Copyright © NISC Pty Ltd
African Journal of AIDS Research 2003, 2(1): 1–8 Printed in South Africa — All rights reserved
AJAR ISSN 1608–5906
Epidemiological and demographic HIV/AIDS projections: South Africa Thomas M Rehle1 and Olive Shisana2* Independent Consultant in International Health and Disease Control, 1426 G Street SE, Washington, DC 20003, USA HSRC/SAHA, Private Bag X9182, Cape Town 8000, South Africa * Corresponding author, e-mail:
[email protected]
1 2
The Epidemic Projection Package (EPP) recently developed by the UNAIDS Reference Group on Estimates, Models and Projections and the Spectrum model program developed by the Futures Group were used to model the South African HIV epidemic, project future trends in HIV/AIDS and estimate the demographic impact of AIDS. The national HIV prevalence surveys among pregnant women from 1990–2001 and the first national, population-based HIV survey in 2002 served as the data sets used to calibrate the input HIV prevalence values for the model. The scenario created by the model showed that a dramatic rise in HIV prevalence during the 1990s has peaked in 2002 with 4.69 million infected people and it is projected that the epidemic in South Africa has now begun to level off. Adult (15–49 years) incidence rates have decreased substantially in the past five years since 1997 (4.2%) and are expected to reach a level of 1.7% in 2003. The annual number of deaths due to AIDS is projected to peak with 487 320 AIDS deaths in the year 2008. By 2020, the total population of South Africa is expected to be 23% smaller than it would be without AIDS, however, a negative population growth rate is not expected during the projection period. Life expectancy at birth is expected to hit a low of 45.6 years in the time period 2005–2010, which is 22 years less than it would have been in the absence of AIDS. Ten years from now over 2.5 million AIDS orphans are projected for South Africa. Models play an important role in estimating HIV variables that are difficult to measure. Projections of the future HIV/AIDS burden in South Africa underscore the importance of acting now to reduce the number of new infections and plan for medical and social care needs. Keywords: AIDS mortality, demographic impact, HIV prevalence, modelling
Introduction South Africa is recorded to have the largest number of persons living with HIV/AIDS in the world. By the end of 2001, this figure was estimated by UNAIDS/WHO to be five million people (UNAIDS/WHO, 2002). The estimate is derived from antenatal data that the Department of Health collects annually through sentinel sites located in each of the nine provinces (Department of Health, 2002). The data are collected from pregnant women who attend antenatal clinics in the public health sector. The Department of Health has collected the data since 1990; they provide a basis for tracking the epidemic. The age-specific results of the antenatal survey are shown in Figure 1. The estimates show that the HIV/AIDS epidemic has been increasing rapidly since 1991, particularly among those aged 20–24 years and 25–29 years. It is encouraging that among the youth the HIV prevalence reached a peak in 1998 and since then has continued to decline. Antenatal clinic prevalence data continue to be the most readily available source of data for modelling the HIV epidemic in many countries. However, antenatal clinic sentinel data are subject to selection biases related to convenience sampling (sites may not always be randomly chosen), usage and coverage of antenatal clinic services, differentials in risk behaviours and contraceptive use, lower fertility rates
among women with HIV-1 infection (Gray, Wawer, Serwadda, Swankambo, Li & Wabwire-Mangen, 1998) and other socio-demographic factors (e.g. age distribution of those attending antenatal clinics, level of education, socioeconomic status, migration patterns). Moreover, there are insufficient data on the relative importance of these factors in different settings, and even less is known about how these factors may vary over time. Comparisons between antenatal clinic sentinel surveillance and general population serosurveys have indicated that data from pregnant women may differ significantly from the general female population data, and the relationship can go in different directions at different stages of the epidemic, and for different age groups (Kigadye, Klokke, Nicoll, Nyamuryekung’e, Borgdorff & Barongo, 1993; Fontanet, Messele, Dejene, Enquselaise, Abebe & Cutts, 1998; Fylkesnes, Ndhlovu, Kasumba, Musonda & Sichone, 1998). The considerable variation in the findings suggests that extrapolations from antenatal clinic data should be made with caution. Population-based studies carried out periodically can help to evaluate the sources of bias in different country settings. This approach allows the necessary ‘calibrations’ of results obtained from pregnant women. In 2002, a national population-based HIV survey was
2
Rehle and Shisana
