Linking prevention and access to treatment
As efforts to provide treatment are scaled up, concerns have been voiced regarding its potential impact on preventive behaviours. The availability of treatment, some fear, may lower people's perception of risk and hence lead to lower vigilance; in contrast, others argue that the strengthening of health-related interventions will encourage testing and counselling, and that the knowledge of HIV status may increase protective behaviour. Studies of people living with HIV in developing countries indicate that treatment is, indeed, associated with increased sexual activity but, at the same time, it is associated with more consistent condom use. In developed countries, an increase in risky behaviour was documented among certain population groups after the introduction of effective antiretroviral therapy, without clear evidence of why this happens (2, 3). The epidemiological data are only suggestive, and trends need to be documented across settings, over time, and among key subpopulations, in particular in people living with HIV/AIDS and other highly exposed groups.
Treatment may directly contribute to prevention of new infections to the extent that lowering of viral load decreases the probability of sexual transmission. Decreased infectiousness is likely to be counterbalanced by increases in the life expectancy of patients. Research must contribute to the adaptation of interventions in order to ensure the sustainability of prevention over the long term (4, 5).
More generally, better evidence is needed on how preventive behaviours can be promoted across age groups, sexes, social strata and different categories of serological status, especially in the context of scaling up access to treatment. The mechanisms linking risk perceptions to behavioural change are shaped by contextual variables and are situation-specific and partner-specific (6). Better evidence is also needed on the extent to which the findings from developed countries are valid for developing countries. There are also indications that gender influences the selection of strategies to reduce risk (7). It is important to investigate whether such differences in perceived risks and protection represent a general gendered pattern.
Microlevel studies indicate that medical information and public health recommendations are not automatically accepted by the general public and that people interpret professional advice in the light of local notions and past experience with health care (8). Investigations of beliefs and practices that are contrary to public health recommendations can suggest ways to communicate better with the public. The availability of effective treatments should contribute to furthering confidence in medicine and public health.
Preventing transmission from mother to child
Among the issues that urgently require further research are better methods of ensuring the prevention of HIV transmission from mother to child, particularly in developing countries and in the postnatal period. Every year, an estimated 700 000 children become infected with HIV. The overwhelming majority acquire the virus through mother-to-child transmission, which can occur either during pregnancy and delivery or postnatally during breastfeeding. In the absence of any intervention, rates of this form of transmission can vary from 15% to 30% without breastfeeding, and reach as high as 45% with prolonged breastfeeding (9). Transmission during the peripartum period accounts for one-third to two-thirds of overall numbers infected, depending on whether breastfeeding transmission occurs or not, and this period has therefore become a focus of prevention efforts.
Transmission of HIV from mother to child can be prevented almost entirely by anti-retroviral drug prophylaxis (usually now given in combinations), elective caesarean section before onset of labour and rupture of membranes, and refraining from breastfeeding (10, 11). In resource-poor countries, however, elective caesarean section is not a safe option, and refraining from breastfeeding is often not feasible or acceptable. In Africa, no more than 5% of HIV-infected women and neonates who could benefit from interventions are receiving them.
Antiretrovirals, either alone or in combinations of two or three drugs, have been shown to be highly effective in reducing mother-to-child transmission of HIV. Evidence of the efficacy of antiretroviral prophylaxis from Africa, Europe, Thailand and the USA has been demonstrated for short-course drug regimens. The substantial efficacy of triple combinations has been shown in observational studies in industrialized countries (10, 11), where rates of transmission are now below 2% in the absence of breastfeeding. There is an urgent need for evidence from breastfeeding mothers in sub-Saharan Africa, one of the most affected populations. Short-term safety and tolerance of the prophylactic regimens have been demonstrated in all the controlled clinical trials on mother-to-child transmission (12). Further studies of these issues are required, especially on the long-term implications of potential antiretroviral resistance for HIV-infected mothers and their children.
Preventive interventions with antiretrovirals have not yet been successfully implemented on the scale required (13). Even where antiretroviral treatment is applied peri-partum, infants remain at substantial risk of acquiring infection in the breastfeeding period. These facts also require investigation; they demonstrate once again the need to strengthen health systems, while integrating HIV/AIDS interventions with reproductive and maternal and child health services.
Protecting women with microbicides
Protecting women against HIV infection is another important area for researchers. Microbicides are anti-infective products such as gels, creams, impregnated sponges and similar devices that women apply before sexual intercourse to prevent HIV transmission and other sexual infections. Unlike the condom, their use is controlled by the woman and they will not necessarily be contraceptive. Attempts are also being made to incorporate microbicides into silicone intravaginal rings that are left in the vagina for several weeks to ensure contraception, with sustained release of the agent providing continuous protection against infection.
Microbicides could make a very substantial difference by widening people's choice of protective interventions. To achieve high levels of use will require a continuing education process aimed at women as well as health policy officials and providers. Epidemiological modelling based on data from over 70 low-income countries suggests that even a partially effective microbicide is likely to have a significant impact on the epidemic: a product that is only 60% effective in protecting against HIV could avert 2.5 million new infections over a three-year period, even if it is used in only 50% of sex acts not protected by condoms, and assuming it is used by only 20% of people easily reached by existing health services (14). However, the microbicide concept has only recently received sufficient support to allow progress to be made. Pharmaceutical companies have not so far regarded microbicides as providing economic incentives for substantial investment, though the Bill and Melinda Gates Foundation is now giving serious consideration to this matter.