Group 3A

The following questions / concerns were discussed:

1. Would the methods reviewed in this workshop be accepted for use in regulatory assessments under current guidance? If not, what steps would be needed to facilitate their acceptance in the future? What are the opportunities to update technical guidance?

The use of SSDs is already widespread in regulation for data rich substances (e.g. for water quality standard derivation in various regulatory regimes), less common is the use of interspecies correlation approaches such as Web Ice and SSD when limited data are available. Some statistical instruments, for compiling SSDs from toxicity data and deriving threshold values, are already routinely in use for these purposes.

A possible limitation to the current application of SSD approaches within regulatory processes is the limited guidance available in some areas, although the availability of toxicity test data which fulfils the taxonomic diversity requirements established under several regulatory regimes is also an important potential limitation. Whilst the minimum data requirements for the use of an SSD differ between different regulatory regimes several European regimes require a minimum of 10 species, representing at least 8 different taxa and 5 different phyla. There is a need for balance in regulatory guidance between prescriptive approaches, which give more consistent outcomes, and flexible approaches based on best scientific practice. It is especially important that where professional or expert judgement is used the justification for the approach taken must be scientifically defensible and clearly documented.

The focus is on fulfilling the required number of species required for an SSD for it to be acceptable. Some flexibility around the number of species/taxa might be better with more emphasis on the uncertainty associated with the derived HC5. Additional guidance is required on the most appropriate ways to derive the confidence interval around the HC5. Furthermore, some attention should also be given to the uncertainty in HC5 that are hard or impossible to quantify (e.g. uncertainties associated with model assumptions). Currently the focus is on those sources of uncertainty that can be quantified, and more work is required to better address those areas of uncertainty which cannot be readily addressed statistically at present.

Extrapolation and estimation techniques are used within some regulatory areas, and some limited general guidance exists for establishing their validity and applicability from the OECD. Again there is a need to understand the uncertainties introduced through the use of extrapolated data.

2. Should current guidance on the use of SSDs be revised in the light of the issues and approaches discussed in this workshop, e.g. number of species?

Some aspects of current guidance could be reviewed, particularly in light of the practical experience gained through the more widespread application of SSDs in regulation. Whilst the taxonomic diversity criteria defined by the London Workshop (EC, 2001*) could be updated it is important to recognise that different applications of SSD have different requirements. It is likely that where there is an existing requirement for taxonomic diversity this would still be maintained. Knowledge concerning the mechanism of toxic action should be used when evaluating the appropriate number of species/taxa required for an SSD.

* EC. 2001. Report of the expert consultation workshop on statistical extrapolation techniques for environmental effects assessment. London Workshop 18-19 January 2001. Despite extensive efforts to track this report, we have only been able to locate a draft copy thanks to some participants. This draft is available by contacting the ECETOC Secretariat.

Guidance in the form of a decision tree would assist in identifying the most appropriate approach for any particular situation. There should be a principle of including as much information as possible, in an intelligent manner so that less reliable or relevant information makes a smaller overall contribution to the overall weight of evidence.

The assessment can be an iterative process but this will not always be the case.

3. What implications are there for the interpretation of SSDs and HC5s in risk assessment and risk management?

Flexibility requires an intelligent approach, which should be fully documented with the supporting scientific justification for any decisions taken.

A distinction needs to be made between protective and predictive applications of SSDs, as this can have implications for the data requirements of different models and approaches. The use of an SSD to derive a PNEC, or similar threshold, for a substance is typically a protective application, and aims to derive an HC5 for the overall community. The use of an SSD to assess impacts at a contaminated site undergoing risk management, where the Potentially Affected Fraction of the overall community might be derived, would typically be a predictive application of an SSD.

SSDs should ideally be applied within approaches which employ multiple lines of evidence, and are updated to include new information as it becomes available.

* EC. 2001. Report of the expert consultation workshop on statistical extrapolation techniques for environmental effects assessment. London Workshop 18-19 January 2001. Despite extensive efforts to track this report, we have only been able to locate a draft copy thanks to some participants. This draft is available by contacting the ECETOC Secretariat.

4. What are the research needs?

The applicability of toxicity extrapolation methods should be further validated for acute effects, and should also be evaluated for chronic effects.

There is a need to better understand the uncertainties within the assessment which are currently unquantifiable.

Further validation of SSDs derived from laboratory data against field and mesocosm studies is required, as is guidance on the different approaches (including their limitation) which can be taken.