REPORTED CONDOM USE IS NOT ASSOCIATED WITH INCIDENCE OF SEXUALLY-TRANSMITTED DISEASES IN MALAWI

Objectives: To establish frequency of reported condom use and validate reliability of self-reporting among urban women in Malawi. Design: Cr oss-sectional survey in antenatal women in 1989 and 1993. Prospective study in cohort first surveyed in 1989. Methods: A total of 6561 women in 1989 and 2460 women in 1993 answered questions about condom use an d sexual activity, had a physical examination and were screened for HI V. A subset of women from the 1989 screening were administered a quest ionnaire and tested for syphilis, Neisseria gonorrhoeae and Trichomona s vaginalis infections every 6 months. Results: Although between the t wo cross-sectional studies intermittent condom use increased from 6 to 15% (P < 0.001) with no difference according to HIV infection, consis tent use was reported by less than 1%. In the prospective study, women reported a higher condom use at any visit than either group assessed cross-sectionally. Consistent condom use peaked at 62% in the first 6 months, but declined to as low as 8% in the second year of follow-up. Condom use at each visit, either intermittent or consistent, was highe r in HIV-seropositive than HIV-seronegative women. Overall, the incide nce of gonorrhea, trichomoniasis and syphilis did not decline in women reporting consistent condom use. Conclusions: In prospectively follow ed women reports of consistent condom use was substantially higher tha n in cross-sectional surveys, but rapidly decreased over time, irrespe ctive of HIV status. The presence of new sexually transmitted diseases suggests that this population of urban women overreports condom use o r underreports sexual activity, or both. Intensive and sustained educa tion is needed to achieve consistent condom use. Biologic markers of s exual activity are useful in interpreting reported condom use.