[Last update: 12/2/03]
We adopt the acronym RD to refer to any rate difference, risk difference, or prevalence "rate" difference. This parameter quantifies the absolute effect of the exposure by estimating the excess (or deficiency) in the number of cases in a population of a given size. Thus, if we let R1 represent the rate (or risk) in the exposed group and R0 represent the rate or risk in the nonexposed group, RD = R1 - R0.
Illustrative example. Suppose an exposed group has a risk (R1) of 2 per 100 and a non-exposed group has a risk (R0) of 1 per 100. Thus, RD = (2 per 100) - (1 per 100) = 1 per 100. This predicts one additional case per 100 exposures.
Notice that the baseline RD of 0 indicates no association between the expose and disease, a positive RD signifies a positive association (risk), and a negative RD signifies a negative association (benefit).