Article Text


Social and behavioural aspects of prevention poster session 1: Adolescents
P2-S1.18 Accuracy of self-reported condom use assessed by the semen Y-chromosome biomarker for unprotected sex
  1. J Rosenbaum1,
  2. J Melendez2,
  3. E Rose3,
  4. G Wingood3,
  5. J Zenilman2,
  6. R DiClemente3
  1. 1University of Maryland School of Public Health, Washington, USA
  2. 2Johns Hopkins Medical Institutions, USA
  3. 3Emory School of Public Health, USA


Background Adolescents may use condoms inconsistently or incorrectly, and may over-report their condom use. Biomarkers offer a potential solution to over-report, allowing interventions to target populations at risk from inconsistent condom use. This study uses a biomarker for unprotected sex to evaluate the accuracy of adolescent womens reports of condom use.

Methods The sample comprised 715 African-American female participants in an HIV prevention trial who were ages 15–21, sexually active, and recruited from three urban Atlanta family planning and STI clinics in 2002–2004. At baseline, 6, and 12 months, participants completed a 40 min ACASI interview and were tested for semen Y-chromosome with PCR from a self-administered vaginal swab. Data analysis used non-parametric tests for trend and clustered logistic regression to predict self-reported pregnancy.

Results At the three waves respectively 30%, 20% and 15% of participants who reported 100% condom use tested positive for semen Y chromosome. Suspected over-reporters of condom use were more likely to report being pregnant at the following wave than any other category of condom and birth control users: at wave 2, 20.4% of wave 1 suspected over- reporters were pregnant vs 14.2% of wave 1 condom never-users and 9.8% of those reporting no birth control at last sex; at wave 3, 16.2% of wave 2 suspected over-reporters were pregnant vs 11.8% of condom never-users and 10.8% reporting no birth control at last sex. Suspected over-reporters of unprotected sex were more likely to report pregnancy at the following wave, controlling for frequency of sex, condom use, and oral contraception (OR 3.95 (1.63 to 9.58), 2.64 (0.92 to 7.58)) see Abstract P2-S1.18 Table 1.

Abstract P2-S1.18 Table 1

Prediction of pregnancy with logistic regression. Analysis was limited to those who participated in all 3 waves of the survey (n=560). Did not report sex past 60 days are those who participated in the wave but did not answer the question how many times they had sex in the past 60 days. Condom use in the past 60 days was imputed from condom use in the past 14 days for these individuals. Predictors of pregnancy are measured in the wave previous to pregnancy

Conclusions Respondents who report 100% condom use but test positive for semen Y-chromosome are at higher risk than other women. These women may be over-reporting their condom use and under-reporting their frequency of sex.

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.