I. Introduction and overview.

A. INTRODUCTION

1. The Sahel (1) is perhaps the last frontier in Sub-Saharan Africa (SSA) in the spread of the AIDS epidemic. Although available data are inadequate, they suggest that the overall prevalence of the human immunodeficiency virus (HIV, the causal agent of AIDS) is still very low in the Sahel. The devastating economic and social effects of this disease, which could negate achievements in all sectors, can still be averted if countries act vigorously to prevent HIV transmission. The relatively late arrival of HIV in the Sahel presents governments with an opportunity not enjoyed in other parts of Africa, where the caseload of HIV-induced illness and AIDS was already high by the time the presence of the virus was recognized. Governments in the region are in a position to draw from the body of knowledge and experience already accumulated about the epidemic both to prepare, at the earliest opportunity, to meet its consequences, and to take urgent action to control its spread. In the absence of a vaccine against HIV and affordable, widely applicable therapies for AIDS, at present the only means of preventing HIV transmission is to encourage people to adopt self-protective measures. Even if effective vaccines and treatment regimes become available in the foreseeable future, however, prevention would still be the first and foremost priority.

2. The objective of this report is to define strategies and actions to assist Sahelian countries in taking advantage of this window of opportunity and carry the fight against the epidemic to a higher level. The report discusses the main issues and manifestations of the epidemic, identifies constraints and impediments to effective policy and program development and implementation, and proposes an agenda for action. It also spells out the key elements of donor collaboration and identifies areas for Bank intervention. The report draws on lessons learned in other regions, particularly the experience gained in those regions where the disease is now endemic. It is intended to enhance Bank staffs awareness and understanding of the dynamics of the epidemic in the Sahel and to assist them engage in effective dialogue with government agencies, industry, NGOs, and community organizations.

B. EPIDEMIOLOGY OF THE DISEASE

General

3. The epidemiology of HIV infections is becoming more and more complex as it spreads without recognizing any geographical, political, social, economic or religious boundaries. In SSA the pandemic is characterized by its variability and unpredictability, with each population group having its own dynamics and patterns of transmission. As elsewhere, but even more predominant in SSA heterosexual transmission is by far the most important mode of acquiring HIV, accounting for 80% to 90% of all infections. In addition, numerous studies indicate the predominant role of sexually transmitted diseases (STDs) as biological cofactors in the transmission and virulence of HIV. This association is strongest with genital ulcer diseases (GUDs) such as syphilis, chancroid, and genital herpes, the first two being highly prevalent in SSA. This relationship has been demonstrated in both directions; first GUDs foster HIV transmission (median risk estimate of approximately 10 times), and second, HIV causes more severe clinical manifestations of GUDs. Other studies have shown that gonococcal and chlamydial infections increase the risk of HIV transmission by 3-5 times. This is a major concern considering the very high prevalence of these infections among pregnant women in Africa, where gonococcal infections are 10-15 times higher, chlamydial infections 2-3 times higher, and syphilis more than 10 times higher, compared with similar groups in developed countries (Wasserheit and Holmes, 1992).

4. The progression of HIV infection to clinical onset of symptoms of AIDS appears to be different between African and industrialized countries, with African HIV-infected patients progressing more quickly to AIDS than patients in industrialized nations. Although more research is still needed to understand the specific reasons for this difference, it appears that differences in clinical management of opportunistic pathologies and individual health-seeking behavior may play a role in the spread of HIV infection. For example, over the last ten years, a dramatic increase of some pathologies such as tuberculosis has been noted in Africa, exacerbating the already high morbidity and mortality levels in the region. It has also been documented that poor nutritional status has an adverse effect on the clinical manifestation of HIV-related diseases.

Nature of the data

5. Estimates of HIV-infected persons and AIDS cases in the Sahel rely almost entirely on two sources: (a) various small-scale HIV prevalence surveys conducted beginning in 1985-86; and (b) data gathered from sentinel surveillance systems among specific population groups. With the exception of Senegal, and to a lesser extent The Gambia and Burkina Faso, most countries do not have reliable data on the levels and trends of HIV infection and even less so with regard to STDs. Moreover, AIDS case reporting is known to be highly incomplete, varying from 10% to 30% of the actual level in any country. This lack of reliable data on STDs, HIV, and AIDS is principally due to the weak and inadequate health information systems existing in these countries, and the absence of adequate epidemiological surveillance systems and poor laboratory capacity (paras 41, 43). In addition, there has been considerable reluctance to report AIDS cases, particularly in the early years of the epidemic (para. 19). Although data collection and analysis have improved markedly in recent years, many biases still remain and estimates should be interpreted cautiously, particularly for more "conventional" STDs. A proven rule of thumb, however, is that in areas where HIV infection rates are still low, the prevalence rates of other STDs, as well as the incidence of tuberculosis, are good indicators of the potential spread of HIV infection.

Levels, trends, and patterns

6. The Sahel can generally be characterized as having relatively low levels of HIV seroprevalence when compared with other sub-regions in SSA. However, current estimates based on surveys and reported cases, suggest a rapidly increasing spread of the disease. And while HIV-2 infections are still predominant in countries such as Senegal, The Gambia and Guinea Bissau, HIV-1 infections are the most prevalent in the rest of the Sahel and are increasing at a much faster rate in all countries (the former is known to have a longer latency period and a lower pathogenicity than the latter). Thus, morbidity and consequent mortality rates are likely to rise rapidly over the next few years. Furthermore, although the age pattern of the spread of the disease suggests higher levels of infection among populations over 30 years of age, (2) there is evidence of a rapidly declining median age at infection, particularly among women. There is no conclusive evidence regarding the gender differentials in the rate of infection in the Sahel, however, estimates from elsewhere in SSA suggest a narrowing gender gap (almost a 1:1 ratio).

7. Levels. For the whole of the Sahel, as of December 1992, there were about three quarters of a million individuals (782,500) estimated to be already infected with HIV. Burkina Faso is estimated to have over 50% (425,000) of those infected, followed by Mali and Chad respectively (Figure l).3 The cumulative number of declared AIDS cases for the region as of December 1992, was only 6,047.4 This figure, however, grossly underestimates the actual number of AIDS cases by a factor of about 3-5. Estimates based on the level of HIV seroprevalence suggest that there were about 4 times more AIDS cases (21,508) than were actually declared, of which 50% (11,682) was accounted for by Burkina Faso. Mali, Chad, Niger, and Senegal, all had over 1,000 cases of AIDS during this period (Figure 2).

8. Compared with other sub-regions in SSA, however, levels of HIV seroprevalence are much lower in the Sahel. Current estimates of HIV seroprevalence for the general urban population in the Sahel are less than 5%, compared to over 10% in coastal West Africa, and between 15% and 20% in certain central and eastern African cities. Estimates of HIV infection among pregnant women attending maternal and child health (MCH) clinics (considered representative of the general population) are much lower, currently in the neighborhood of 1% to 3%, although in some "pockets" in a few countries (e.g., Sikasso in Mali, Gaoua in Burkina Faso and Moundou in Chad) prevalence rates have already exceeded 5%. Much higher levels of HIV seroprevalence (over 20%) have been estimated among selected groups engaged in high-risk sexual behavior (such as commercial sex workers and STD patients); these levels are comparable to those in other parts of SSA (Table 1).

9. Trends. Since the first few hundred cases of AIDS were reported in the Sahel in the mid-1980s, the level of HIV infection and number of AIDS cases have been increasing and spreading rapidly among all groups. Short-term projections (1992-97) (5) so far attempted for the Sahel indicate that by 1997 there will be almost 2 million individuals infected with HIV, almost tripling the estimated number of infections in 1992 (782,500). HIV seroprevalence among the general population, while still modest when compared to other subregions (para. 8), is expected to increase from about 2% to 4% during this period. Among the countries, the most striking case is Burkina Faso, where the estimated level of HIV seroprevalence is projected to increase from its current level of 7% to 10% by 1997, with over 1 million individuals infected with the virus (Annex 2).

10. The trend is even more alarming when the current estimates and projected number of AIDS cases are considered. Given the five- to seven-year lag between infection and manifestation of AIDS, there is tremendous momentum for the rapid increase in the number of AIDS cases in the Sahel over the next five years. Based on HIV seroprevalence data, in 1994 there are about 89,000 AIDS cases in the Sahel; a four-fold increase from 1992. This figure is expected to rise to over 300,000 by 1997 (Figure 3). Thus, in only five years (1992-97) the total cumulative number of AIDS cases is expected to increase more than 15-fold. The likely consequences of this rapidly increasing trend in HIV/AIDS are discussed in the next section.

[FIGURE 3 OMITTED]

11. Patterns. The pattern of the spread of HIV infection in the Sahel mirrors that in the rest of Africa, where predominately older cohorts of men and younger cohorts of women are infected, reflecting the early marital age for girls. While the evidence is not conclusive regarding gender differentials in the rate and level of infection, (6) the importance of male seasonal out-migration to the highly endemic coastal countries of Cote d'Ivoire, Ghana, Togo, and Benin suggests relatively higher levels of infection among men than women. There is, however, evidence of a rapidly increasing rate of infection, particularly among younger women. There is also growing evidence of an increasing rate of mother-to-child transmission, particularly of HIV-1, with an estimated transmission rate of 35% to 40%. This rate is about twice that estimated for developed countries for the same strain of virus. Four hypotheses have been postulated to explain this marked variation: (a) more virulent strains in Africa; (b) increased viral load; (c) high prevalence of other STDs; and (d) breastfeeding (although it is still, and will remain, one of the major public health measures to encourage even in HIV hyper-endemic areas).

C. CONSEQUENCES

12. The rapidly increasing trend in HIV infection and the number of AIDS cases do not augur well for the future in the Sahel. The limited analytical work undertaken so far in selected countries (Burkina Faso in particular) demonstrates that the socio-economic consequences of the epidemic would be devastating, exacerbating the already fragile economies in the sub-region. (7)

13. Direct demographic effects. Population projections incorporating the effects of AIDS prepared for selected Sub-Saharan African countries indicate that even in the worst case scenario, the long-term direct demographic effect of AIDS would be a reduction of about 1% or less per annum in the rate of population growth (Bulatao and Bos, 1992). In the Sahel, given the high rate of population growth (about 3% per annum), even without the slowing of HIV infections, the population will continue to grow rapidly--over 2 % annually--and will more than double in size during the next 25 years.

14. Mortality and morbidity effects. The most obvious and direct impact of the epidemic is on the health status of the population. Given the recent nature of the epidemic in the Sahel, the mortality effects of the epidemic are still very modest. About 10,000 deaths were attributable to AIDS in 1992, accounting for less than 2% of the total crude death rate. This figure is expected to more than triple this year to about 34,000 and would increase to 92,000 by 1997, by which time almost 10% of total deaths in the sub-region each year would be attributable to AIDS. When adult mortality is examined, it becomes even more evident that the mortality implications of the AIDS epidemic during the next 3 years would be far more devastating than that of any single illness, in particular due to the resurgence of pulmonary infections such as tuberculosis. By 1997, deaths due to AIDS would account annually for no less than 30% of all adult mortality in the Sahel.

15. The burden of the disease. The rapidly rising trend in long-term morbidity and mortality would have a serious negative impact on the provision and financing of health care, "crowding out" other health services. A recent rapid assessment of the economic impact of HIV conducted in Burkina Faso estimated that nearly half of the beds in some wards at the National Hospital in Ouagadougou are now occupied by HIV/AIDS patients. The situation in Burkina Faso is reaching alarming proportions, very much reflecting the situation in Cote d'Ivoire. Based on the estimated total lifetime average cost of AIDS care per patient of US$416 (8) (derived from the Burkina Faso study), the estimated total direct cost of HIV/AIDS was US$446,000 in 1992. (9) While this amount may appear modest, it already represents approximately 4.1% of the total expenditures (recurrent and investment) of the Ministry of Health in 1992. But this is a very conservative cost estimate as it is based on the total number of reported AIDS cases in 1992, which clearly understates the actual number of AIDS cases by a factor of about five. Thus, the direct lifetime cost of AIDS in Burkina Faso could already exceed 20% of the total health budget. This percentage is likely to increase rapidly during the next few years as the number of AIDS cases rises (Figure 3). In addition, estimates of the indirect cost of AIDS (the value of healthy life years lost from the disease) derived from the study (by calculating the discounted healthy years of life lost) is about US$8 million in 1993, based on 1,073 reported AIDS cases. A more plausible estimate of the indirect cost based on the estimated number of new AIDS cases would be about US$60 million.

16. These figures are substantial and predict the likely scenario in other Sahelian countries, particularly, Mali, Chad, Niger, and The Gambia, where the number of AIDS cases is rapidly increasing. However, this analysis represents only one aspect of the global burden of the disease, namely the impact on the health sector. The epidemic would have severe negative effects on household welfare and productivity, increase costs to firms, and affect major economic sectors, including education and agriculture. One major social consequence of the growing number of AIDS deaths is the rapid increase in the number of orphans. Estimates from the short-term projections cited above, indicate that by the end of 1994, for the Sahel as a whole, there would be almost 70,000 cumulative AIDS deaths. This figure would more than triple to over a quarter of a million (282,000) by 1997, resulting in over 1.7 million orphans. (10) In addition, with a relatively higher percentage of mother-to-child transmission observed in Africa, the number of pediatric AIDS cases would increase significantly over the next three years, creating even greater pressure on already limited public health resources. At the household level there would be tremendous disruption of families as household resources are diverted to care for the growing pediatric AIDS cases, prolonged morbidity among the adult members, and the growing number of AIDS orphans.

17. Without effective intervention to mitigate the spread of HIV/AIDS, the epidemic, as currently observed in many Eastern African communities, would create new poverty groups, exacerbate the already difficult economic situation and pose a considerable challenge in coping with the situation. Moreover, it would alter the distribution of income, thwart efforts to develop sectors that rely on skilled or scarce manpower, and will likely reduce the growth rate of per capita income below levels that would have been feasible without AIDS. For the Sahel, further analytical work is urgently required to identify and measure quantitatively the distribution of the economic burden of the disease, particularly the potential impact on individuals, households, firms and sectors, in order to identify policies to mitigate the impact and implement cost-effective prevention programs in sectors most sensitive to HIV infection.