Cover Image
close this bookEmerging Patterns of HIV Incidence in Uganda and other East African Countries (International Center for Migration and Health - ICMH, 1997, 97 pages)
close this folderAN ASSESSMENT OF EMERGING PATTERNS OF HIV INCIDENCE IN UGANDA AND OTHER EAST AFRICAN COUNTRIES
close this folder4.0 Simulation modelling of HIV incidence dynamics in Uganda with validation to Kampala ANC sentinel surveillance.
View the document(introductory text...)
View the document4.1 Basic principles of HIV dynamics: the relationship of HIV incidence, prevalence and mortality illustrated in 3 simulations for Kampala, Uganda.
View the document4.2 HIV prevalence in ANC clinic and by age: Kampala baseline and incidence reduction simulations.
View the document4.3 HIV mortality and demographic impacts: Kampala baseline and incidence reduction simulations.
View the document4.4 HIV incidence modelling simulations for Kampala women and validation to Kampala ANC sentinel surveillance trends.

(introductory text...)

The model was developed using an epidemiological model previously developed to test hypotheses of antecedent HIV incidence patterns consistent with surveillance trends in AIDS and HIV prevalence by sex, age, birth cohort, and year in Uganda. The model simulates three processes, namely: (a) HIV incidence; (b) demographics of mortality, fertility, and ageing; and (c) the natural history of HIV from infection to death. The model retains some aspects of the earlier developed WHO Epimodel, including empirical inputs, while defining an underlying age and sex profile of risk of HIV infection, and epidemic processes of infection saturation and renewal of susceptibles.

In brief, the model describes an underlying population at risk of HIV infection that is age and period dependent. HIV incidence is generated using an epidemic curve. In heterosexual epidemics, for example, this curve is applied to an age dependent risk that is characterised by a distribution of risk of STD in the general population. The epidemic curve has similar inputs to the Epimodel, and combines point prevalence, the estimated year of extensive HIV spread, and the position on the curve relative to the period of epidemic growth. In contrast to Epimodel, the epidemic curve of HIV incidence is extended over an age distribution and can be modified to fit a gamma, logistic, linear, or other function. Furthermore, it is demographically dynamic, and HIV risk and infection rates are applied to cohorts from a real population. Thus, the population size of the cohorts to which HIV infection rates are applied changes as new cohorts enter ages of risk of infection and older cohorts experience AIDS and natural mortality.

The model produces outputs of HIV incidence and prevalence, AIDS incidence and prevalence, and AIDS deaths, population size by age, sex, and cohort. In addition, and because of its capacity to generate age-specific incidence, a fertility-weighted female HIV prevalence picture can also be produced to approximate HIV prevalence rates in antenatal clinic attendees. These features are important for the empirical validation of incidence scenarios to ANC sentinel surveillance data.

The model can be run for up to six population groups with a global template to aggregate results. The alpha version of the model runs in Windows and is written in Visual Basics. The data output is produced in an Access database format. The model's primary applications are for forecasting scenarios and for the evaluation of determinants of changing HIV prevalence trends, particularly with respect to effectiveness of interventions. A more complete description of the model and its applications has been published or is in press: Human Immunodeficiency virus infection dynamics in East Africa deduced from surveillance data. Am. J. Epidemiol. 1996;144:682-95; An age-and sex-structured HIV epidemiological model: features and applications. Bull. of WHO; Empirical evidence for severe but localised impact of AIDS on population structure. Nature Medicine, May 97.