The impact of model uncertainty on benchmark dose estimation

Abstract

We study the popular benchmark dose (BMD) approach for estimation of low exposure levels in toxicological risk assessment, focusing on dose-response experiments with quantal data. In such settings, representations of the risk are traditionally based on a specified, parametric, dose-response model. It is a well-known concern, however, that uncertainty can exist in specification and selection of the model. If the chosen parametric form is in fact misspecified, this can lead to inaccurate, and possibly unsafe, lowdose inferences. We study the effects of model selection and possible misspecification on the BMD, on its corresponding lower confidence limit (BMDL), and on the associated extra risks achieved at these values, via large-scale Monte Carlo simulation. It is seen that an uncomfortably high percentage of instances can occur where the true extra risk at the BMDL under a misspecified or incorrectly selected model can surpass the target BMR, exposing potential dangers of traditional strategies for model selection when calculating BMDs and BMDLs.

Information‐theoretic model‐averaged benchmark dose analysis in environmental risk assessment

“…Alternatively, had the analyst chosen the model based on the minimum AIC, which here is the probit (M2), the 95% BMDL M2 would be 1.9266 − (1.645)(0.2414) = 1.5295 mg/kg/day. While slightly larger than our FMA BMDL, this limit (i) is unadjusted for its data-based model selection and (ii) if based on an incorrect model may likely suffer from non-trivial under-coverage/overestimated extra risk, as we identified in West et al (2012). Corroborating reports by many who have come before, we find that FMA adjustment frees risk assessors from the selection biases, model inadequacies, and inferential uncertainties one encounters when committing to only a single parametric model to perform the benchmark analysis.…”

“…Whether the chosen form is actually correct is uncertain, however, and there is growing recognition that this level of model uncertainty is more extensive and pernicious than has previously been imagined with BMD estimation. For instance, we recently studied each of the eight models in Table 1, and found that model misspecification can drastically affect the coverage performance of the Wald-type BMDL in (2.1): for most misspecification pairings—e.g., selecting a logistic model, M1, when the data are actually generated from the similar probit model, M2—severe undercoverage from the nominal 95% coverage level often occurred, and could spiral down to 0% as sample size increased (West et al, 2012). [As expected, when the correct parametric model was employed to build the BMDL, large-sample coverage patterns were stable; cf.…”

Correlation of Noncancer Benchmark Doses in Short‐ and Long‐Term Rodent Bioassays

“…So, data sets having BMD to BMDL ratios above 10 are not typical of a biological response, and any such models were removed. Filter 5. The remaining best‐fit models were then averaged together, although generally these instances resulted in BMD and BMDLs that were virtually or completely identical Filter 6.…”

Bayesian Model-Averaged Benchmark Dose Analysis via Reparameterized Quantal-Response Models

“…To adjust, we explore a Bayesian model averaging (BMA) framework for the estimation process (Hoeting et al, ). Previous parametric model averaging schemes for BMDs usually focused on at most three or four different forms, although some writers have expanded to as many as eight different dose‐response functions (Wheeler and Bailer, ; West et al, ; Piegorsch et al, ). These correspond to popular choices in the U.S. EPA's BMDS software (Davis et al, ), and are cataloged in Table .…”