Abstract
Executive summaryBackground and objectivesSouth Africa has one of the highest HIV incidence rates in the world. Although much research has focused on developing biomedical strategies to reduce HIV incidence, there has been less investment in prevention strategies that address the social drivers of HIV spread. Understanding the social determinants of HIV is closely related to understanding high-risk populations (‘key populations’), since many of the factors that place these key populations at high HIV risk are social and behavioural rather than biological.Mathematical models have an important role to play in evaluating the potential impact of new HIV prevention and treatment strategies. However, most of the mathematical modelling studies that have been published to date have evaluated biomedical HIV prevention strategies, and relatively few models have been developed to understand the role of social determinants or interventions that address these social drivers. In addition, many of the mathematical models that have been developed are relatively simple deterministic models, which are not well suited to simulating the complex causal pathways that link many of the social drivers to HIV incidence. The frequency-dependent assumption implicit in most deterministic models also leads to under-estimation of the contribution of high-risk groups to the incidence of HIV.Agent-based models (ABMs) overcome many of the limitations of deterministic models, although at the expense of greater computational burden. This study presents an ABM of HIV in South Africa, developed to characterize the key social drivers of HIV in South Africa and the groups that are at the highest risk of HIV. The objective of this report is to provide a technical description of the model and to explain how the model has been calibrated to South African data sources; future publications will assess the drivers of HIV transmission in South Africa in more detail.MethodsThe model is an extension of a previously-published ABM of HIV and other sexually transmitted infections (STIs) in South Africa. This model simulates a representative sample of the South African population, starting from 1985, with an initial sample size of 20 000. The population changes in size as a result of births and deaths. Each individual is assigned a date of birth, sex and race (demographic characteristics). This in turn affects the assignment of socio-economic variables. Each individual is assigned a level of educational attainment, which is dynamically updated as youth progress through school and tertiary education, with rates of progression and drop-out depending on the individual’s demographic characteristics. Each individual is also assigned to an urban or rural location, with rates of movement between urban and rural areas depending on demographic characteristics and educational attainment.The model assigns to each individual a number of healthcare access variables that determine their HIV and pregnancy risk. These include their ‘condom preference’ (a measure of the extent to which they wish to use condoms and are able to access condoms), use of hormonal contraception and sterilization, use of pre-exposure prophylaxis (PrEP), male circumcision, HIV testing history and uptake of antiretroviral treatment (ART). Access to these healthcare services changes over time, and is also assumed to depend on demographic and socioeconomic variables, as well as on the individual’s health status.Sexual behaviour is simulated by assigning to each individual an indicator of their propensity for concurrent partnerships (‘high risk’ individuals are defined as individuals who have a propensity for concurrent partnerships or commercial sex). Each individual is also assigned a sexual preference, which can change over their life course. Three types of relationship are modelled: sex worker-client contacts, short-term (non-marital) relationships and long-term (marital or cohabiting) relationships. Individuals are assumed to enter into short-term relationships at rates that depend on their risk group and demographic characteristics. Each time a new short-term partner is acquired, the individual is linked to another individual in the population, with the probability of linkage depending on the individual’s sexual preference and preference for individuals of the relevant age, risk group, race, location and educational attainment. Individuals marry their short-term partners at rates that depend on their demographic characteristics. Frequencies of sex are assumed to depend on demographic characteristics and relationship type, and migrant couples are assumed to have reduced coital frequency. Probabilities of condom use also depend on demographic characteristics and relationship type, and are assumed to be strongly associated with levels of educational attainment.Women’s risk of falling pregnant is assumed to depend on their sexual behaviour, natural fertility level, contraceptive usage and breastfeeding status. Adoption and discontinuation of hormonal contraception is assumed to depend on demographic characteristics, sexual behaviour and past pregnancy and contraceptive experience. Girls who fall pregnant while in school are assumed to be less likely to complete their schooling than those who do not fall pregnant.Probabilities of HIV transmission per act of sex are assumed to depend on several biological factors, including the viral load of the HIV-positive partner, whether the HIV-positive partner is on ART, the presence of other STIs, the type of contraceptive used, the age and sex of the susceptible partner, male circumcision, the type of relationship, and the use of new HIV prevention methods such as PrEP. If an individual acquires HIV, they are assigned a CD4 count and viral load, both of which change dynamically over the course of HIV infection. The HIV mortality risk is determined by the individual’s CD4 count. HIV-positive individuals are diagnosed at rates that depend on their demographic characteristics and CD4 count, and if they disclose their HIV status to their sexual partners after diagnosis, this is assumed to lead to increased rates of condom use. Assumptions about HIV transmission probabilities have been set in such a way that the model matches the observed trends in HIV prevalence, by age and sex, in national South African antenatal and household surveys.The model also simulates male incarceration. Rates of incarceration are assumed to depend on men’s demographic characteristics and educational attainment, and are also assumed to be higher in men who have previously been incarcerated.Results and conclusionsThe model matches reasonably closely the observed levels of HIV prevalence in South Africa by age and sex, as well as the observed changes in HIV prevalence over time. The model also matches observed patterns of HIV prevalence by educational attainment, by urban-rural location and by history of recent migration. Estimates of HIV prevalence in key populations (sex workers, MSM and prisoners) are roughly consistent with surveys. The model has also been calibrated to match total numbers of HIV tests and male circumcision operations performed in South Africa. The model estimates of levels of HIV diagnosis and ART coverage are consistent with the Thembisa model, an HIV model that has been calibrated to South African HIV testing and ART data.Although many of the phenomena simulated in the MicroCOSM model have been simulated in previously-published HIV models, MicroCOSM is the first model that systematically describes all of these phenomena in a fully integrated model. This makes it possible to use the model to describe complex interactions between socio-economic and behavioural factors, and their influence on disease and health-seeking behaviour. It also provides a framework for understanding socio-economic and racial inequality in health outcomes in South Africa, and for assessing the potential impact of strategies to reduce these inequalities.
Publisher
Cold Spring Harbor Laboratory