HIV sero-discordance among married HIV patients initiating anti-retroviral therapy in northern Vietnam
There was a high sero-discordance rate among the married couples in our study, and it was higher among men than among women. The sero-discordance rate shown in this study is higher than that reported in other countries, e.g. Thailand (58%) [7] or Nigeria (25%) [8].
High sero-discordance rate was associated with male sex, history of IDU and diagnosis of TB. Patients with a history of IDU might be less sexually active, hence posing less of a risk of transmitting the virus to their spouse, also shown in an earlier study [9]. Patients with a history of IDU were commonly diagnosed early as part of the on-going sentinel surveillance program among risks groups. This may explain the counterintuitive result that patients with known HIV infection for more than 2 years had a higher sero-discordance rate than patients with known infection less than 2 years. Early HIV detection through sentinel surveillance or VCT before development of severe immunosuppression or indication for ART might sensitise the patients to preventive measures such as condom use or abstinence, as shown in a study from Ethiopia [10]. In Dong Trieu, where no sentinel surveillance activities or VCT were available until 2002 compared to 1995 in Ha Long City, the sero-concordance rate was significantly higher.
As shown, patients with higher education had a significantly lower sero-discordance rate. This could be due to a more favourable social status in the community, making the individual more socially vulnerable to accidental disclosure and stigmatisation. This factor may be a strong disincentive to HIV testing [11, 12], thereby delaying diagnosis and increasing the duration of exposure and risk for HIV transmission to the partner. In Vietnam, a high proportion of men have visited female sex workers [13]. As the latter are a vulnerable group with a high prevalence of HIV and Sexually Transmitted Diseases (STDs) [14], the male partners may have been infected not only with HIV, but also STDs such as syphilis, herpes type 2 and gonorrhoea, all of which increase the risk for HIV transmission [15].
There was a much higher rate of sero-discordance among male married patients than female married patients. A study in Italy concluded that the efficiency of male-to-female transmission was 2.3 times greater than that of female-to-male transmission [16]. Thus, a higher discordance rate would be expected among married male patients with an equal sex distribution of index patients. This factor, combined with the finding that men were more severely immunosuppressed than women, indicating a longer duration of HIV infection, may lead to the conclusion that the male partner was the index case in most cases and infected their female partner. This idea is also supported by the high proportion of widows (40%) in the study population.
High plasma viral load that may increase the risk of transmitting HIV to their sexual partners [17]. This idea is supported by the general trend of increasing HIV prevalence among women, particularly those who live in areas with a high prevalence of male PWID [18, 19]. In our cohort we could not find a significant correlation between high baseline viral load and level of sero-discordance (Table 2).
Factors related to HIV sero-discordant/concordant status
As all the patients were included in the study at the start of ART, the effect of viral suppression on transmission rates is not a likely cause of the high sero-discordance rate, although this has been shown to be the case in other studies [20]. TB may occur at a modest level of immunosuppression, and individuals diagnosed with TB are tested for HIV. These factors might have led to earlier diagnosis and hence a reduction in the exposure time for the spouses. The deterioration in health, as well as common stigmatisation of TB, may have decreased the sexual activity and hence the exposure of the spouse [9].
The high sero-discordance rates of the HIV+ married men may indicate that a large part of the population was aware of preventive measures, especially in the areas where VCT has been carried out for a longer duration. However, the high number of female spouses at risk of exposure emphasises the need for continued and increased measures to prevent transmission such as Information, Education and Communication programs and provision of condoms in the community to prevent HIV transmission among couples [21, 22]. In Zambia, it has been reported that the ideological stance in the U.S. President’s Emergency Plan for AIDS Relief to promote abstinence and faithfulness before condoms (the ‘ABC’ approach) might have deterred the organisations involved in the implementation of the program from giving information and providing condoms to at-risk populations [23]. The ABC approach has not been so influential in Vietnam where condoms have been promoted as a core part of the prevention program. This may have influenced the comparably high rates of sero-discordance reported, in combination with other factors such as the prevalence of IDU and the rate of sexual HIV transmission.
Early initiation of ART reduce sexual transmission, especially among sero-discordant couples, the incidence of Opportunistic Infections (OIs) and death [24–31]. The risk for transmission is proportional to the viral load in combination with other factors that increase the risk for blood to blood contact as STD’s, genital lesions and anal intercourse. Hence viral load monitoring is important to assess both the effect of treatment as well early detection of increased risk for HIV transmission, especially among sero-discordant couples (27). In seen in Table 2 there were no significant difference in viral load between sero-concordant and discordant or male and females. In many low-income countries, patients are generally diagnosed at a late stage of the disease and the ART coverage is in many places still low [28]. As seen in our study, there was limited access to ART, even for those patients who were severely immunosuppressed and with high viral load. The lack of access was mainly due to the program’s regulation, under which the number of patients started on ART each month was limited to avoid overloading health staff.
There are some methodological considerations in this study. Non-naïve was an exclusion criteria, hence patients that had used ART prior to enrolment were not asked to join the RCT. Therefore, this cohort might not be representative of HIV patients in Vietnam. Indicators such as IDU and history of TB were self-reported, and this might have caused an underestimation. Information regarding number of years married, number of sex partners or sexual intercourse behaviour was not collected. Thus, the data might not represent the magnitude of all predictors of sero-discordance among married patients. As the study included only patients who registered at OPCs, information on patients who have not yet registered is lacking. Thus, the study does not fully reflect the entire characteristics of People Living with HIV in Vietnam.