Concluding Notes

HIV/AIDS has caused a global health shock of extraordinarily large proportions, and the global response has been very effective in reversing the epidemic and mitigating its consequences. The worst country-level health shocks recorded since 1950 are dominated by HIV/AIDS. Of twenty-two episodes on record featuring a drop in life expectancy exceeding 4 years, twelve are primarily caused by HIV/AIDS; if the persistence of the health shock is also taken into account, eight of the ten worst reversals in life expectancy observed on the country level are associated with HIV/AIDS (Table 2.2).

These numbers of course already reflect the impact of the global response to HIV/AIDS, and especially the scaling up of treatment, which has reversed the health impacts of the epidemic. Without it, the bottom of the global distribution of health outcomes would feature a cluster of countries with life expectancy at about 50 years or below, dominated by countries facing a severe HIV/ AIDS epidemic (Figure 6.4). In contrast, people living with HIV in South Africa can now expect a near-normal life span, provided that they initiate treatment relatively early, and in Botswana, where life expectancy overall was reduced by almost 20 years because of HIV/AIDS, the loss in life expectancy is now down to 5 years (Figure 2.8).

The global response has in part been motivated by concerns about the economic repercussions, based on the observations that HIV/AIDS primarily affected young (productive) adults and that the epidemic could reverse economic progress in countries where the level of economic development was already low. These concerns, however, have not materialized.

On the microeconomic level, the impacts of HIV/AIDS on affected individuals and families are well documented. 'People have undoubtedly suffered terrible personal loss and distress, but those who have survived have got by' (Seeley et al., 2010), high national HIV prevalence has not been associated with an increase in poverty rates (Figure 3.4), and the evidence for any impact of HIV/AIDS on GDP per capita is very weak. There is, however, a broad consensus that HIV/AIDS has slowed down GDP growth overall, as population growth has declined, and that—conversely—the HIV/AIDS response contributes to higher growth because a larger population can sustain higher output. Such economic gains, however, are of a different kind from the financial costs of the HIV/AIDS response—most of the GDP gains owing to reduced HIV/AIDS-related mortality are absorbed by the costs of living of people surviving longer because of the HIV/AIDS response, and only a small proportion of it is available to offset the economic costs of the HIV/AIDS response.

In summary, the epidemic has caused a devastating shock to the lives of people affected by HIV/AIDS and, in many countries, to aggregate national health outcomes, but the implications for national poverty rates or economic growth do not fundamentally add to the evaluation of the consequences of HIV/AIDS across countries.

The larger part of the book deals with the effectiveness and cost- effectiveness of HIV/AIDS interventions and policies. Cost-effectiveness was considered as somewhat disreputable in global HIV/AIDS policies, at least at the highest levels of policy formulation, where the impetus came from human rights and medical imperatives.¹ This changed around 2011 with the launch of the UNAIDS investment framework, which framed global HIV/AIDS strategies in terms of 'investment... that will yield long-term dividends', efficiency gains, and cost-effectiveness (Schwartlander et al., 2011).

Cost-effectiveness analysis for HIV/AIDS programmes and interventions, though, presents unusual challenges, resulting from the very slow disease progression and transmission dynamics:

• • The direct impacts of HIV prevention interventions are very different from the population-level impacts, and the latter evolve over time. The effects of an HIV prevention intervention can be much diminished by subsequent infection risk, but the effect of an HIV infection averted can also be magnified by subsequent 'downstream' infections averted (see, e.g., Table 9.3).
• • Formal optimization approaches are frustrated by the excessive interdependencies over time—it is generally not possible to establish optimal spending allocations one period at a time, and there is no obvious measure to value the state of HIV/AIDS and the HIV/AIDS response at the end of the policy period.
• • Over the last decade, the health consequences of HIV infections have diminished (because people living with HIV are likely to progress to treatment, and remain on it for decades), and the financial effects have become (at least relatively) more important and more pervasive (Figures 8.2 and 9.8).

The analyses reviewed and developed in this book respond to these challenges by consistently interpreting investments in the HIV/AIDS response as financial as well as health investments. Policy commitments under the HIV/ AIDS response are interpreted as financial commitments, and new HIV infections give rise to financial liabilities as the resulting costs of HIV/AIDS services absorb fiscal space. The costs of an HIV/AIDS policy over a specific period are interpreted as spending over that period, plus the financial liability posed by the needs of people living with HIV at the end of that period. This interpretation links programme effectiveness and the contributions of a policy towards the financial sustainability of the HIV/AIDS response, and yields more precise estimates of the cost-effectiveness of HIV/AIDS policies than the most common approaches to programme evaluation.

With regard to specific HIV interventions, the book develops two aspects of a forward-looking analysis of cost-effectiveness. First, it illustrates the use of modelling to translate estimates of the direct effects of HIV prevention interventions into estimates of total effects (including, through HIV transmissions averted, beyond the groups targeted by an intervention). Examples include the discussion of the consequences of HIV infections for different sexual risk behaviour (Tables 9.2 and 9.3), and the analysis of the population-level effects of male circumcision (Table 9.4).

Second, the book applies the analysis of the financial liabilities caused by new HIV infections to HIV prevention interventions. Fully accounting for these financial returns, and counting them against the financial costs of an intervention, accentuates differences in cost-effectiveness between HIV prevention interventions (Figure 10.2, Table 10.3), and some HIV interventions come out as good financial investments, even before taking into account any positive health outcomes (see Table 9.1 on condoms, Table 9.4 on male circumcision). This approach also provides a powerful tool for programme evaluation, by testing the marginal contributions of specific interventions to the programme, in terms of costs, health outcomes, and the fiscal space absorbed by the programme, providing pointers on how to improve the programme by reallocating resources or additionally invest in interventions with high cost- effectiveness or a potential to contain the fiscal space absorbed by the programme and contribute to financial sustainability.