SciELO - Scientific Electronic Library Online

vol.79 issue5Interpretation of the Mwanza and Rakai STI trialsThe world in 2050: more crowded, urban and aged author indexsubject indexarticles search
Home Page  

Bulletin of the World Health Organization

Print version ISSN 0042-9686

Bull World Health Organ vol.79 n.5 Genebra Jan. 2001 

Interpretation of the Mwanza and Rakai STD trials

Editor – There has been considerable discussion of the contrasting results of the Mwanza and Rakai trials of interventions against sexually transmitted diseases (STDs). The Mwanza trial showed that improved STD treatment services as recommended by WHO led to a 38% reduction in HIV incidence (1), while in the Rakai trial mass treatment for STDs failed to reduce HIV infection (2). Several thoughtful reviews have discussed these findings, and in June last year scientists from the two trials combined forces to present their views (3). A number of explanations have been put forward: STDs may play a less important role in HIV spread in mature HIV epidemics; incurable STDs, including herpes, may have been more common in Rakai; and rounds of mass treatment in Rakai may have been followed by rapid reinfection.

In a recent contribution to the Bulletin, Dr Hudson puts forward an alternative view (4), arguing that the trial results failed to demonstrate that STD treatment reduces HIV transmission, and that the HIV effect seen in Mwanza was due instead to behavioural change. Unfortunately, while we welcome open debate, we do not think Hudson’s paper will do much to illuminate the present controversy, since most of his assertions are inconsistent with published data.

Hudson argues that the HIV effect in Mwanza could not have been due to improved STD treatment because the effect on STDs was only ‘‘modest’’. However, the prevalence of symptomatic urethritis was cut by half, and newly acquired syphilis by at least 40% (5). The prevalence of chancroid was not measured, but the syphilis data suggest that the prevalence of genital ulcer disease (GUD) was reduced very substantially. There is now overwhelming evidence that GUD plays a key role in HIV spread in Africa, and it seems very likely that the reduction in GUD in Mwanza made an important contribution to the observed decrease in HIV incidence. Interestingly, the Rakai intervention had no effect at all on newly acquired syphilis. The reported effect on syphilis refers to the prevalence of positive serology, and this fell due to treatment of positives at mass treatment rounds. However, most of these were low titre cases, probably representing latent infections that are of little relevance to HIV transmission.

Hudson suggests the HIV impact in Mwanza resulted from behavioural change due to counselling of STD patients at clinics or to information campaigns in the villages. The latter focused on informing the community about the improved treatment services, and it seems very unlikely that this would have had a measurable effect on sexual behaviour. Hudson identifies one variable in the published data that seems to show a difference between the intervention and comparison communities in women, but this is an isolated and non-significant finding and is not borne out in other variables or in men. The overwhelming impression from the many behavioural variables examined was that there was no difference between the trial arms. Promoting sexual abstinence or condom use until treatment is complete is an important component of STD case management. However, the idea that this could have a substantial impact on the overall level of HIV transmission in the community, even if some patients were in the viraemic phase of primary HIV infection, seems fanciful. Moreover, condom uptake in these rural areas at the time of the trial was extremely low, despite counselling, and only 0.9% of STD patients accepted free condoms.

Finally, Hudson misinterprets recently published data from Mwanza on population-attributable fractions (PAF) of HIV due to symptomatic STDs (6). The fact that in the comparison arm 40% of HIV seroconversions in men were attributable to self-reported STD syndromes, but only 12% in the intervention arm, strongly supports the hypothesis that reductions in symptomatic STDs led to a reduction in HIV. Calculations show that this difference would explain nearly 80% of the observed effect on HIV (6). The striking contrast between the strong association of GUD and HIV in the comparison arm (OR = 14.8) and lack of association in the intervention arm (OR = 1.1) is very difficult to explain other than through an effect of treatment on GUD duration and hence on HIV acquisition. Data from Rakai showed a much lower PAF due to treatable STDs.

Hudson suggests a trial of counselling of STD patients alone versus counselling and improved STD treatment. This is ethically unacceptable as it would mean actively withholding care for harmful infections from patients attending improved facilities in the control communities.

Drs O’Reilly & Gerbase, in their accompanying commentary (7), stress that STD treatment is not a ‘‘magic bullet’’, but one of many approaches needed for HIV prevention. We fully agree. As stated in the original Mwanza report: ‘‘Educational interventions, aimed at modifying risk behaviour ... remain an important priority for national AIDS control programmes. However, our results suggest the importance of complementing these activities with improvements to STD treatment services’’ (1). That STDs are an important risk factor for HIV transmission is beyond question. Effective STD management has been identified as a fundamental component of primary care, not only because of the effect of STDs on HIV but their many other adverse effects. While continuing to promote behavioural change, AIDS control programmes must continue to ensure that this basic provision is available to all.


Richard Hayes
Professor of Epidemiology and International Health


Heiner Grosskurth
Senior Lecturer in Infectious Disease Epidemiology


David Mabey
Professor of Communicable Diseases
London School of Hygiene and Tropical Medicine,
Keppel Street, London WC1E 7HT, England


1. Grosskurth H et al. Impact of improved treatment of sexually transmitted diseases on HIV infection in rural Tanzania: randomised controlled trial. Lancet, 1995, 346: 530–536.

2. Wawer MJ et al. Control of sexually transmitted diseases for AIDS prevention in Uganda: a randomised community trial. Lancet, 1999, 353: 525–535.

3. 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: 1981–1987.

4. Hudson CP. Community-based trials of sexually transmitted disease treatment: repercussions for epidemiology and HIV prevention. Bulletin of the World Health Organization, 2001, 79: 48–58.

5. Mayaud P et al. Improved treatment services significantly reduce the prevalence of sexually transmitted diseases in rural Tanzania: results of a randomised controlled trial. AIDS, 1997, 11: 1873–1880.

6. Orroth K et al. Syndromic STD treatment reduces the proportion of incident HIV infections attributable to these diseases in rural Tanzania. AIDS, 2000, 14: 1429–1437.

7. O’Reilly KR, Gerbase AC. STI care: one of many necessary approaches for prevention of HIV infection. Bulletin of the World Health Organization, 2001, 79: 58–59.