Search

Objective: The objective of this study was to assess the validity of women's reports of recent unprotected sex by testing for prostate-specific antigen (PSA) in vaginal samples., Study Design: The authors conducted prospective research with 332 female sex workers attending 2 public dispensaries in Madagascar., Results: Among women who reported no sex or protected sex only within the past 48 hours, 21% and 39%, respectively, tested positive for PSA. Among those testing positive for PSA, no differences in PSA concentrations were found among those reporting no sex, protected sex only, or at least one unprotected act., Conclusions: The substantial disagreement between self-reports and measurement of a biologic marker of semen exposure in vaginal specimens substantiates that self-reports of sexual behavior cannot be assumed to be valid measures. Future sexually transmitted infection/HIV and pregnancy prevention studies should confirm the validity of self-reports or use end points that do not rely on self-reported data.

Molecular methods that characterize the Neisseria gonorrhoeae porin protein Por are needed to study gonococcal pathogenesis in the natural host and to classify strains from direct clinical samples used with nucleic acid amplification-based tests. We have defined the capabilities of por variable region (VR) typing and determined suitable conditions to apply the method to direct clinical specimens. Nested PCR from spiked urine samples detected 1 to 10 copies of template DNA; freezing spiked whole urine greatly reduced the ability to amplify porB. In a laboratory model of mixed gonococcal infections, the por type of one strain could be determined in the presence of a 100-fold excess of another. por VR typing was used to examine clinical samples from women enrolled in studies conducted in Baltimore, Md., and Madagascar. por type was determined from 100% of paired cervical swab and wick samples from 20 culture-positive women from Baltimore; results for eight individuals (40%) suggested infection with more than one strain. In frozen urine samples from Madagascar, porB was amplified and typed from 60 of 126 samples from ligase chain reaction (LCR)-positive women and 3 samples from LCR-negative women. The por VR types of 13 samples (21%) suggested the presence of more than one gonococcal strain. Five por types, identified in >45% of women with typed samples, were common to both geographic areas. Molecular typing is an important adjunct to nucleic acid amplification-based diagnostics. Methods that utilize direct clinical samples and can identify mixed infections may contribute significantly to studies of host immunity, gonococcal epidemiology, and pathogenesis.