OBJECTIVE: To build a deterministic compartmental model for exploring the effects on the transmission of human immunodeficiency virus (HIV) of a population abstaining from sex or practising only "safe" sex for one month each year.
METHODS: A model of HIV transmission was built to simulate the effects of the intervention (i.e. an annual no-sex or safe-sex month in which no transmission occurred) in three countries, under several optimistic assumptions. The reduction in the modelled annual incidence of transmission that was attributable to this "test" intervention was compared with that seen with an alternative intervention. In the latter, monthly incidences of transmission were each reduced by one twelfth, so that, essentially, the month-long interruption was spread evenly across a full year.
FINDINGS: Over the first modelled year, the test intervention averted only 2.5% (Kenya), 3.3% (South Africa) and 1.6% (Swaziland) more HIV infections than the alternative interruption. According to the model, if the test intervention were repeated each January, it would avert only 2% (Kenya), 2% (South Africa) and 1% (Swaziland) more HIV infections over 5 years than the alternative intervention.
CONCLUSION: Although it did not appear markedly more effective than the alternative intervention, the test intervention may still be more feasible and therefore worthwhile. Before the test intervention can be recommended, the cost-effectiveness and feasibility of such an annual month-long break in HIV transmission need to be assessed and compared with those of other interventions that may reduce new HIV infections, such as circumcision and concurrency-reduction campaigns.
