bmd: an R package for benchmark dose estimation

Abstract

The benchmark dose (BMD) methodology is used to derive a hazard characterization measure for risk assessment in toxicology or ecotoxicology. The present paper’s objective is to introduce the R extension package bmd, which facilitates the estimation of BMD and the benchmark dose lower limit for a wide range of dose-response models via the popular package drc. It allows using the most current statistical methods for BMD estimation, including model averaging. The package bmd can be used for BMD estimation for binomial, continuous, and count data in a simple set up or from complex hierarchical designs and is introduced using four examples. While there are other stand-alone software solutions available to estimate BMDs, the package bmd facilitates easy estimation within the established and flexible statistical environment R. It allows the rapid implementation of available, novel, and future statistical methods and the integration of other statistical analyses.

Fathead Minnows Exposed to Organic Compounds from Oil Sands Tailings as Embryos Have Reduced Survival, Impaired Development, and Altered Behaviors That Persist into Larval Stages

“…Exposure–response relationships for embryonic mortality, hatch success, heart rate, and twitching rate, as well as larval mortality, were modeled using log‐logistic (logit) regression with the drc package (Ritz et al, 2015), and the bmd package (Jensen et al, 2020) was used to estimate the benchmark concentration (to the 10th percentile with 95% confidence interval [CI]) for NAFCs (Jenson et al, 2020; Ritz & Streibig, 2005). The four‐parameter log‐logistic function (LL.4) was used in exposure–response models for binary endpoints (embryonic and larval mortality, embryonic hatch viability), and the three‐parameter log‐logistic function (LL.3) was used in exposure–response models for all other endpoints (embryonic heart rate and twitching rate).…”

A Nonmechanistic Parametric Modeling Approach for Benchmark Dose Estimation of Event‐Time Data

“…Model checking may also be carried out separately for each step. Subsequently, BMD estimation may be performed in the usual way (Jensen et al., 2020).…”

“…The procedure is implemented using the open source statistical programming language R (http://www.r-project.org) (R Core Team, 2019) and in particular the extension package bmd (Jensen, Kluxen, Streibig, Cedergreen, & Ritz, 2020). The entire R code used in the examples below is provided as part of the supplementary material.…”

Biostatistics and its impact on hazard characterization using in vitro developmental neurotoxicity assays

“…Inverse regression estimates both BLL and BUL directly from the regression fit around the BMC (Buckley, Piegorsch & West, 2009;Fang, Piegorsch & Barnes, 2015) and therefore puts high emphasizes on a successful regression fit in terms of robustness and reliability. The delta method is an asymptotic approach which combines information of the estimated model parameters to derive a Wald-type interval (Jensen et al 2020) To simplify the model averaging method, only three regression models were considered (four-parameter loglogistic, four-parameter Weibull and three-parameter exponential model).…”

“…The BMC estimation consists of various statistical decisions to be made in the concentration-response analysis, which dependent largely on the experimental design, the concentration-response data and assay endpoint features, and which require statistical knowledge that is usually only warranted by experienced biostatisticians. In current data practice, the experimenter is often responsible for the data analysis and therefore relies on statistical software without being aware about the software default settings and how they can impact the outputs of data analysis (Jensen et al ., 2020). Existing guidelines for concentration response data analysis are often too general (OECD, 2006; EFSA, 2016), and no clear consensus on a common and standardized biostatistical method for in vitro toxicity data have been achieved (Wheeler et al, 2015; Sand et al, 2017).…”