Risk assessment retrospective

The third category of methods addressed in this chapter are error analysis and reduction methodologies. Error analysis techniques can either be applied in a proactive or retrospective mode. In the proactive mode they are used to predict possible errors when tasks are being analyzed during chemical process quantitative risk assessment and design evaluations. When applied retrospectively, they are used to identify the underlying causes of errors giving rise to accidents. Very often the distinction between task analysis and error analysis is blurred, since the process of error analysis always has to proceed from a comprehensive description of a task, usually derived from a task analysis. 

Studies of working women present the potential for additional bias, because some factors that influence employment status may also affect reproductive end-points. For example, because of child care responsibilities, women may terminate employment, as might women with a history of reproductive problems who wish to have children and are concerned about workplace exposures (Joffe, 1983 Lemasters Pinney, 1989). Thus, retrospective studies of female exposure that do not include terminated female workers may be of limited use in risk assessment, because the level of risk for some of the outcomes is likely to be overestimated (Lemasters Pinney, 1989). 

The different models show bias. Concentration addition implies that effects always increase with each additional compound (even when present at extremely low concentrations), and there is clear evidence that this occurs with similarly acting compounds, but it is not yet clarified with dissimilarly acting compounds (see reviews, above). Thus, applying concentration addition to mixtures of dissimilarly acting compounds might overestimate effects and risks, and this may be an undesired feature for various risk assessments (e.g., for retrospective, diagnostic risk assessments). Response addition implies that mixture effects only occur when at least 1 compound induces toxic effects. However, as cited in the review of data (Section 5.3.1), mixtures with... 

Experimental data or field observations on mixture toxicity and the responses in species assemblages are rarely available, with some exceptions (Korthals et al. 2000 Backhaus et al. 2004 Arrhenius et al. 2004). Nonetheless, risk assessment and legislation often focus on the protection of community-level endpoints in both prospective and retrospective risk assessments. 

Species sensitivity distributions (SSDs) are used for both prospective and retrospective risk assessments (Posthuma et al. 2002b). In prospective risk assessments, the concept is used to derive hazardous concentrations (e.g., HC5), which are used to derive environmental quality criteria. In retrospective risk assessments, the SSD approach is used to determine the local toxic pressure in terms of the potentially affected fraction (PAF) of species for each compound separately. Subsequently the multisubstance (ms)PAF, or optionally the combi-PAF, for the local mixture can be calculated. Originally, the combi-PAF concept was developed by Hamers et al. (1996) and assumes that only compounds exerting narcotic effects... 

Prospective and Retrospective Risk Assessment at the Landscape Level... 

The monitoring studies mentioned above can be considered as valuable retrospective tools to verify the field relevance of risk assessment procedures currently in use, including the extrapolation tools used to predict chemical stress at the ecosystem and landscape levels. 

There are many extrapolation methods, of different complexities, and with different purposes and suitabilities for prospective and retrospective risk assessments. A compilation of the methods is insufficient to guide the choice of procedures to use when assessors need to conduct risk assessments. Therefore, a practical and pragmatic guide to extrapolations and their everyday use is provided in the last chapter. It defines a general stepwise approach to identifying the types of extrapolation (matrix and media, (Q)SARs, mixtures, etc.) that are most relevant for an assessment problem, and it defines an overall approach to the assignment of tiers. 

Extrapolation methods are used for various types of risk assessment. Methods may be used in the process of deriving environmental quality objectives, in the registration of new substances, and in the process of site-specific risk assessment. Suter (1993) called these approaches prospective (the former 2) and retrospective (the latter) risk assessments. The specific process in which extrapolation methods are used has implications for the concepts to be applied and the data to be used as input in extrapolation. Strictly described approaches are in place for the derivation of environmental quality criteria (EQCs) and the registration of pesticides and newly developed substances. The prescribed approaches for deriving EQCs can differ between jurisdictions. The approaches for retrospective investigations have more degrees of freedom. A characteristic of the latter approach is that the methods can make use of measured local exposure levels and can estimate local risk with known precision (or known uncertainty ). The latter is uncommon for EQCs. 

When the scope is known and the extrapolation methods are defined, the data for the risk assessment steps (and extrapolations) can be collected. For tiered systems, this implies that choices need to be made on the manner in which uncertainties are addressed and what to do when these are only addressed by simple methods. A distinction was already made between prospective use and retrospective use of extrapolation methods. In both cases, extrapolations are being applied, but the way in which existing methods are selected for an assessment problem can differ. 

In retrospective risk assessment, the output should be as accurate as possible at any tier. Tiers differ by the level of sophistication and realism. 

When there is no guidance to choose the best method, it is always possible to use the window of prediction. Especially for retrospective risk assessments that should result in a decision, 2 or more extrapolation approaches can be used for the same problem (for example, the 2 mixture models), and both results can be compared to the decision criterion. In case both predictions are (by far) lower or higher than the decision criterion, the decision would remain the same, irrespective of the model Only in cases where the window of prediction overlaps with the decision criterion would further work be necessary. 

Ankley GT, Mount DR. 1996. Retrospective analysis of the ecological risk of contaminant mixtures in aquatic sediments. Human Ecol Risk Assess 2 434-440. 

Ecological risk assessment of chemical mixtures thus has to deal with a variety of field phenomena, a possible range of assessment endpoints, and a variety of assessment approaches. Moreover, there exists a huge variety in the regulatory questions and problem formulations addressed in ecological risk assessment of chemical mixtures. Examples include the protection of specific species against well-defined mixtures (like PCBs and PAHs), the protection of an undefined concept like the ecosystem, and retrospective assessments for highly or diffusely contaminated systems. 

The review presented in the previous sections shows an enormous diversity in risk assessment methods and procedures for chemical mixtures. This diversity is characteristic for the current state of the art. The awareness that mixtures may cause risks that are not fully covered by single compound evaluations is growing, but the knowledge required to accurately assess the risks of chemical mixtures is still limited. The scientific community is attempting to unravel the mechanisms involved in mixture exposure and toxicity, and over recent decades, a multitude of new techniques to assess mixture risks have been developed. However, a comprehensive and solid conceptual framework to evaluate the risks of chemical mixtures is still lacking. The framework outlined in Section 5.4 can be considered a first step toward such a conceptual framework. The framework recognizes that the problem definitions vary greatly (between protective and retrospective assessments, for humans and ecosystems), and that each question has resulted in a different type of approach. 

Ecological risk assessments can either be predictive or retrospective. Predictive risk assessments are used... 

HPDP explicitly incorporates, in the case of the Cherry Point herring, these levels of scale and grain size that are critical to consider in a risk or retrospective assessment (Figure 2.6). A framework that applies the components of HPDP immediately places an endpoint or assessment endpoint into an ecologically relevant contextual framework including spatial scale, grain size, and temporal relationships. Wu and David (2002) have demonstrated that the HPDP framework can also incorporate anthropogenic features such as land use boundaries, roads, and urbanization. 

Forbes, V. and P. Calow. 2002. Applying weight-of-evidence in retrospective ecological risk assessment when quantitative data are limited. Human. Ecol. Risk. Assess. 8 1625-1639. 

Suter, G.W. (1998) Retrospective assessment, ecoepidemiology and ecological monitoring. Handbook of Environmental Risk Assessment and Management, Calow, P. (ed.), pp. 170-217. Blackwell Science, Oxford. 

---