The 130th Annual Meeting of APHA

Abstract #39378

Victor J. Schoenbach, PhD, Department of Epidemiology, University of North Carolina at Chapel Hill, 2104D McGavran-Greenberg Hall, Chapel Hill, NC 27599-7435, 919-966-7436, victor_schoenbach@unc.edu, Kenneth G Keppel, PhD, Division of Health Promotion Statisics, CDC, NCHS, 6525 Belcrest Rd., Rm 770, Hyattsville, MD 20782, Chester Scott, Division of Vital Statistics, CDC, NCHS, 6525 Belcrest Rd., Rm 820, Hyattsville, MD 20782, and John W. Lynch, PhD, School of Public Health, University of Michigan, 109 Observatory, Room M3116, Ann Arbor, MI 48109-2029.

A difference or disparity is the quantity that separates two measures. Public health is particularly interested in disparities, because their existence often indicates opportunities for improving health status of population groups through the application of existing knowledge and identification of new causal or preventive factors. A basic methodological question is whether to quantify a disparity as an arithmetic difference (absolute comparison) or as an arithmetic ratio (relative comparison). An absolute comparison presents the disparity in the units of the measures being compared. A relative comparison, in contrast, may have either no units or entirely different units.

Both absolute and relative disparity measures are legitimate, and both are informative. They represent complementary perspectives and provide different kinds of information. Each has limitations, however, and ratio comparisons present problems of interpretation akin to changes of scale.

This paper compares absolute and relative measures for monitoring disparities in rates and proportions over time. We propose a hybrid measure, in which a relative difference is based on an externally-derived denominator, as a way of combining some of the advantages of both absolute and relative perspectives while avoiding some of their disadvantages. Although we focus on comparisons between rates for two groups, similar principles apply to situations where more than two groups are being compared.

Learning Objectives:

• Articulate the difference between absolute and relative comparisons. Describe why both absolute and relative comparisons provide useful information. Describe one way to combine infromation based on absolute and relative comparisons.

Keywords: Minority Health, Statistics

Presenting author's disclosure statement:
I do not have any significant financial interest/arrangement or affiliation with any organization/institution whose products or services are being discussed in this session.