Background: Under-reporting of sexual behavior in HIV prevention efforts compromise the accuracy of predicting sero-discordant couples likely to transmit infections to negative partners. Here we examine the reliability of self-reports of unsafe sexual behaviors using 4 biological measures of unprotected sex and the association between time to unprotected sex and predictors collected at enrollment and during follow up.
Methods: HIV discordant heterosexual couples were followed at a CVCT center from 1994 to 2012. Baseline and time varying factors were measured. The outcome was a time-varying composite measure of any: self-reported unprotected sex, sperm present on a wet prep, incident pregnancy, and incident HIV/STI. A repeated outcomes survival analysis with time-varying covariates explored factors associated with unprotected sex.
Results: Among 3,049 couples followed for an average of two years, any indication of unprotected sex occurred during 62% of study intervals. In the unadjusted analysis, younger age of the male partner ((M-F+: HR 0.995 p=0.002, M+F-: HR 0.99 p <0.0001), female partner (M-F+: HR 0.995 p=0.011, M+F-: HR 0.992 p =0.003), fewer number of years co-habiting (M-F+: HR 0.992 p=0.008, M+F-: HR 0.991 p =0.003), age difference (M-F+: HR 0.995 p=0.04, M+F-: HR 0.991 p=0.02), male partner use of alcohol (M+F-: HR 1.114 p=0.01), female partners number of lifetime sex partners (M-F+: HR 1.092 p=0.002), circumcision of the male partner (M+F-: HR 0.886 p=0.032), self-reported condom use by female partner (M-F+: HR 0.812 p<0.0001, M+F-: HR 0.756 p <0.0001), being pregnant at interval visits (M-F+: HR 1.304 p<0.0001, M+F-: HR 1.341 p <0.0001) were significant predictors of unprotected sex. Multivariate analyses showed that a higher number of lifetime sex partners for the female partner (aHR: 1.003, 95% CI: 1.002, 1.004), a positive history of STI in the past year for the male partner (aHR: 1.149, 95% CI: 1.062, 1.243) and being pregnant at any interval visit (aHR: 1.191, 95% CI: 1.08, 1.314) were significant predictors for the M-F+ couples while consistent protected sex with the study partner (aHR: 0.795, 95% CI: 0.658, 0.961), being pregnant at any interval visit (aHR: 1.253, 95% CI: 1.113, 1.41) and use of either oral (OCP) or injectable (INJ) contraceptives at interval visits were important predictors of unprotected sex for the M+F- couples.
Conclusions: Our study provides further evidence that gross under-reporting of unsafe sexual practices is very likely by couples in discordant relationships and caution must be exercised when using self-reports to determine a discordant couple's risk of HIV transmission to the negative partner.
Table of Contents
Abstract Cover Page
Table of Contents
BACKGROUND AND LITERATURE REVIEW. 3
STRENGTHS AND WEAKNESSES. 26
About this Master's Thesis
|Subfield / Discipline|
|Committee Chair / Thesis Advisor|
|Incidence and Predictors of Any Incident Indicator of Unprotected Sex among a Cohort of HIV sero-discordant Couples in Lusaka, Zambia (1994 2012) ()||2018-08-28 13:09:55 -0400||