Article Text

Download PDFPDF

S05.5 Network formation as a determinant of spatial disparity
  1. Ian Spicknall,
  2. HW Chesson,
  3. CG Patel,
  4. TL Gift,
  5. KT Bernstein,
  6. SO Aral
  1. Centers for Disease Control and Prevention, Atlanta, USA


Introduction Racial disparities in the burden of sexually transmitted disease have been documented and described for decades. Several disparity measures, such as the black-to-white rate ratio, the Index of Disparity (ID), and the Gini coefficient, have been used to quantify disparities in STDs. Although these measures are generally useful and informative, in some scenarios the results of these measures may seem incongruous with reasonable, practical assessments of disparity. The purpose of this study was to provide examples of such incongruities in the context of racial disparities in STDs.

Methods We examined a population consisting of 312.5 million people overall, with 200 million non-Hispanic Whites, 40 million non-Hispanic Blacks, 55 million Hispanics, 15 million Asian/Pacific Islanders, and 2.5 million American Indian/Alaskan Natives (AI/AN), similar to the actual US population. Case numbers of a hypothetical STD for each racial group were chosen arbitrarily (yet purposefully) to yield illustrations of potential incongruities among disparity measures.

Results We found several hypothetical examples of incongruities among disparity measures. For example, the ID was about ten times higher when all STD cases occurred among AI/AN than when all cases occurred among non-Hispanic Blacks. As another example, the ID indicated that disparity was less when all STD cases occurred among non-Hispanic Blacks than when each of the five racial groups accounted for one fifth of all STD cases.

Conclusion Relative measures of racial disparity in STDs can be useful to illustrate the burden of disparity, to assess trends, and to inform the targeting of prevention resources. However, in some scenarios the disparity measures can be incongruous with reasonable, practical assessments of disparity, such as when the ID is biased against non-Hispanic Blacks. The ID is more prone to these incongruities than measures which account for population size, such as the Gini coefficient or the weighted ID.

Disclosure of interest statement The authors have no conflicts to declare. No pharmaceutical grants were received in the development of this study.

Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.