Biological validation of self-reported condom use among sex workers in Guinea
Aho, Jéphine.; Koushik, A.; Diakité, Sïla.Laye.; Loua, K.Marcel.; Nguyen, V-Kim.; Rashed, Sélim.
Aids and Behavior 14(6): 1287-1293
2010
ISSN/ISBN: 1090-7165 PMID: 19680799 DOI: 10.1007/s10461-009-9602-6Document Number: 207025
Self-reported condom use may be prone to social desirability bias. Our aim was to assess the validity of self-reported condom use in a population of female sex workers using prostate specific antigen (PSA) as a gold standard biomarker of recent unprotected vaginal intercourse. We collected data on 223 sex-workers in Conakry, Guinea in order to assess the sensitivity and specificity of self-reported condom use as well as to examine the predictors of discordance between self-report and PSA presence. PSA was detected in 38.4% of samples. Sensitivity of self-reported condom use was 14.6% and its specificity was 94.7%. Self-perceived high risk of HIV infection was the only significant independent predictor of misreported condom use. PSA could be useful to validate self-reported condom use in surveys and to allow a better understanding of factors associated with social desirability in sexual behaviour reporting.
Document emailed within 0-6 h
