Sexual behaviour of people living with HIV attending a tertiary care government hospital in Kathmandu, Nepal: a cross sectional study

Rapid development in the field of ART has led to reduction in morbidity and mortality
among PLHIV. However, there is a growing concern over sexual risk behaviour of PLHIV
on ART. The changes in sexual risk behaviour of PLHIV with the initiation of ART might
have important implications on the way counseling may be designed and implemented.
An increase in sexual risk behaviour with therapy will be a matter of great concern
from a prevention of transmission perspective. If not intervened, it may further fuel
the epidemic.

In this study, only around half of the respondents reported to be sexually active
in the past six months. It is possible that lower sexual activity may be a result
of symptomatic disease or a result of adverse effects of ART 23]. The psychological trauma resulting from a positive diagnosis might also reduce sexual
desire. A majority of the respondents in both groups (ART experienced and ART naïve)
had sex only with their regular partners in the past six months. Those who had other
types of partners always used condoms during sex except for one male ART naïve PLHIV
who never used condom with his casual partner. This meant that sexual risk behaviour
was mostly dependent upon the respondents’ condom use with their regular partners.
This is similar to other studies where most reported sex was with regular partners
17], 30], 44]. In India, 96 % of those who were sexually active in the past six months reported
sex with a regular partner 17]. However, we should consider the influence of both recall and social desirability
bias on responses related to sexual behaviour when interpreting these results.

In regular partner relationship, non-disclosure of self’s HIV status, lack of knowledge
of partner’s HIV status in combination with unprotected sex lays out favourable circumstances
for transmission of HIV to serodiscordant partners. Furthermore unprotected sex with
regular partner also carries the risk of unwanted pregnancy and the transmission of
HIV virus to the child. Nearly a fifth (20 %) of the respondents who had sex with
regular partners were in a discordant relationship with an HIV negative partner but
a majority of them (93 %) had protected sex. Likewise, 58 % were in a concordant relationship
with an HIV positive partner but a relatively lower percentage (66 %) reported protected
sex. Unprotected sex with HIV positive partners carries the risk of becoming infected
with resistant viral strains or reinfection with novel strains 30].

With regular partners, ART experienced PLHIV were more likely to have protected sex
compared to ART naïve PLHIV, which was similar to other studies conducted in a developing
country setting 17], 25], 30]. Thus, the study provides no evidence to suggest that sexual risk behaviour may actually
increase with initiation of ART in this context 23], 25], 26], 28]–30]. Belief about the reduced severity and threat of the disease due to the availability
of HAART as well as those about HAART related HIV transmission might have contributed
to increased sexual risk behaviour 24] particularly among ART naïve PLHIV compared to ART experienced PLHIV. On the other
hand, lower risk behaviour among ART experienced may be explained when taking into
account their larger mean time since HIV diagnosis (43 months) compared to ART naïve
PLHIV (29 months). This means a longer time had elapsed since getting tested and diagnosed
with greater chances of having advanced HIV disease as well. Consequently they may
perceive the seriousness of their illness differently and should have had several
contacts with health workers including counsellors who conveyed prevention and protection
messages during the course of treatment. This may have contributed to lower sexual
risk behaviour. In addition, similar to a study in India, the belief that condoms
are needed for ART to be effective might have facilitated consistent condom use with
regular partners 45] among ART experienced PLHIV. Among married couples, the desire to have children might
also be a predictor of unprotected sex with regular partners 46] but it was not investigated in this study. In order to get a clear understanding
of the underlying reasons for unprotected sex, it might be necessary to investigate
individual motives through a qualitative inquiry.

Females were more likely to have unprotected sex compared to males. The decision to
use/not use condom was taken mostly by the males. Females seem to have very little
say when it comes to deciding on condom use. This might have contributed to this finding.
A study from Asia found that wealthier and highly educated married women were more
likely to report that they can refuse sexual intercourse and ask their husbands to
use a condom 47]. Additionally the study had also found that women’s overall involvement in making
family decisions empowers them to negotiate safer sex 47].

Alcohol consumption during last sex with regular partners was one of the significant
predictors of unprotected sex. The use of alcohol may decrease the ability to make
rational judgements and provoke risk behaviours as unprotected sex. Studies conducted
in different populations and high risk sub-groups have attempted to examine the relationship
where a number of them report an association between risk behaviour and alcohol consumption
45], 46], 48]–52]. On the contrary, a meta-analysis to assess the relationship of alcohol use and condom
use did not find an association 53].

Similar to the limitations of other previous studies, 44] this current cross-sectional study could not derive causal inferences. Further, the
study defined sex as penile vaginal insertive sexual events which meant that other
risky sexual behaviour including penile-anal sex between men and between men and women
were not accounted for. In the multivariate analysis, the reduction in sample size
may have resulted in wider confidence intervals. Thus, the conclusions may be tentative,
especially with regards to the non-significant findings. The fact that sex of the
person was statistically significant only in the multivariate analysis, once ART status
was controlled for may indicate an interaction effect, but this could not be tested
with the small sample size. These limitations should be taken into consideration when
interpreting the results of this study. Also, since sexual behaviour was self-reported
it may have been underestimated 28], 30], 44]. Furthermore, it is also important to note the possible influence of social desirability
on study results.