Objective: To assess how the impact of sexually transmitted disease (STD) treatment on HIV incidence varies between stages of the HIV epidemic.
Methods: We simulated the spread of curable STD, herpes simplex virus type 2 and HIV in the dynamic transmission model. Parameters were quantified to represent a severe HIV epidemic as in Rakai, Uganda, using demographic, behavioural and epidemiological data from a recent STD treatment trial.
Results: The model fitted the HIV epidemic in Rakai if we assumed a considerable behavioural risk reduction, starting at the end of the Ugandan civil war in 1986. An improvement in STD treatment reduced HIV incidence in this population by 35% over 2 years if implemented in 1981, but only by 11 and 8% in 1988 or 1998. This trend resulted partly from the hypothesized behaviour change, which markedly reduced the prevalences of bacterial STD. In a simulated epidemic without behavioural change, the corresponding treatment impacts in 1988 and 1998 would be 19 and 15%. Enhanced herpetic ulceration in immunocompromised HIV patients contributed little to the reduced impact of treatment of bacterial STD over time.
Conclusion: In HIV epidemics beyond the first decade, the impact of STD treatment programmes on HIV transmission may depend more on behavioural risk reduction than on the stage of the epidemic. Preceding behavioural change associated with restored civil stability may have contributed to the lack of impact of STD treatment on HIV in the Rakai trial. In advanced epidemics with less behaviour change, STD treatment may still be important for HIV prevention.