An official website of the United States government.

This is not the current EPA website. To navigate to the current EPA website, please go to www.epa.gov. This website is historical material reflecting the EPA website as it existed on January 19, 2021. This website is no longer updated and links to external websites and some internal pages may not work. More information »

Using Predictive Methods to Assess Hazard under TSCA

On this page:

What information can these models provide?

A hazard assessment attempts to answer the following questions for a chemical:

  • Will it present health effects of concern to the human population?
  • Will it present health effects of concern to the ecological population?
  • At what level will these adverse effects occur?

How and when to use hazard models?

The first step in completing a hazard assessment is to ensure there is adequate information – empirical and/or predictive -- on each of the toxicological endpoints of interest.  The process for collecting data and information within EPA is shaped by the regulatory framework of individual environmental laws and whether adequate empirical data is available for all endpoints.

Depending on the answers to those questions, then a weight of evidence approach may be used employing predictive approaches to inform the hazard assessment. When experimental data are lacking, predictive models can provide useful insights on potential adverse effects.

Each predictive tool is designed with the goal of estimating activities of untested chemicals based on their structural similarity to chemicals with known activities. Predictive methods have a long history of use both for the industrial design and regulatory assessment of chemicals.

Use considerations: confidence level, weight-of-evidence approach

Predictive approaches are considered one of many potential sources of information to support a weight-of-evidence approach in chemical hazard assessment.

Similar to other sources of data considered, such as laboratory studies, the defensibility of predictions can be related to the consistency of the predictions generated from the various models as well as the consistency between the predictions and the results of other lines of evidence.

The degree of confidence required depends on the purpose of the prediction, and the standard of proof and degree of scientific certainty required will all vary depending on the purpose of the assessment. For example, a different degree of confidence may be required for:

  • Screening and priority-setting of chemicals for further evaluation
  • Hazard identification for risk assessment
  • Classification and labeling
  • Meeting information requirements in different regulatory schemes.

Predictive approaches to assess hazard can be used for all of these as primarily or supporting evidence. EPA also uses hazard predictive models to:

  • Better inform chemical management decisions
  • Devise testing strategies
  • Prioritize activities
  • Conduct alternatives analysis
  • Inform green chemistry
  • Ranking of chemicals for degree of hazard
  • Support chemical risk assessments

Models and tools developed by EPA to assess hazard under TSCA

Ecological Structure-Activity Relationships Program (ECOSAR): Uses quantitative structure-activity relationships (QSAR) to predict the toxicity of untested chemicals based on their structural similarity to chemicals for which aquatic studies are available.  Predictions are based on a library of class-based QSARs within the program overlaid with an expert decision tree for selecting the appropriate chemical class. The program estimates a chemical's toxicity potential based on short and long-term exposure for fish, invertebrates, and aquatic plants along with a limited number of methods for terrestrial and marine organisms.

OncoLogic: Uses mechanism-based structure-activity relationships (SAR) analysis to evaluate cancer potential of untested chemicals based on their structural similarity to chemicals for which studies have been conducted. The rules are incorporated into decision trees that are used along with user input for analysis.

Non-Cancer Health Assessment: Provides user with a procedure of identifying health effects data for the chemical of interest from various public databases and gives insight into various chemical classes associated with health effects of concern.

Analog Identification Methodology (AIM) : Designed to facilitate a data search on a chemical of interest in addition to identification of potential structural analogs and associated data which can be used to support read across approaches and data gap filling.

Chemical Assessment Clustering Engine (ChemACE): Instantly “clusters” chemicals in a large user defined chemical list based on structure. The tool is useful for identifying structural diversity in a chemical inventory and instantly highlighting analogous chemicals for potential read across.

Model Averaging for Dichotomous Response Benchmark Dose (MADr-BMD) Tool: The Model Averaging for Dichotomous Response Benchmark Dose (MADr-BMD) tool is a downloadable program that implements model averaging based upon dichotomous dose-response data.

EPA's New Chemical Categories Document: Designed to highlight EPA's accumulated experience in assessing particular substances with shared chemical and toxicological properties. Chemicals whose physical-chemical, toxicological and ecotoxicological properties are likely to be similar or follow a regular pattern as a result of structural similarity may be considered as a group, or ‘category’ of chemicals.