Validity of self-report: Data from a sexual risk-reduction trial with African-American female adolescents

Background: Self-reported behavior has been the cornerstone of measurement in sexual health research. Self-report of sexual behaviors may be prone to biases, including recall and social desirability. In this study we compared adolescent females' self-reported condom use in the past 14 days to a biological indicator of sexual activity.

Methods: Participants were enrolled in an HIV prevention program. Only data from baseline, prior to randomization, were used for this analysis. African-American females, aged 15-21, reporting sexual activity in the past 60 days were eligible to participate (N=715). The primary analysis examined concordance between adolescents' reported condom use, assessed by ACASI and YcPCR assay, which detects the presence of sperm in vaginal fluid for 14 days following unprotected vaginal sex.

Results: Of participants with a valid YcPCR assay and reporting vaginal sex in the past 14 days (N=484), 186 reported 100% condom use. Of these, one-third (N=63) had a positive YcPCR. Participants reporting consistent condom use with a history of STDs were 2.4 times more likely to have a positive YcPCR (adjusted OR=2.4, 95% CI=1.2-4.8; P=.01).

Conclusions: We observed a significant degree of discordance between self-report of condom use and YcPCR positivity. There are rival explanations for the observed discordance. One is that adolescents inaccurately report condom use. Another is that condoms were used, but were not used correctly, resulting in user “error”. And a third explanation is that adolescents responded with socially desirable answers. Using an objective biological measure may provide one strategy for validating adolescents' self-reported condom use.