TABLE 4.

Detailed ORs for the logistic regression model of the association between virulence factors and STEC isolates from serotypes reported in humans, including the eae^*stx₂ interaction (set 1)^a

Comparison	OR
eae positive stx2 negative vs eae negative stx2 negative	1.74
eae negative stx2 positive vs eae negative stx2 negative	2.36
eae positive stx2 positive vs eae positive stx2 negative	13.20
eae positive stx2 positive vs eae negative stx2 positive	9.68
eae positive stx2 positive vs eae negative stx2 negative	22.87

^aeae positive, isolate carrying the eae gene; eae negative, isolate lacking the eae gene; stx₂ positive, isolate carrying the stx₂ gene; stx₂ negative, isolate lacking the stx₂ gene. These results indicate, for instance, that an eae-positive and stx₂-positive STEC isolate is 13.2 times more likely to be from a serotype previously reported in humans than an eae-positive but stx₂-negative isolate.