Approaches for describing and communicating overall uncertainty in toxicity characterizations: U.S. Environmental Protection Agency's Integrated Risk Information System (IRIS) as a case study

Abstract

Single point estimates of human health hazard/toxicity values such as a reference dose (RfD) are generally used in chemical hazard and risk assessment programs for assessing potential risks associated with site- or use-specific exposures. The resulting point estimates are often used by risk managers for regulatory decision-making, including standard setting, determination of emission controls, and mitigation of exposures to chemical substances. Risk managers, as well as stakeholders (interested and affected parties), often have limited information regarding assumptions and uncertainty factors in numerical estimates of both hazards and risks. Further, the use of different approaches for addressing uncertainty, which vary in transparency, can lead to a lack of confidence in the scientific underpinning of regulatory decision-making. The overarching goal of this paper, which was developed from an invited participant workshop, is to offer five approaches for presenting toxicity values in a transparent manner in order to improve the understanding, consideration, and informed use of uncertainty by risk assessors, risk managers, and stakeholders. The five approaches for improving the presentation and communication of uncertainty are described using U.S. Environmental Protection Agency's (EPA's) Integrated Risk Information System (IRIS) as a case study. These approaches will ensure transparency in the documentation, development, and use of toxicity values at EPA, the Agency for Toxic Substances and Disease Registry (ATSDR), and other similar assessment programs in the public and private sector. Further empirical testing will help to inform the approaches that will work best for specific audiences and situations.

Risk and Uncertainty Communication

“…There appears to be no attempt at communicating what these terms mean in practice, although there have been recent suggestions for graphics showing how uncertainty leads to the move from PoD to RfD (Beck et al 2016), and of using the GRADE quality of evidence scale in environmental and occupational health (Morgan et al 2016). Using examples including acrylamide and BPA, Lofstedt (2013) argued that regulatory agencies need to invest heavily in their communications, with which it would be difficult to disagree.…”

Guidance on Communication of Uncertainty in Scientific Assessments

“…Only four studies (Patt and Dessai, 2005;Wardekker et al, 2008;Pappenberger et al, 2013;Beck et al, 2016) experimentally dealt with the clearly identifiable target audiences with whom EFSA interacts on a regular basis, namely decision-makers, policy advisers and stakeholder representative groups such as NGOs, consumer groups or industry associations.…”

Recommendations to address uncertainties in environmental risk assessment using toxicokinetics-toxicodynamics models

“…With the 10 data sets we used and the 20 fittings we performed, the four measures of goodness-of-fit showed similar outputs for the GUTS-RED-SD and GUTS-RED-IT models under both constant and time-variable exposure profiles. The percentage of observed data falling within the 95% predicted credible interval, %P P C, has the advantage of being linked to visual graphics, i.e., PPC plots, and is therefore easier for risk assessors and stakeholders to interpret than the Bayesian WAIC and LOO-CV measures [12].…”

“…Quantifying uncertainties or levels of confidence associated with toxicity endpoints is undoubtedly a way to improve trust in risk predictors and to avoid decisions that could increase rather than decrease the risk [12,14,24]. The Bayesian framework has many advantages for dealing with uncertainties since the distribution of parameters and thus their uncertainties is embedded in the inference process [36].…”