relprev

Relprev, short for relative prevalence, is a statistical measure used to compare how common a particular feature, condition, or outcome is across two groups. In practice, relprev is defined as the ratio of prevalences: RelPrev = p1 / p0, where p1 is the prevalence in the target group and p0 is the prevalence in the reference group. A relprev greater than 1 indicates that the feature is more prevalent in the target group; less than 1 indicates it is less prevalent; equal to 1 indicates parity.

Relprev is commonly used in cross-sectional analyses, epidemiology, and market research to assess disparities or differences in frequency between populations. It is closely related to the prevalence ratio and is conceptually similar to the risk ratio, though it is typically framed in terms of prevalence rather than incidence. It should be distinguished from the odds ratio, which compares odds rather than direct probabilities and can diverge from relprev when prevalences are high.

Calculation considerations include handling zero prevalence in the reference group, which makes the ratio undefined; researchers may apply continuity corrections or alternative measures in such cases. Confidence intervals for relprev can be obtained through standard binomial methods or regression-based approaches.

Example: if 12% of a treated group and 6% of a control group exhibit a condition, RelPrev = 0.12 / 0.06 = 2.0, suggesting the condition is twice as prevalent in the treated group.

Limitations include sensitivity to baseline prevalence and potential confounding factors. Relprev is one tool among several for understanding group differences in prevalence.