Patients who did not disclose their HIV status and who did not use condoms were more likely to be in relationships in which their spouse seroconverted. A study from South Africa reported that non-disclosure of HIV status to partners was associated with increased HIV transmission risk-taking behaviours [44]. Although rates of condom use in the current study increased during the 12 months of follow-up, more patients in seroconverting relationships did not use condoms than patients who were in serodiscordant relationships. The increasing use of condoms may be related to the regular risk reduction

counselling and free condoms provided by counsellors selleck chemical to HIV-infected patients and their spouses at each clinic visit. Earlier studies have documented inconsistent condom use among

serodiscordant couples [3], and that women in particular may find it difficult to negotiate condom use [4]. In India, female sterilization has historically been used as a means of family planning rather than broader reproductive health programmes that include contraception, prevention of STIs and addressing sexual violence against women [45]. Future interventions among serodiscordant couples will need to develop strategies to decrease alcohol consumption, promote HIV disclosure and normalize the use of condoms. The current study shares some of the methodological limitations OSI906 of similar observational studies related to sexual risk behaviour assessment based on self-reported behaviours, which may be affected by social desirability. Also the proportion of infections acquired from outside of marital relationships cannot be quantified. ADAMTS5 The current analysis only included data collected from the index patient who first enrolled into care, and similarly

to other epidemiological studies it was assumed that the index patient had first been infected with HIV and had consequently infected his/her partner. Certain factors that could affect HIV transmission, such as socio-economic characteristics, sexual violence and circumcision, were not systematically collected by clinicians and counsellors at every visit, and hence were not included in the present study. HIV status was assessed using antibody-based tests, and hence detection of acute infection using HIV-1 RNA quantification techniques was not done. Although patients in this study period may have been excluded, this is unlikely as the serostatus of spouses in serodiscordant relationships was examined on follow-up clinic visits. The present study was not designed to examine the pattern of exact transmission. It is very unlikely that transmission occurred outside the partner dyad as most individuals who seroconverted were women and our prior data have shown that most Indian women testing HIV positive at our clinic are in monogamous relationships [24]. The heterosexual transmission of HIV involves the complex interaction of both biological and behavioural factors.