Introduction

Forecasting the evolution of the AIDS epidemic in different situations is a primary problem which has to be solved in order to perform cost-benefit analyses and evaluate different policies which can be adopted to control the epidemic.

Several mathematical and statistical models have been proposed for these purposes in recent years. Most of them are deterministic models which try to take into account very complex situations. Others are statistical models which simply try to fit data and extrapolate the observed beaviours (some references can be found in Rossi (1991)). The most suitable approach to modelling AIDS epidemic is via stochastic models, based on an understanding of the transmission dynamics of HIV.

In the following, a stochastic compartmental model, recently proposed (Rossi 1991), is used to perform some scenario analyses and evaluate the impact of different policies.

The model used here concentrates, in particular, on the proportion of stayers, st, throughout, i.e. the proportion of individuals of the population who are not at risk of infection by HIV, either through behaviour or by possible immunity, and on the proportion of infected individuals who can be considered removed from the transmission process due to their ‘prudent’ behaviour.

The flow chart and the equations of the model can be found in Rossi (1991). In the present contribution we concentrate on the influence on the behaviour of the epidemic of the parameters s0 and vij where v12, v23, v32 a nd v21 are four parameters that represent the effects of screening and associated treatment programmes.