According to a May 2015 progress report to Congress, the Environmental Protection Agency’s (EPA) Integrated Risk Information System (IRIS) is better off than it was a year ago, but not as effective as it needs to be to deliver timely, high-quality and credible chemical risk assessments.
EPA does deserve credit, first and foremost, for following through on several of the National Research Council’s (NRC) recommendations to fix IRIS, which has struggled to meet various stakeholder expectations. At the same time, more tangible signs of progress are needed before everyone can be assured that the Agency is doing the very best job it can to build greater confidence that its assessments can be used to inform federal and state chemical management decisions.
Keep in mind, IRIS assessments are highly influential for guiding public health policy, so it’s important that “close enough” not be the standard by which the Agency evaluates chemicals under IRIS.
In order for IRIS to produce better results, EPA needs to make more headway in three key areas:
1. IRIS needs a more robust problem formulation and scoping process
In its progress report, EPA states that changes to problem formulation and scoping have been implemented. However, ACC has some concerns about transparency in the decision-making processing regarding which assessments are initiated, and why.
EPA’s own Human Health Risk Assessment Framework describes important elements that should be included in the process, but which are not referenced in the IRIS scoping and problem formulation documents for ethylbenzene, naphthalene and PCB’s. Examples of these important elements include:
- What risk assessment products (qualitative and quantitative) are needed by management?
- Are screening values sufficient?
- Are best estimates and distributions needed?
- Finally, what will the schedule be for developing the assessment?
2. EPA must establish a clear & objective criteria for reviewing evidence
While it is discouraging that recent IRIS assessments have not included clear and objective criteria, ACC continues to be optimistic that the Chemical Assessment Advisory Committee (CAAC) review panels will urge EPA to create transparent benchmarks to judge data quality and relevance, and clearly note the rating of each critical study according to these benchmarks.
At the recent CAAC review for benzo[a]pyrene, reviewers struggled to understand why EPA chose to rely on some studies that experts found to be low quality. Simple intermediate steps (e.g., Klimisch scoring for non-cancer end points) for documenting the quality of studies would go a long way to improve the logic and decision making in IRIS assessments.
3. EPA must be more responsive to peer review and stakeholder input
While EPA has implemented more stakeholder meetings — which indeed have led to more stakeholder feedback — commenters can cite very few examples, if any, in which EPA actually used the feedback to inform their decisions.
For example, there have been two meetings to discuss draft assessments before peer review (step four meetings for ethylene oxide and benzo[a]pyrene) where EPA took public comment; however EPA made little if any changes in response to these discussions. Where there has been engagement before draft assessments are released, EPA has not provided any feedback to stakeholders regarding what information is helpful and what is not.
EPA should facilitate a two-way dialogue at the bimonthly IRIS meetings so stakeholders leave the meetings with a clear sense of how EPA believes scientific differences should be resolved. Currently, EPA collects information but is reticent to share what they think about the information with the public.
A solution for sound chemical assessments
We understand these problems can take time to resolve. However, some solutions are available that are not difficult to implement and which EPA can begin to move forward on today.
Completing work in these three areas will help EPA ensure that objective scientific analysis and transparency are at the core of how the Agency evaluates the safety of chemicals. Fixing the flaws in the process will help improve the productivity of IRIS and eliminate unnecessary delays in completing the assessments – and in turn, shorten the time needed to finalize the assessment. Furthermore, addressing these issues will lead to higher quality assessments and increase their overall utility for making risk management decisions, a goal that everyone can agree is important for restoring the public’s faith that chemicals are being managed effectively.
This is why we support an inclusive process that leverages everyone’s knowledge and expertise for improving how assessments are conducted, and why we are committed to helping EPA fill in the scientific gaps so the cracks plaguing IRIS can finally fade away.