A scoping review of evaluations of and recommendations for default uncertainty factors in human health risk assessment

Abstract

Uncertainty factors (UFs) are used to account for uncertainties and variability when setting exposure limits or guidance values. Starting from a proposal of a single UF of 100 to extrapolate from an animal NOAEL to a human acceptable exposure, the aspects of uncertainty and number of UFs have diversified and today there are several risk assessment guidelines that contain schemes of default UFs of varying complexity. In the present work, we scoped the scientific literature on default UFs to map developments regarding recommendations and evaluations of these. We identified 91 publications making recommendations for one or several UFs and 55 publications evaluating UFs without making explicit recommendations about numerical values; these were published between 1954 and 2021. The 2000s was the decade with the largest number of publications, interspecies differences and intraspecies variability being the most frequent topics. The academic sector has been the most active (76 out of 146 publications). Authors from the private sector more often presented UF recommendations, but differences between sectors regarding size of recommendations were not statistically significant. The empirical underpinning of the reviewed recommendations ranges from four to 462 chemicals, that is, relatively low numbers compared with the range of chemicals these default UFs are expected to cover. The recommended UFs have remained remarkably constant, with merely a slight decrease over time. Although chemical specific UFs are preferable, the widespread use of default UFs warrants further attention regarding their empirical and normative basis.

Prioritization of control measures in leakage scenario using Hendershot theory and FBWM-TOPSIS

“…Researchers in process safety analysis mainly face three types of uncertainty: completeness, modeling, and parameter [15]. In these studies, they have tried to use different methods to reduce these uncertainties [16][17][18]. Hajiaghaei et al used the Pythagorean Fuzzy Technique for Order Preference by Similarity to the Ideal Solution (PF-TOPSIS) method to select the best Green Supplier Selection (GSS) [19].…”

Human–mouse comparison of the multistage nature of radiation carcinogenesis in a mathematical model

“…1 There is great uncertainty, however, in quantitatively predicting the dose-effect relationship in humans from the results of mouse experiments. 2 For example, current risk assessment of genotoxic chemicals uses a linear dose-response model, unless evidence of a threshold-like dose response is available, and estimates an exposure level that gives a certain lifetime probability of cancer (e.g., 10 À5 ) based on evidence from epidemiology or long-term animal experiments. 3 Therein, the lifetime probability of cancer in animals is assumed to have quantitatively equivalent significance in humans.…”

“…Many agents have been identified to be carcinogenic in mouse models and classified as probable and possible human carcinogens 1 . There is great uncertainty, however, in quantitatively predicting the dose–effect relationship in humans from the results of mouse experiments 2 . For example, current risk assessment of genotoxic chemicals uses a linear dose–response model, unless evidence of a threshold‐like dose response is available, and estimates an exposure level that gives a certain lifetime probability of cancer (e.g., 10 −5 ) based on evidence from epidemiology or long‐term animal experiments 3 .…”