Reliability of self-reported sexual behavior in human immunodeficiency virus (HIV) concordant and discordant heterosexual couples in northern Thailand.

A partner study was conducted in northern Thailand between March 1992 and June 1996 which included data that allowed an assessment of the reliability of self-reports of sexual behavior and contraceptive use among heterosexual couples. The authors enrolled 529 couples among whom all male subjects were human immunodeficiency virus (HIV) seropositive voluntary blood donors and their female sexual partners were either HIV infected (n = 246) or HIV seronegative (n = 283). The levels of agreement within couples were assessed for recency of last sexual intercourse, sexual activity in the prior year, and contraceptive practices. For HIV discordant couples, a prospective study was conducted to examine risk factors for HIV transmission, the primary goal of the study. This allowed assessment of reliability of inter-partner reports over 6-12 months. Overall, agreement among couples was good for common sexual practices, especially vaginal intercourse and time since last intercourse, but was lower for condom use. Anal and oral sex were infrequently reported by these couples and there was greater disagreement for the occurrence of these practices. Partner agreement for contraceptive histories was good to excellent. Prospective data showed less frequent intercourse and more condom use but reliability remained good. Common sexual practices may be reliable for both HIV concordant and discordant couples in studies estimating prevalent infection. Estimates of incident heterosexually transmitted HIV may be made with greater reliability by studies which include assessment of reports of risk behavior by each member of a couple than studies of individuals.