A semiparametric approach to risk assessment for quantitative outcomes.

Characterizing the dose-effect relationship and estimating acceptable exposure levels are the primary goals of quantitative risk assessment. A semiparametric approach is proposed for risk assessment with continuously measured or quantitative outcomes which has advantages over existing methods by requiring fewer assumptions. The approach is based on pairwise ranking between the response values in the control group and those in the exposed groups. The work generalizes the rank-based Wilcoxon-Mann-Whitney test, which for the two-group comparison is effectively a test of whether a response from the control group is different from (larger than) a response in an exposed group. We develop a regression framework that naturally extends this metric to model the dose effect in terms of a risk function. Parameters of the regression model can be estimated with standard software. However, inference requires an additional step to estimate the variance structure of the estimated parameters. An effective dose (ED) and associated lower confidence limit (LED) are easily calculated. The method is supported by a simulation study and is illustrated with a study on the effects of aconiazide. The method offers flexible modeling of the dose effect, and since it is rank-based, it is more resistant to outliers, nonconstant variance, and other departures from normality than previously described approaches.

[1]  S. Zeger,et al.  Longitudinal data analysis using generalized linear models , 1986 .

[2]  D. Barnes,et al.  Reference dose (RfD): description and use in health risk assessments. , 1988, Regulatory toxicology and pharmacology : RTP.

[3]  Kenny S. Crump,et al.  Calculation of Benchmark Doses from Continuous Data , 1995 .

[4]  D W Gaylor,et al.  Incidence of developmental defects at the no observed adverse effect level (NOAEL). , 1992, Regulatory toxicology and pharmacology : RTP.

[5]  Ronald E. Gots.  Toxic Risks: Science, Regulation, and Perception , 1993 .

[6]  D. Gaylor,et al.  Analysis of trinomial responses from reproductive and developmental toxicity experiments. , 1991, Biometrics.

[7]  David W. Gaylor,et al.  Dose-response modeling of quantitative response data for risk assessment , 1992 .

[8]  Haseman Jk,et al.  Analysis of dichotomous response data from certain toxicological experiments. , 1979 .

[9]  D W Gaylor,et al.  Issues in qualitative and quantitative risk analysis for developmental toxicology. , 1988, Risk analysis : an official publication of the Society for Risk Analysis.

[10]  Ralph L. Kodell,et al.  Statistical methods of risk assessment for continuous variables , 1993 .

[11]  L. Ryan,et al.  Statistical model for fetal death, fetal weight, and malformation in developmental toxicity studies. , 1993, Teratology.

[12]  K S Crump,et al.  A new method for determining allowable daily intakes. , 1984, Fundamental and applied toxicology : official journal of the Society of Toxicology.

[13]  J. Hanley,et al.  The meaning and use of the area under a receiver operating characteristic (ROC) curve. , 1982, Radiology.

[14]  L Ryan,et al.  Quantitative risk assessment for developmental toxicity. , 1992, Biometrics.

[15]  Distribution-free Confidence Intervals for Pr(X1 < X2) , 1987 .

[16]  R L Kodell,et al.  Upper confidence limits on excess risk for quantitative responses. , 1993, Risk analysis : an official publication of the Society for Risk Analysis.

[17]  W Slikker,et al.  Risk assessment for neurotoxic effects. , 1990, Neurotoxicology.

[18]  Irwin Guttman,et al.  Statistical inference for Pr(Y < X): The normal case , 1986 .

[19]  W. Piegorsch.  Biometrical methods for testing dose effects of environmental stimuli in laboratory studies , 1993 .

[20]  Douglas A. Wolfe,et al.  On Constructing Statistics and Reporting Data , 1971 .