|
LLNA Report Appendix N Page 1 of 3: NICEATM Assessment of the Effect of Different Stimulation Index (SI) Levels on Performance of the LLNA Previous Appendix 1 2 3 (WARNING - SLOW LOADING - ~900K) Next Page Return to the NICEATM-ICCVAM Document Search Page Data on maximal dose tested and maximal SI response for each test substance included in Appendix A were obtained, when available, and used to generate a database capable of being analyzed for the effect of different SI criteria on sensitivity, specificity, positive predictivity, negative predictivity, and accuracy for the LLNA. The revised list, containing only chemicals were SI data were located, and for which guinea pig and/or human data were available, is attached. Multiple entries (highlighted in the list) for the same test substance were included where multiple tests had been conducted. Where the same data were present in multiple citations, only the earliest citation is provided. Arbitrary fold-increase SI criteria for a positive call (i.e., 4.0, 3.5, 2.4, 2.0) in addition to the standard increase SI criteria of 3.0 were used to distinguish a positive response from a negative one. The resulting calls were used to compare the sensitivity, specificity, positive predictivity, negative predictivity, and accuracy of the LLNA versus:
The results of these analyses are presented in the accompanying table. In making these comparisons and to be consistent with the previous evaluation, (1) discordant LLNA results (i.e., where multiple tests were conducted, with some positive and some negative calls) which could not be reconciled by inspection, were classified as negative; (2) equivocal HMT results were classified as positive; and (3) in cases where a negative result was recorded for the HMT but the substance was used as a HPTA, the chemical was classified as positive for human senistization. In regard to item (1), one data set was omitted from each of 3 chemicals (cinnamic aldehyde, formaldehyde, sodium lauryl sulfate) as indicated in accompanying data list, because the low response was associated with a maximum dose considerable lower than that used in the other tests. The resulting analyses indicates that an SI of 3.0 is a reasonable criteria for classifying an LLNA response as positive. Previous Appendix 1 2 3 (WARNING - SLOW LOADING - ~900K) Next Page Return to the NICEATM-ICCVAM Document Search Page |
