Determination of the intra- and interlaboratory reproducibility of the low volume eye test and its statistical relationship to the Draize eye test.

The reproducibility of toxicologic test methods, including alternative tests, is a key scientific and regulatory concern. In the present work, historic rabbit eye irritation data were used to determine the intra- and interlaboratory reproducibility of the low volume eye test (LVET). The standard Draize eye irritation test was used as the basis for comparison. The LVET and Draize tests had similar degrees of intra- and interlaboratory reproducibility as determined by examination of their coefficients of variation, although the variability in LVET results was directionally lower. Results from 70 parallel Draize and LVET tests indicated a strong positive association between results from the two tests, for corneal, iridial, conjunctival, and maximum average scores (MAS). Correlation coefficients were 0.60, 0.73, 0.69, and 0.73, respectively (P </= 0.0001). The linear relationship between LVET and Draize MAS values was examined by regression analysis and found to follow the relationship LVET MAS = 0.522 (Draize MAS). Thus, the LVET is at least as reproducible as the Draize test and gives responses that are (linearly) correlated to the Draize. The previous findings that the LVET is more predictive of human eye responses than the Draize test lends additional support for its use as a refined alternative to the Draize test.