John F. Fox
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Research Interests
John F. Fox has a total of 75 co-authors. Here are the top co-authors by number of shared publications:
- Leonid Kopylev
Center for Public Health and Environmental Assessment, Office of Research and Development, US Environmental Protection Agency, Washington DC, USA
3 shared publications Chao Chen
NCEA, ORD, U.S. Environmental Protection Agency, Pennsylvania Ave. NW, Washington, DC, USA
2 shared publicationsCheryl Siegel Scott
U.S. Environmental Protection Agency, Office of Research and Development, National Center for Environmental Assessment, Washington, DC 20460, United States
2 shared publications- Debra L. Denton
U.S. Environmental Protection Agency, Region IX, Sacramento, California 95814, USA
2 shared publications Jennifer Jinot
U.S. Environmental Protection Agency, Office of Research and Development, National Center for Environmental Assessment, Washington, DC 20460, United States
2 shared publicationsJohn P. Bryant
University of Alaska
2 shared publications
Latest Publications
- 25 August 2025
- Published by Oxford University Press (OUP) in Toxicological Sciences
- Vol. 208 (1) , 167-175
Estimation of benchmark dose ratio distributions for subchronic-to-chronic extrapolation using meta-analysis
Recently, the International Programme on Chemical Safety (IPCS) developed a unified probabilistic framework for deriving reference values and a software tool, Approximate Probabilistic Analysis (APROBA), to help implement this framework. The distributions of multiple sources of uncertainty and variability were estimated, including uncertainty when extrapolating from subchronic to chronic data. The subchronic-to-chronic distribution was estimated using ratios between subchronic and chronic benchmark doses (BMDs) and was determined to be approximately lognormal, with parameter values reported by IPCS. These parameters were estimated largely from historical data on body and organ weights from toxicological studies. We estimated the distribution using a larger collection of data, including histopathological and clinical endpoints. Our analysis determined that key assumptions of the method and the default values in APROBA are consistent with the results from the new data. However, the uncertainty of predictions for dichotomous response data was greater than assumed in APROBA, and the reference values derived using our new results were lower than those derived from APROBA (by 25% in an example case). Also, APROBA's default parameter values do not account fully for the uncertainty of predicted chronic BMDs. Most importantly, the uncertainty of the prediction can be much greater than assumed in APROBA if BMDs are accepted when they fall well outside the observed dose range or when an upper confidence limit is not quantifiable. Careful evaluation of dose–response model fit, including a number of indicators of model suitability in addition to standard goodness-of-fit statistics, is necessary to improve quantification of uncertainty.
- 25 August 2025
- Published by Oxford University Press (OUP) in Toxicological Sciences
- Vol. 208 (1) , 167-175
Estimation of benchmark dose ratio distributions for subchronic-to-chronic extrapolation using meta-analysis
Recently, the International Programme on Chemical Safety (IPCS) developed a unified probabilistic framework for deriving reference values and a software tool, Approximate Probabilistic Analysis (APROBA), to help implement this framework. The distributions of multiple sources of uncertainty and variability were estimated, including uncertainty when extrapolating from subchronic to chronic data. The subchronic-to-chronic distribution was estimated using ratios between subchronic and chronic benchmark doses (BMDs) and was determined to be approximately lognormal, with parameter values reported by IPCS. These parameters were estimated largely from historical data on body and organ weights from toxicological studies. We estimated the distribution using a larger collection of data, including histopathological and clinical endpoints. Our analysis determined that key assumptions of the method and the default values in APROBA are consistent with the results from the new data. However, the uncertainty of predictions for dichotomous response data was greater than assumed in APROBA, and the reference values derived using our new results were lower than those derived from APROBA (by 25% in an example case). Also, APROBA's default parameter values do not account fully for the uncertainty of predicted chronic BMDs. Most importantly, the uncertainty of the prediction can be much greater than assumed in APROBA if BMDs are accepted when they fall well outside the observed dose range or when an upper confidence limit is not quantifiable. Careful evaluation of dose–response model fit, including a number of indicators of model suitability in addition to standard goodness-of-fit statistics, is necessary to improve quantification of uncertainty.
- 18 February 2019
- Published by Oxford University Press (OUP) in Environmental Toxicology and Chemistry
- Vol. 38 (3) , 511-523
Comparison of false-positive rates of 2 hypothesis-test approaches in relation to laboratory toxicity test performance
We compared 2 statistical hypothesis‐test approaches (no‐observed‐effect concentration [NOEC] and test of significant toxicity [TST]) to determine the influence of laboratory test performance on the false‐positive error rate using the US Environmental Protection Agency's Ceriodaphnia dubia reproduction whole‐effluent toxicity (WET) test endpoint. Simulation and power calculations were used to determine error rates based on observed control coefficients of variation (CV) for 8 laboratories over a range of effect levels. Average C. dubia control reproduction among laboratories was 20 to 40 offspring per female, and the 75th percentile CV was 0.10 to 0.31, reflecting a range in laboratory performance. The 2 approaches behave similarly for CVs of 0.2 to 0.3. At effects CV decreases, TST is less likely to declare toxicity and NOEC is more likely to do so. Laboratory performance affects whether a sample is declared toxic and influences the probability of false‐positive (and −negative) error rates using either approach. At the 75th percentile control CV observed for each laboratory, 4 laboratories would achieve approximately a 5% false‐positive rate using 13 or fewer replicates for this test method. For the remaining 4 laboratories, more replicates would be needed to achieve a 5% false‐positive rate. The present analyses demonstrate how false‐positive rates are influenced by laboratory performance and WET test design. Environ Toxicol Chem 2019;9999:1–13. Published 2019 Wiley Periodicals Inc. on behalf of SETAC. This article is a US government work and, as such, is in the public domain in the United States of America.
- 18 February 2019
- Published by Oxford University Press (OUP) in Environmental Toxicology and Chemistry
- Vol. 38 (3) , 511-523
Comparison of false-positive rates of 2 hypothesis-test approaches in relation to laboratory toxicity test performance
We compared 2 statistical hypothesis‐test approaches (no‐observed‐effect concentration [NOEC] and test of significant toxicity [TST]) to determine the influence of laboratory test performance on the false‐positive error rate using the US Environmental Protection Agency's Ceriodaphnia dubia reproduction whole‐effluent toxicity (WET) test endpoint. Simulation and power calculations were used to determine error rates based on observed control coefficients of variation (CV) for 8 laboratories over a range of effect levels. Average C. dubia control reproduction among laboratories was 20 to 40 offspring per female, and the 75th percentile CV was 0.10 to 0.31, reflecting a range in laboratory performance. The 2 approaches behave similarly for CVs of 0.2 to 0.3. At effects CV decreases, TST is less likely to declare toxicity and NOEC is more likely to do so. Laboratory performance affects whether a sample is declared toxic and influences the probability of false‐positive (and −negative) error rates using either approach. At the 75th percentile control CV observed for each laboratory, 4 laboratories would achieve approximately a 5% false‐positive rate using 13 or fewer replicates for this test method. For the remaining 4 laboratories, more replicates would be needed to achieve a 5% false‐positive rate. The present analyses demonstrate how false‐positive rates are influenced by laboratory performance and WET test design. Environ Toxicol Chem 2019;9999:1–13. Published 2019 Wiley Periodicals Inc. on behalf of SETAC. This article is a US government work and, as such, is in the public domain in the United States of America.
- 23 October 2017
- Published by Wiley in Risk Analysis
- Vol. 38 (5) , 1052-1069
Correlation of Noncancer Benchmark Doses in Short‐ and Long‐Term Rodent Bioassays
This study investigated whether, in the absence of chronic noncancer toxicity data, short‐term noncancer toxicity data can be used to predict chronic toxicity effect levels by focusing on the dose–response relationship instead of a critical effect. Data from National Toxicology Program (NTP) technical reports have been extracted and modeled using the Environmental Protection Agency's Benchmark Dose Software. Best‐fit, minimum benchmark dose (BMD), and benchmark dose lower limits (BMDLs) have been modeled for all NTP pathologist identified significant nonneoplastic lesions, final mean body weight, and mean organ weight of 41 chemicals tested by NTP between 2000 and 2012. Models were then developed at the chemical level using orthogonal regression techniques to predict chronic (two years) noncancer health effect levels using the results of the short‐term (three months) toxicity data. The findings indicate that short‐term animal studies may reasonably provide a quantitative estimate of a chronic BMD or BMDL. This can allow for faster development of human health toxicity values for risk assessment for chemicals that lack chronic toxicity data.
- 23 October 2017
- Published by Wiley in Risk Analysis
- Vol. 38 (5) , 1052-1069
Correlation of Noncancer Benchmark Doses in Short‐ and Long‐Term Rodent Bioassays
This study investigated whether, in the absence of chronic noncancer toxicity data, short‐term noncancer toxicity data can be used to predict chronic toxicity effect levels by focusing on the dose–response relationship instead of a critical effect. Data from National Toxicology Program (NTP) technical reports have been extracted and modeled using the Environmental Protection Agency's Benchmark Dose Software. Best‐fit, minimum benchmark dose (BMD), and benchmark dose lower limits (BMDLs) have been modeled for all NTP pathologist identified significant nonneoplastic lesions, final mean body weight, and mean organ weight of 41 chemicals tested by NTP between 2000 and 2012. Models were then developed at the chemical level using orthogonal regression techniques to predict chronic (two years) noncancer health effect levels using the results of the short‐term (three months) toxicity data. The findings indicate that short‐term animal studies may reasonably provide a quantitative estimate of a chronic BMD or BMDL. This can allow for faster development of human health toxicity values for risk assessment for chemicals that lack chronic toxicity data.
- Open Access
- 27 August 2016
- Published by Wiley in Risk Analysis
- Vol. 37 (5) , 905-917
Dose‐Response Modeling with Summary Data from Developmental Toxicity Studies
Dose-response analysis of binary developmental data (e.g., implant loss, fetal abnormalities) is best done using individual fetus data (identified to litter) or litter-specific statistics such as number of offspring per litter and proportion abnormal. However, such data are not often available to risk assessors. Scientific articles usually present only dose-group summaries for the number or average proportion abnormal and the total number of fetuses. Without litter-specific data, it is not possible to estimate variances correctly (often characterized as a problem of overdispersion, intralitter correlation, or “litter effect”). However, it is possible to use group summary data when the design effect has been estimated for each dose group. Previous studies have demonstrated useful dose-response and trend test analyses based on design effect estimates using litter-specific data from the same study. This simplifies the analysis but does not help when litter-specific data are unavailable. In the present study, we show that summary data on fetal malformations can be adjusted satisfactorily using estimates of the design effect based on historical data. When adjusted data are then analyzed with models designed for binomial responses, the resulting benchmark doses are similar to those obtained from analyzing litter-level data with nested dichotomous models.
- 27 August 2016
- Published by Wiley in Risk Analysis
- Vol. 37 (5) , 905-917
Open AccessDose‐Response Modeling with Summary Data from Developmental Toxicity Studies
Dose-response analysis of binary developmental data (e.g., implant loss, fetal abnormalities) is best done using individual fetus data (identified to litter) or litter-specific statistics such as number of offspring per litter and proportion abnormal. However, such data are not often available to risk assessors. Scientific articles usually present only dose-group summaries for the number or average proportion abnormal and the total number of fetuses. Without litter-specific data, it is not possible to estimate variances correctly (often characterized as a problem of overdispersion, intralitter correlation, or “litter effect”). However, it is possible to use group summary data when the design effect has been estimated for each dose group. Previous studies have demonstrated useful dose-response and trend test analyses based on design effect estimates using litter-specific data from the same study. This simplifies the analysis but does not help when litter-specific data are unavailable. In the present study, we show that summary data on fetal malformations can be adjusted satisfactorily using estimates of the design effect based on historical data. When adjusted data are then analyzed with models designed for binomial responses, the resulting benchmark doses are similar to those obtained from analyzing litter-level data with nested dichotomous models.
- Open Access
- 27 August 2016
- Published by Elsevier in Reproductive Toxicology
- Vol. 65, 321-358
A systematic evaluation of the potential effects of trichloroethylene exposure on cardiac development
- 27 August 2016
- Published by Elsevier in Reproductive Toxicology
- Vol. 65, 321-358
Open AccessA systematic evaluation of the potential effects of trichloroethylene exposure on cardiac development
- 1 June 2015
- Published by Elsevier in Regulatory Toxicology and Pharmacology
- Vol. 72 (1) , 94-101
Evaluation of the ToxRTool’s ability to rate the reliability of toxicological data for human health hazard assessments
- 1 June 2015
- Published by Elsevier in Regulatory Toxicology and Pharmacology
- Vol. 72 (1) , 94-101
Evaluation of the ToxRTool’s ability to rate the reliability of toxicological data for human health hazard assessments
- Open Access
- 1 March 2013
- Published by American Chemical Society (ACS) in Environmental Health Perspectives
- Vol. 121 (3) , 303-311
Human Health Effects of Trichloroethylene: Key Findings and Scientific Issues
A peer-reviewed open access journal published by the National Institute of Environmental Health Sciences
- 1 March 2013
- Published by American Chemical Society (ACS) in Environmental Health Perspectives
- Vol. 121 (3) , 303-311
Open AccessHuman Health Effects of Trichloroethylene: Key Findings and Scientific Issues
A peer-reviewed open access journal published by the National Institute of Environmental Health Sciences
- 22 January 2013
- Published by Wiley in Risk Analysis
- Vol. 33 (9) , 1608-1619
Nonparametric Bayesian Methods for Benchmark Dose Estimation
The article proposes and investigates the performance of two Bayesian nonparametric estimation procedures in the context of benchmark dose estimation in toxicological animal experiments. The methodology is illustrated using several existing animal dose-response data sets and is compared with traditional parametric methods available in standard benchmark dose estimation software (BMDS), as well as with a published model-averaging approach and a frequentist nonparametric approach. These comparisons together with simulation studies suggest that the nonparametric methods provide a lot of flexibility in terms of model fit and can be a very useful tool in benchmark dose estimation studies, especially when standard parametric models fail to fit to the data adequately.
- 22 January 2013
- Published by Wiley in Risk Analysis
- Vol. 33 (9) , 1608-1619
Nonparametric Bayesian Methods for Benchmark Dose Estimation
The article proposes and investigates the performance of two Bayesian nonparametric estimation procedures in the context of benchmark dose estimation in toxicological animal experiments. The methodology is illustrated using several existing animal dose-response data sets and is compared with traditional parametric methods available in standard benchmark dose estimation software (BMDS), as well as with a published model-averaging approach and a frequentist nonparametric approach. These comparisons together with simulation studies suggest that the nonparametric methods provide a lot of flexibility in terms of model fit and can be a very useful tool in benchmark dose estimation studies, especially when standard parametric models fail to fit to the data adequately.
- 31 August 2012
- Published by Wiley
Aquatic Toxicity Tests
- Open Access
- 1 July 2012
- Published by Punjab University Press in Pakistan Journal of Statistics and Operation Research
- Vol. 8 (3) , 441-478
Some New Aspects of Statistical Inference for Multistage Dose-Response Models with Applications
Applications of the results to dose-response multistage models serve as illustrations. The paper considers various scenarios, and is expected to be of interest to practitioners of risk analysis.
- 1 July 2012
- Published by Punjab University Press in Pakistan Journal of Statistics and Operation Research
- Vol. 8 (3) , 441-478
Open AccessSome New Aspects of Statistical Inference for Multistage Dose-Response Models with Applications
Applications of the results to dose-response multistage models serve as illustrations. The paper considers various scenarios, and is expected to be of interest to practitioners of risk analysis.