Abstract of Meeting Paper

Society for Risk Analysis 2002 Annual Meeting

Application of Benchmark Dose Modeling To Reduce Uncertainty in Dose-Response Assessment. M. E. Stelljes and R. R. Wood, SLR International Corp.

An occupational exposure level (OEL) was developed for inhalation exposure to n-propyl bromide (nPB) using EPA’s Benchmark Dose Software (BMDS). NPB is proposed for industrial use as a replacement for certain ozone-depleting chlorinated solvents. A benchmark dose (BMD) was derived on the basis of dose-response data from acute, chronic, and two-generation rodent studies. Data for over 30 endpoints were entered into the BMDS for calculation of a BMD and benchmark dose low (BMDL). The BMD is equal to a target benchmark response (e.g., toxicity in 10 percent of exposed animals) and the BMDL is the lower 95th confidence limit on the BMD. Once values were calculated for all relevant studies and endpoints, they were evaluated using a weight-of-evidence approach to develop a single BMD and BMDL representative of nPB toxicity. The BMDL is similar to the reference dose (RfD). USEPA’s BMD approach represents a departure from the way RfDs were historically developed. In the past, the no-observed adverse effect level (NOAEL) was divided by uncertainty factors to develop an RfD. If a NOAEL was unavailable, a lowest-observed adverse effect level (LOAEL) was used. This approach depends on the doses used in the critical studies. It ignores other dose levels and does not attempt to interpolate response as a function of dose. Advantages of the BMD approach include: (1) the entire dose-response curve is utilized instead of a single dose, (2) a measure of variability based on actual data is incorporated into the calculation (i.e., BMDL), (3) responses within the experimental range are used, rather than extrapolation to low doses not experimentally tested, and (4) a consistent benchmark response level (e.g., 10 percent) is used across studies. As a result, the BMDL is more accurate than a NOAEL or LOAEL, and lower uncertainty factors are needed to adjust the BMDL to a RfD or OEL.

The authors acknowledge the support of Enviro Tech International.


Go to . . .

2002 SRA Annual Meeting Table of Contents
2002 SRA Annual Meeting Author Index
Main Abstracts Menu Page
RiskWorld Home Page