions are made, the risk characterization should be reviewed and it should be confirmed that no significant changes in current or future exposure scenarios and toxicity values of site-related chemicals have occurred. If significant changes in exposure scenarios or toxicity values have occurred, applicable components of the risk assessment should be reviewed and updated if needed so that risk management decisions are based on the latest information (see Section 8.2.1).

7.3.1.1 Option – Provide Information so that the Uncertainties and Bias Can be Understood.

The information and results presented in the risk characterization include the uncertainties and bias inherent in the assumptions, models, and input parameter values used in the risk assessment (as discussed in Chapter 4, Chapter 5, and Chapter 6). Generally, when bias is recognized in selected input parameter values, the intent is to be protective of human health and to overestimate likely exposures and risks. Underestimation of potential site risks is usually related to deficiencies in sampling and analysis or improper modeling of exposure concentrations. A clear explanation of the uncertainties and bias inherent in the risk assessment is key to the transparency of the risk characterization.

Information about the uncertainty and bias in the risk results should be presented in tables and text as appropriate for the project. This clear presentation allows project managers to understand, interpret, weigh, and decide on a course of action for a specific site.

If it is determined that the overall bias may result in underestimating the risks, then multiple alternatives are available to overcome these uncertainties (for example, revisiting the planning and scoping of the data collection; see Section 3.3).

7.3.1.2 Option – Detailed Consideration of Toxicological Assumptions

In general, it is not necessary for the risk assessment to include details regarding specific toxicity values unless adjustments are made (for example, relative bioavailabilityThe fraction of an ingested dose that crosses the gastrointestinal epithelium and becomes available for distribution to internal target tissues and organs (USEPA 2007c).) or site-specific considerations warrant such discussion. It is important, however, to provide insight into the degree of uncertainty and potential bias inherent in the derivation of a specific toxicity values for chemicals driving the risk assessment. Some of the potential issues to address include: 

• uncertainty related to extrapolation from animal studies to human toxicity
• applying dose-response information from homogeneous animal populations or healthy human populations to predict effects that may occur in the general population, including sensitive subpopulations
• using high-to-low-dose linear extrapolation methods to develop toxicity values, which can result in overestimates of risk for chemicals with a threshold dose or nonlinear low-dose responses
• using surrogate chemicals or route-to-route extrapolation to account for the lack of toxicity values
• possible synergistic or antagonistic effects associated with multiple chemical exposure, which could underestimate or overestimate the final risks
• using toxicity values derived for a different exposure period (such as subchronic) to exposure periods evaluated in the risk assessment (such as chronic)

7.3.1.3 Option – Detailed Consideration of Exposure Assumptions

The exposure assumptions used in the risk assessment generally include uncertainties and bias that must be recognized when making risk management decisions. Some of the potential issues to address may include:

• sampling methods, which are often selected to provide the worst case scenario (for example, biased sampling, grab groundwater sampling, or sampling only from the interval of likely highest chemical concentration)
• sampling methods that are designed to evaluate the site-wide average concentration (for example, composite sampling or incremental sampling methods) and may not be appropriate to identify potential hot spotsHot spots are considered to be soil volumes with relatively high concentrations that could be present at a site but whose locations and dimensions cannot be anticipated prior to sampling (ITRC 2012a). (especially if the receptorAn individual (for example, residential adult, residential child, worker, trespasser, or recreator) who has the potential to be exposed to a chemical in environmental media.’s long-term exposure could be consistent with the areal extent of any hot spots)
• sensitive analytical methods, which typically provide detection limits below risk-based screening values, even if such limits are below a site-specific background concentration developed to screen out chemicals
• EPCs, which can be calculated for large areas where exposure could regularly occur over the assumed exposure period; or EPCs, which can be calculated for very small areas where regular exposure would not likely occur over the exposure period
• exposure scenarios, exposure pathways, and exposure parameters evaluated in the assessment, which may not completely account for land use assumptions and restrictions, or parameters that overestimate the RME (such as applying fish ingestion rates to a small water body when these rates came  from a combination of rates in multiple water bodies)
• common fate and transport models used in the risk assessment, which assume steady-state or assume simultaneous complete phase change and  infinite source for different simultaneous routes of exposure

7.3.1.4 Option – Include Multiple Descriptors of Risk

The risk assessment calculations can incorporate multiple risk descriptors in addition to RME, such as the central tendency exposure, population-level risk, and risk for important subgroups (for example, children, and subsistence anglers). Evaluating multiple risk descriptors provides the project managerAn individual from a regulatory agency (for example, federal, state, or local), or a consulting company, or responsible party company, who is coordinating the site cleanup including the risk assessment. with information on a range of possible exposures (and thus risk) to consider when using the risk assessment results to manage a site.