versão impressa ISSN 0042-9686
Bull World Health Organ vol.79 no.1 Genebra Jan. 2001
STI care: one of many necessary approaches for prevention of HIV infection
Challenging orthodoxy always has an appeal to some, and we must confess we are part of that group. In his article (1), Hudson is challenging a hastily built orthodoxy derived from the results of the Mwanza trial that spread rapidly and largely without question throughout the public health world. Following publication of the trial results, which seemed to indicate that management of sexually transmitted infection (STI) could result in significant reduction of HIV incidence, many were quick to conclude that an important AIDS-prevention tool was at hand (2). The fact that the original hypothesis of the study, linking STI management and HIV transmission, was not proven was eclipsed by an important decline in HIV incidence in Mwanza at a time when the public health community needed hope. Perhaps STI prevalence did not decline as much as HIV incidence, reasoned the authors, because of the high prevalence of asymptomatic infection among women. It is not possible to address asymptomatic infection with syndromic management, the main STI management strategy used in the Mwanza trial.
Enter the results of the Rakai trial, which was meant to address the problem of asymptomatic STI by presumptive mass treatment. However, for reasons made clear by Hudson and others (1, 35), the Rakai trial failed to reduce HIV incidence at all. Hudson suggests this failure may stem from a misreading of the results of the Mwanza trial. To assume that the link between STI management and HIV transmission was conclusively proven at Mwanza is to ignore the nuances and context of the trial: these must be considered in order to make sense of the complicated epidemiology of HIV/AIDS.
Many facts argued for caution in interpreting the Mwanza results: the well-known observation that first trials commonly produce more positive results than subsequent trials; the fact that the original hypothesis was not really proved; and the seemingly inexorable drive to find a single intervention, a magic bullet, that will make control of the epidemic possible. It has long been known that no magic bullet exists nor is likely to for some time to come, and that a multiplicity of approaches is needed (6). One key approach must be to encourage behavioural change. Changing sexual behaviour, the first element of any public health programme to control HIV, is a strategy that often enjoys too little confidence in such programmes, despite the fact that there are no national examples in which the spread of HIV has been diminished or brought under control without behavioural change playing an important, if not the key role.
Efforts to prevent HIV infection by encouraging behavioural change are either being increasingly ignored (7) or are being forced to compete against other approaches for already insufficient resources. Successful interventions involving behavioural change, are we know now not notable for their sophistication, but rather for their consistency, intensity, and duration. More research is needed to determine if there are ways to achieve the same ends with less effort. Currently, however, we know how to prevent HIV infection, we lack only the will.
If the architects of these two key trials in Mwanza and Rakai can be blamed for anything, it is for paying insufficient attention to behaviour and behavioural change, either in their efforts to reduce sexual risk (8) or, as Hudson points out (1), by failing to consider important behavioural changes already taking place. If the interpreters of the Mwanza trial can be held accountable for anything, it is for overstating the results and trying to justify STI management through HIV control. STI care certainly has a role in HIV prevention, but it is also an important public health activity. Overselling STI management as the sole intervention for HIV control will only damage efforts to address an epidemic that has long preceded HIV/AIDS.
The history of public health efforts in AIDS prevention will undoubtedly show the folly of ignoring what we know in favour of what we might prefer. Hudson has reminded us (1), as we must be reminded frequently, it seems, that there is no single answer, that multiple approaches must be used probably everywhere, and that behavioural change remains a key component everywhere. To place undue effort on any one intervention is ill-advised and certainly not justified by the evidence produced to date.
1. Hudson CP. Community-based trials of sexually transmitted disease treatment: repercussions for epidemiology and HIV prevention. Bulletin of the World Health Organization, 2001, 79: 4858.
2. Laga M. STD control for HIV prevention. Lancet, 1995, 346: 518519.
3. Hitchcock P, Fransen L. Preventing HIV: lessons from Mwanza and Rakai. Lancet, 1999, 353: 513515.
4. Grosskurth H et al. Control of sexually transmitted diseases for HIV-1 prevention: understanding the implications of the Mwanza and Rakai trials. Lancet, 2000, 355: 19811987.
5. Wawer M et al. Preventing HIV-1: lessons from Mwanza and Rakai. Lancet, 1999, 353: 15231524.
6. Cates W, Hinman AR. AIDS and absolutism the demand for perfection. New England Journal of Medicine, 1992, 327: 492494.
7. Merson M, Rosenfeld A. The AIDS epidemic: lessons learned? Lancet, 2000, 356: 1204.
8. OReilly K et al. Impact of improved treatment of STD on HIV. Lancet, 1995, 346: 1158.
1 Scientist, Department of Reproductive Health and Research, World Health Organization, 1211 Geneva 27, Switzerland. Correspondence should be addressed to this author.
2 Team Coordinator on Prevention, HIV/AIDS/STI Initiative, World Health Organization, Geneva, Switzerland.
Ref. No. 00-1122
Much debate has followed the publication of the results of the Rakai trial in 1999 (1). Given the outcome of the Mwanza trial published a few years earlier (2), most of the AIDS scientific community was expecting positive results from Rakai. Indeed, how could periodic mass treatment for sexually transmitted disease (STD) fail to prevent HIV-1 transmission at the community level when a much more modest intervention (appropriate STD treatment for people with STD symptoms attending primary health care centres) had led to a 38% reduction in HIV-1 incidence?
The main reason evoked up until now for these discrepant results is the difference in the stage of the HIV epidemic at the study sites (3, 4). Indeed, as confirmed by a sub-analysis of the Rakai data (5), the population attributable fraction for STD in HIV-1 transmission will be quite low when the HIV epidemic reaches a high but stable prevalence with a low to moderate incidence. Other reasons put forward have been the increase in herpes simplex virus type 2 infection (an incurable STD) at the population level in mature epidemics, which could then contribute to further HIV-1 transmission; the possibility of reinfection occurring between mass treatment cycles; the possibility that symptomatic STDs play a more important role in HIV-1 transmission than asymptomatic STDs; and chance 4), with imbalance between the study arms at baseline resulting in wide confidence intervals (6).
In this issue of the Bulletin of the World Health Organization, Hudson presents new arguments (7), adding to the debate. He draws three main points: firstly, in contrast to most previous analyses, he considers that the Rakai trial should be considered as the gold standard, and it is the Mwanza results which need to be explained, mainly because reinfection between mass treatment rounds is unlikely to be a major factor accounting for the negative results from Rakai. Secondly, behavioural counselling, rather than STD treatment, could explain the Mwanza results, mainly because STD symptoms, in people contracting HIV-1 and an STD simultaneously, may prompt attendance at a clinic prior to manifestations of primary HIV-1 infection; this would lead to abstinence or consistent condom use during the earliest period of viraemia associated with primary infection. Thirdly, a community-based randomized trial of improved counselling alone versus improved counselling combined with improved syndromic management of STDs is urgently needed.
It is likely that the counselling component of STD syndromic management has a more important impact on HIV-1 transmission than the STD treatment per se. Indeed, the results of a recent mathematical modelling exercise mimicking an intervention which started in 1993 in Cotonou, Benin, suggest that increased condom use averted many more cases of HIV-1 infection than cases of STD treated (8). Additional data from Cotonou also suggest that simultaneous exposure to HIV and an STD is frequent, especially among female sex workers (9). However, I consider that STD management at the primary health care level should include both counselling and effective STD treatment. It is unacceptable to set these two interventions in opposition. STD treatment must reach the highest standards possible, even if only for the proper management of STDs, which have their own important impact on the populations health, especially that of women. The real urgency is to increase the scope of interventions that are currently far too modest. For instance, it is very likely that the impact of STD treatment on HIV-1 transmission would be more significant if interventions had greater intensity and population coverage (8). There is no need for more trials at the community level: instead, comprehensive interventions, taking into account the context in which they will be applied (e.g. urban versus rural settings; late versus early stage of the epidemic), should be designed, appropriately funded, implemented, and evaluated.
Furthermore, reinfection between mass treatment rounds could, in fact, explain the Rakai trial results. Both the surveys and the mass treatment interventions in Rakai were household based. This type of survey tends to miss people whose sexual behaviour puts them more at risk (e.g. female sex workers), because they also tend to be the most mobile or not to live in a standard household. We have recently shown that, although the coverage of the population in the Rakai trial was high, differential coverage of the low- and high-risk sectors of the population could well have led to a very pronounced decrease in the impact of the intervention, as a result of reinfection and persistent enhancement of HIV-1 transmission by untreated STDs in the high-risk group (10). This raises the issue of whether interventions should be targeted at the most vulnerable sectors of the population. Although the importance of core groups has been recognized since the early phase of the HIV-1 epidemic (11), interventions specifically targeting these sectors of the population have not been sufficiently promoted. In a recent extensive review of the data available from Cotonou, Benin, we have shown that most of the new HIV-1 cases in the general population were linked to sexual networks between female sex workers, their clients, and the other, mostly regular, female sexual partners of the clients (9). Thus we, like others (12), strongly believe that interventions aimed at the prostitution milieu including condom promotion, individual counselling and STD care should be of the utmost priority for the control of the HIV-1 epidemic in developing countries. This is particularly important in places where the epidemic is increasing, as it is in Asia and still is in many countries of sub-Saharan Africa (e.g. West Africa). If there is one community-based randomized trial yet to be performed, it should aim to assess the impact of such interventions on HIV-1 transmission at the population level.
The author would like to thank Catherine M. Lowndes and Marie-Claude Boily for their helpful comments on the initial version of this commentary.
1. Wawer M et al. Control of STD for AIDS prevention in Uganda: a randomised community trial. Lancet, 1999; 353: 525535.
2. Grosskurth H et al. Impact of improved treatment of STD on HIV infection in rural Tanzania: randomised controlled trial. Lancet, 1995; 346: 530536.
3. Hitchcock P, Fransen L. Preventing HIV: lessons from Mwanza and Rakai. Lancet, 1999; 353: 513515.
4. Grosskurth H et al. Control of sexually transmitted diseases for HIV-1 prevention: understanding the implications of the Mwanza and Rakai trials. Lancet, 2000; 355: 19811987.
5. Gray R et al. Relative risks and population attributable fraction of incident HIV associated with symptoms and treatable STD. AIDS, 1999; 13: 21132123.
6. Korenromp EL, De Vlas S, Habbema M. Effect of imbalances at baseline in community-randomized trials. XIII International AIDS Conference, Durban, South Africa, 914 July 2000 (Abstract WePpC1313).
7. Hudson CP. Community-based trials of sexually transmitted disease treatment: repercussions for epidemiology and HIV prevention. Bulletin of the World Health Organization, 2001, 79: 4858.
8. Boily MC, Lowndes CM, Alary M. Factors influencing the spread of HIV in sub-Saharan Africa and their implications for prevention at different stages of the HIV epidemic: Insights from mathematical models. Paper presented at: Symposium on Phase-Specific Strategies for the Prevention, Control and Elimination of Sexually Transmitted Diseases: Implications for Research, Policies and Programs, Rome, Italy, 36 October 2000.
9. Lowndes CM et al. Phases of HIV/STD epidemics in developing countries: the case of Cotonou, Benin. Paper presented at: Symposium on Phase-Specific Strategies for the Prevention, Control and Elimination of Sexually Transmitted Diseases: Implications for Research, Policies and Programs, Rome, Italy, 36 October 2000.
10. Boily MC, Lowndes CM, Alary M. Complementary hypotheses concerning the community STD mass treatment puzzle in Rakai, Uganda. AIDS, 2000; 14: 25832592.
11. Plummer FA et al. The importance of core groups in the epidemiology and control of HIV-1 infection. AIDS, 1991; 5 (Suppl. 1): S169S176.
12. Ainsworth M, Teokul W. Breaking the silence: setting realistic priorities for AIDS control in developing countries. Lancet, 2000, 356: 5560.
1 Visiting Professor, Department of Infectious and Tropical Diseases, London School of Hygiene and Tropical Medicine, Keppel Street, London, WC1E 7HT, England.
Ref. No. 00-1121