OBJECTIVE: The objective of this study was to determine concordance in sex partner reports of frequency of vaginal and oral intercourse and condom use. STUDY DESIGN: Self-reported behaviors among 112 women treated for urinary tract infection (UTI), sore throats, or receiving birth control at a student health service were compared with that of their current sex partner. RESULTS: Correlation was excellent for frequencies of vaginal intercourse (r = 0.88), cunnilingus (r = 0.72), fellatio (r = 0.76), and condom use (r = 0.86), but women with UTI compared with women without UTI reported greater frequency of sexual activity than their sex partner. This effect overwhelmed the modest increases in correlation associated with duration of relationship and marital status. CONCLUSIONS: Women with genitourinary symptoms may over-recall frequency of sexual behavior compared to women without these symptoms. Knowing the extent of this recall bias will assist in the interpretation of data and appropriate correction of parameters when included in prediction models.
OBJECTIVE: The objective of this study was to determine concordance in sex partner reports of frequency of vaginal and oral intercourse and condom use. STUDY DESIGN: Self-reported behaviors among 112 women treated for urinary tract infection (UTI), sore throats, or receiving birth control at a student health service were compared with that of their current sex partner. RESULTS: Correlation was excellent for frequencies of vaginal intercourse (r = 0.88), cunnilingus (r = 0.72), fellatio (r = 0.76), and condom use (r = 0.86), but women with UTI compared with women without UTI reported greater frequency of sexual activity than their sex partner. This effect overwhelmed the modest increases in correlation associated with duration of relationship and marital status. CONCLUSIONS:Women with genitourinary symptoms may over-recall frequency of sexual behavior compared to women without these symptoms. Knowing the extent of this recall bias will assist in the interpretation of data and appropriate correction of parameters when included in prediction models.