AN UPDATE ON THE AIDS EPIDEMIC IN BOTSWANA
HIV prevalence The 2004 UNAIDS Global Report states that the adult prevalence rate in Botswana was 37.3 percent in 2003. This figure is based on anonymous, random testing of pregnant women at 21 antenatal clinics throughout the country. There are now two new sources of information on HIV prevalence levels in the Botswana. The first is from the newly established national network of voluntary counselling and testing centres. The adult (15-49) prevalence rate for females was 37.4 percent and 23.9 percent for males in 2003, which gives an overall adult prevalence rate of 31.0 percent. The second source is a national population-based HIV survey that was undertaken in mid-2004 by the Central Statistical Office. Unfortunately, adverse publicity in the early stages of the survey resulted in around one-third of those sampled refusing to be tested. However, it was mainly in urban areas where refusal rates were especially high. According to this survey, the national HIV prevalence rate is 25.3 percent, which is around one-third lower than the UNAIDS estimate. Other recent national population-based surveys in Africa have also produced national HIV prevalence estimates that are much lower than those based on antenatal clinic testing (see Table 1).
Table 1: Population-based and ante-natal clinic HIV estimates Population Ante-natal Country based Clinic Botswana 25.3 37.3 Cameroon 5.5 11.8 Kenya 6.7 13 Mali 1.7 1.9 Niger 0.9 1.2 Sierra Leone 0.9 7 South Africa 16 25 Notes: Estimates are for same years
ANC estimates for 2001-2003 show that there has no been statistically significant increase or decrease in the national HIV prevalence rate. The prevalence rate for 15-19 year olds declined from a high of 32.4 percent in 1995 to 21.0 percent in 2002. Prevalence rates for the 20-24 year group peaked at 42.8 percent in 1998, but the rate is still very high (at 37.4 percent in 2002). The 2004 population-based HIV survey estimates that the prevalence rate for 15-19 year olds is 6.6 percent 1 . Gender-disaggregated data have not yet been made available, but the testing results from the VCT centres indicate that the 15-19 prevalence rates are likely to be around 10
1
Recent population-based testing in Swaziland also shows that the HIV prevalence rate for 15-19 year old women was 6.0 percent in 2004 compared to the ANC-based prevalence rate for this age category of 32.5 percent (in 2002). Nearly one-third of the pregnant women sampled at ANC clinics were in the 15-49 age category.
1
percent for females and 1 percent for males. From a prevention perspective, the key message for young people is that, even at this late stage in the epidemic, the very large majority of teenagers remain infection free. It is their behaviour once they become young adults in their 20s that determine whether they avoid becoming infected. Morbidity and mortality Information published in the 2003 national HIV/AIDS Surveillance Report shows, that despite much higher reported levels of HIV prevalence among females, the overall gender difference in the number of reported new AIDS cases between 2000 and 2002 is minimal (51.4 percent female and 48.6 percent male). The 2001 Population and Housing Survey also shows that female mortality rates are lower than male mortality rates for all adults except the 20-24 and 25-29 age cohorts (see Table 2), which is surprising, given much higher HIV prevalence rates among females. In Botswana, there is a strong negative relationship between AIDS-related mortality levels and socio-economic status. In particular, mortality rates are much higher in rural than urban areas (see Table 2). Since HIV prevalence rates are similar in rural and urban locations, it would appear that better access to medical treatment and better nutrition are resulting in lower levels of mortality in urban areas.
Table 2: Mortality rates for females and males by location and age, 2001 RURAL AGE Female Male 15 to 19 0.24 0.18 20 to 24 1.24 0.65 25 to 29 2.52 1.64 30 to 34 2.73 3.12 35 to 39 2.33 3.50 40 to 44 1.89 3.10 45 to 49 1.54 3.09 50 to 54 1.37 2.69 55 to 59 1.38 2.56 Source: Population Census URBAN Female 0.17 0.63 1.24 1.48 1.49 1.37 1.31 1.42 1.53
Male 0.18 0.39 0.94 1.77 1.96 2.25 2.50 2.24 2.75
According to the 2003 Surveillance Report, there was a 51 percent increase in the number of people who tested HIV positive between 2000 and 2002 (from 6001 to 9081). However, the number of new AIDS cases that were reported fell by 37 percent (from 1141 to 717) during the same period, which is completely the opposite of what one would expect. Clearly, not all individuals with AIDS are being picked up and reported, but the public health care system in Botswana is well resourced with comprehensive coverage. The mass provision of anti-retroviral drugs is likely to have a significant impact on morbidity and mortality rates. By late 2004, around 35,000 people were accessing anti-retroviral drugs free of charge as part of the government’s
2
national ARV programme 2 . It is planned to increase the total number of beneficiaries to 65,000 by mid 2006. Around 110,000 individuals are estimated to be in need of these drugs. Public servants have been accessing ARVs since the late 1990s as part of the government’s medical aid scheme. The number of teachers who were taking ARVs as part of this scheme increased from 62 in 1998 to over 474 in Aril 2002. As a result, annual mortality rates among teachers remained in the range 0.6-0.8 percent during this period, which is two-four times lower than total projected mortality rates (AIDS and non-AIDS related). Around 10 percent of the staff of a major parastatal organisation are currently taking ARVs and the mortality rate (from all causes) was only 0.6 percent in 2004. Orphans The 2001 Population Census collected fairly detailed information on the parental status of children and the school attendance of orphans and nonorphans. In that year, the incidence of orphans was as follows: lost both parents 2.0 percent, maternal orphans 3.6 percent, and paternal orphans 9.6 percent. These percentages are much less than the 2001 estimated numbers and incidence of these three groups of orphans published in the high profile Children on the Brink publication (see Table 3) 3 . COB projections are generally relied upon in most analyses of the ‘orphan crisis’ in Africa.
Table 3: Comparison of orphan population figures from Children on the Brink and Population Census, 2001 MATERNAL Number % 69000 10.6 27000 3.6 PATERNAL Number 91000 71000 DOUBLE Number 62000 15000
Source COB Census
% 14 9.6
% 9.5 2
Notes: Census figures cover children 0-17 years old while Children on the Brink estimates are for 0-14 year olds only. COB estimates 0-14 population as 650,000.
There were no statistically significant differences in the school attendance rates of orphans and non-orphans in 2001 (see Table 4). As reported by Bennell et al 4 , the main reasons for this are the significant levels of support provided by the government to ‘needy’ orphans, the provision of school meals (which have improved noticeably during the last two years), and the strong schooling culture in Botswana with orphans wanting to be like everybody else.
2
Another 9600 individuals have been assessed as requiring ARVs, but were still awaiting treatment in late 2004. There is serious over-inflation of orphan estimates by COB in numerous other African countries. See Paul Bennell, 2005, The impact of the AIDS epidemic on the schooling of orphans and other directly children in sub-Saharan Africa. Journal of Development,Studies, vol. 41, no. 3, April P.S. Bennell,, B.Chilisa, K. Hyde, A. Makgothi, A. Molobe and L. Mpotokwane, 2002, The impact of HIV/AIDS on primary and secondary schooling in Botswana. DFID Education Research Series no. 47, DFID, London.
3
4
3
Table 4: Schooling attandance by parental status, 2001 AGE 5 TO 9 never attended 29.1 30.4 25.1 33.4 33.0 FEMALE in school 69.5 68.3 73.0 65.6 65.9 left School 1.4 1.3 1.9 1.0 1.1 never attended 30.8 38.7 30.5 36.2 36.1 MALE in school 67.8 59.6 67.6 62.6 62.7 MALE in school 90.0 90.2 89.1 91.2 91.0 left school 1.5 1.6 1.8 1.1 1.2
Mother dead Father dead Both dead Both alive All children AGE 10-14
FEMALE never in left attended school School Mother dead 2.6 94.1 3.3 Father dead 2.5 93.7 3.8 Both dead 2.8 93.5 3.8 Both alive 2.5 94.4 3.1 All children 2.5 94.2 3.2 Source: Population and Housing Census 2001
never attended 3.9 3.7 4.3 3.8 3.8
left school 6.1 6.1 6.6 5.0 5.2
Paul Bennell, Senior Partner, Knowledge and Skills for Development, Brighton. January 2005
4