Generation of LRs

We generated positive LRs and confidence intervals based on the rate of the entity under study in positives (true positive rate) compared with the rate of entity under study in negatives (false positive rate), (a/a + c)/(b/b + d), where a = true positive, b = false positive, c = false negative, and d = true negative.²² We generated a negative LR based on the rate of absence of the entity under study in negatives (true negative rate) compared with the rate of absence of the entity under study in positives (false negative rate), (d/b + d)/(c/a + c). When 1 or more cells contained 0 counts, we universally applied to those analyses a Haldane correction (adding 0.5 to each cell): this correction dampens a signal of association toward the null and thus is inherently conservative.