Assessing the Performance of Three HIV Incidence Risk Scores in a Cohort of Black and White MSM in the South (vol 44, pg 297, 2017)

• Type of Material: Article
• Authors: Jeb Jones, Emory University; Martin Hoenigl, University of California, San Diego; Aaron J Siegler, Emory University; Patrick S Sullivan, Emory University; Susan Little, University of California, San Diego; Eli S Rosenberg, Emory University
• Language: English
• Date: 2018-03-01
• Publisher: Lippincott, Williams & Wilkins
• Publication Version: Author Accepted Manuscript (After Peer Review)
• Copyright Statement: © 2017 American Sexually Transmitted Diseases Association
• Final Published Version (URL): https://journals.lww.com/stdjournal/FullText/2017/05000/Assessing_the_Performance_of_3_Human.8.aspx
• Title of Journal or Parent Work: Sexually Transmitted Diseases
• ISSN: 0148-5717
• Volume: 45
• Issue: 3
• Start Page: E13
• End Page: E13
• Grant/Funding Information: Sources of support: National Institutes of Health
• Keywords: Science & Technology; Infectious Diseases; Life Sciences & Biomedicine
• Research Categories: Health Sciences, Immunology; Health Sciences, Epidemiology

Abstract

Background Risk scores have been developed to identify men at high risk of HIV seroconversion. These scores can be used to more efficiently allocate public health prevention resources, such as pre-exposure prophylaxis. However, the published scores were developed with datasets that comprise predominantly white men who have sex with men (MSM) collected several years prior and recruited from a limited geographic area. Thus, it is unclear how well these scores perform in men of different races or ethnicities or men in different geographic regions. Methods We assessed the predictive ability of three published scores to predict HIV seroconversion in a cohort of black and white MSM in Atlanta, GA. Questionnaire data from the baseline study visit were used to derive individual scores for each participant. We assessed the discriminatory ability of each risk score to predict HIV seroconversion over two years of follow-up. Results The predictive ability of each score was low among all MSM and lower among black men compared to white men. Each score had lower sensitivity to predict seroconversion among black MSM compared to white MSM and low area under the curve values for the ROC curve indicating poor discriminatory ability. Conclusions Reliance on the currently available risk scores will result in misclassification of high proportions of MSM, especially black MSM, in terms of HIV risk, leading to missed opportunities for HIV prevention services.