The expansion of combination antiretroviral therapy (ART) in Africa has dramatically changed HIV/AIDS from an illness with almost certain short-term death to a chronic disease.1 Early antiretroviral therapy guidelines for resource-limited settings recommended eligibility for treatment when a patient's CD4+ T-cell count reached 200 cells per microliter or below.2 The majority of patients initiated therapy with much lower CD4+ counts.3,4 In 2010, the World Health Organization (WHO) issued guidance to resource-constrained settings to expand the eligibility of the treated population by recommending the initiation of ART earlier, when a patient's CD4+ T-cell count reached 350 cells per microliter or less or was clinically necessary.5 These recommendations were in line with recommendations adopted by developed countries several years earlier.6
CD4+ cell count is well understood to be an important prognostic factor in ART patients, with lower CD4+ counts strongly associated with increased risk of mortality.7,8,9,10 However, expanding eligibility to a larger number of patients is associated with short-term feasibility challenges and costs. For these reasons, donors were initially reluctant to support the shift to earlier initiation, and many countries maintained eligibility at ≤200 cells per microliter or demonstrated a progressive improvement to ≤250 cells per microliter.11 Most, but not all, countries have now applied the earlier initiation threshold, but the relative cost-benefit analysis of earlier initiation is still debated,12 and in many places, the policy is initiated passively, with little attempt to encourage people to present to care earlier.13
In addition to improved survival after early initiation of ART, other benefits of early initiation are becoming apparent. Early initiation of ART (1) improves life expectancy, (2) decreases the likelihood of developing serious coinfections, including tuberculosis, and (3) decreases the need for hospitalization.7,9,14 There is convincing evidence that ART also reduces transmission of HIV to sexual partners.15,16 Yet increasing the number of patients on therapy to improve their disease status represents a fundamental challenge to optimal public health, due primarily to financial or structural constraints.17 Multilateral funding for HIV/AIDS is decreasing, and many countries are reticent to provide new slots for patients.18,19,20 The aim of this article is to determine if increasing the eligibility threshold for the initiation of ART from 200 to 350 cells per microliter can be cost-effective, as defined by WHO-CHOICE, in a developing country.
Our study used data collected by The AIDS Support Organization (TASO). The TASO cohort has been described elsewhere.21 Briefly, TASO was founded in 1987 and has since provided care to more than 200,000 patients in Uganda, that is, providing counseling, free access to ART, and regular health care for the treatment of opportunistic infections (OIs), while actively attempting to retain patients and avoid loss to follow-up. It began free distribution of antiretrovirals in 2004 and currently provides services to 24,000 patients at 11 sites throughout Uganda. We have previously reported on clinical outcomes within this cohort, including CD4+ cell counts, mortality data, and life expectancy aggregated by sex and age.21,22,23,24 The TASO cohort is one of the oldest and largest of its kind in Africa and is generalizable to many resource-limited settings because it captures data nationwide across a variety of patient experiences and provides low-cost HIV care based on care provided by a mix of physicians, clinical officers, community workers, and expert patients. In addition, laboratory services are minimal and offer CD4 evaluations nonroutinely, and ART is delivered in a decentralized manner.21
The cohort for this analysis includes 22,315 patients initiated since 2004, older than 14 years, and with a median baseline CD4+ cell count at presentation of 142 cells per microliter (interquartile range, 80–230 cells per microliter). Overall, 69.4% were female, and the median age was 37 years, and the median follow-up period was 31 months. More than 85% of cohort members maintained at least 95% adherence to their prescribed ART regimen. Patients initiating antiretroviral therapy typically received a non–nucleoside reverse transcriptase inhibitor with first-line treatment with fixed-dose combinations comprising nevirapine or efavirenz plus lamivudine and stavudine; second-line therapy consisted of boosted lopinavir, didanosine, and zidovudine.
Our study was conducted from the perspective of the state ministry of health. We included the expected cost of death and excluded indirect costs, including transportation to and from health centers and costs of foregone productivity (lost wages due to illness or death, etc)25 from the primary analysis. These costs are examined in the sensitivity analysis. Time horizons selected are of use to a decision maker because they incorporate the normal time horizon for planning (5 years) and include the likely longevity of expenses and benefits associated with the comparators, over the expected life span of a patient in this setting (30 years).14
The majority of ART recipients in the southwestern Rakai district in a recent report had monthly incomes of <US$6 and spent between 10%–50% of their income on travel to and from ART clinics.26 The same source article reported a mean waiting time at the clinic of 8.36 hours, discouraging participation in ART programs. A truly societal perspective would account for these lost hours and expenses, which we have included in the sensitivity analysis.
Our main outcome measures are years of life saved (YLS), budget impact of expanded ART provision, and the cost-effectiveness of the new initiation threshold. We define the cost-effectiveness of <350 cells per microliter as the incremental cost per YLS. All costs and YLS were discounted 3%, with various discount rates tested in sensitivity analysis, as per accepted practice.27 In the absence of data with which to conduct a cost-utility analysis, this cost-effectiveness model was deemed appropriate.
In 2010, 39% of currently eligible Ugandan persons living with HIV/AIDS were receiving ART (201,400 patients), compared with 53% at the end of 2009 under the previous guidelines, as the total number of eligible patients has increased from 380,000 to 520,000.28 We assume that coverage will reach the same level as the previous regimen meaning that the total number of patients receiving ART will be 275,600 (520,000 × 0.53) increased from 201,400. Our primary results reflect the budget impact of such an increase in coverage.
Survival data from the TASO cohort formed the basis of the analysis, using survival curves stratified by initial CD4+ cell count.22 We used the survival curves from the 100 to 199 per microliter range and 200 to 350 cells per microliter. The survival curves estimate mortality for the first 12 years for patients initiating under the <200 cells per microliter threshold to be 4.782%, increasing to 6.387% for the subsequent 12 years, and then 7.04% for the remaining years in the model. For patients initiating at the <350 cells per microliter threshold, initial mortality is estimated at 4.017% for 12 years, then 4.972% for the subsequent 12 years, and 5.729% for the remaining years in the model. The incremental cost-effectiveness ratio (ICER) results will be conservative because the <200 cells per microliter does not include the very low CD4+ cell counts (<100 cells). All mortality rates were converted to transition probabilities for use in a deterministic Markov model, using the standard transformation.29 A Markov model was selected due to the number of variables considered and the time horizon employed.
One of the main expected clinical differences between CD4+ strata is the incidence of OIs. The most common OI of clinical concern is tuberculosis, and we include the expected costs per incident in our analysis. The number of incident tuberculosis cases for patients initiating at <200 cells per microliter is 20.5 per 1000 patients annually. Based on the published data from Haiti, earlier ART initiation illustrated a protective effect, reducing incident cases of tuberculosis by 50% (with a range tested in sensitivity analysis),7 which was thus applied to the earlier-initiating group. The rate was converted to a transition probability for use in the model.
We also assumed a preventive role of ART, as treatment of index patients dramatically reduces the likelihood that they transmit to another patient.16 The benefits of ART in preventing new infections are assumed to increase when a patient is initiated at an early stage because they are more likely to be healthy enough to continue active sexual behaviors. Following methods used at TASO previously, we included the cost savings associated with HIV infections averted, estimating 86.3 infections averted per 1000 patients in the first 2 years, and limiting the preventive benefit from ART to that period. Our estimate of the annual savings from infections averted was calculated as the annual cost of ART × the number of infections averted in year 2, with the conservative assumption that that number would not increase, and with each group's respective mortality rate applied to the number of remaining infections averted each year. Univariate sensitivity analysis was used to assess the impact of this factor on the incremental cost per year-of-life-saved (YLS).
Recent data has estimated the cost of ART provision in PEPFAR programs in Uganda at $843, expressed in 2009 USD, and this represents our base-case assumption. Costs per episode of tuberculosis ($46.02) are based on hospitalization costs from a microcosting study conducted at TASO and are inflated to 2009 US$ for our primary analysis, as is the cost of death, $614.30 The most recent UNAIDS data cite a cost of US$791 per patient on ART in Uganda in 2005, and this figure is considered in sensitivity analysis.31
WHO defines an intervention as “highly cost-effective” if the cost per disability adjusted life year (DALY) is less than per capita gross domestic product (GDP), “cost-effective” if the cost is less than 3 times GDP per capita, and “not cost-effective” if the cost is more than 3 times GDP per capita.32 In the absence of relevant utility data, we have restricted our analysis to YLS. Uganda's purchasing-parity-adjusted per capita GDP for 2009 is estimated at $1200 in the CIA Factbook. The WHO benchmark must also consider the context of the intervention in relation to other public health programs being funded from a finite supply of resources: only when other, more cost-effective programs have been fully implemented should the expansion of a given program be funded. We used a deterministic analysis in Microsoft Excel for all cost-effectiveness analyses (version 2011. 14.1.2).
In the primary analysis, the goal of 53% coverage of the eligible population of 520,000 under earlier initiation will cost an additional $261,651,942 (budget impact) and yield an additional 329,794 extra YLS for the 5-year analysis. Over 30 years, earlier initiation will cost an additional $872,685,561, yielding an additional 1,113,689 YLS. The ICER for earlier initiation is $695/YLS for 5 years, and it is $769 at 30 years. Both ICER values fall within the WHO-CHOICE definition of highly cost-effective, because they represent less than per-capita GDP.
ART Program costs varied around the 2009 PEPFAR estimate ($843) by ±25%, from $632 to $1054. The 5-year ICER at a program cost of $632 was $491/YLS and $574 at 30 years. At a program cost of $1054, the 5-year ICER was $899, and the 30-year ICER was $964 (see Table 2).
Cost of treatment for tuberculosis varied from 50% to 150%, separately from the overall program costs. At the lower values, the 5-year ICER was $705/YLS, and at 30-years, it was $771. At the higher values, the 5-year ICER was $685/YLS and at $767 30-years.
The base-case assumption for incidence of tuberculosis was 20.5 per 1000 patients annually. Protection against tuberculosis attributed to ART was assumed to be 50%.7 Varying the protection from 25% to 75%, the ICER ranged from $684 to $705 at 5 years and $767 to $772 at 30 years. Excluding the cost of death resulted in a 5-year ICER of $794 and $775 at 30 years.
We also considered the impact of a broader perspective on the results. Based on a mean unskilled annual wage rate of $2507,33 we estimated the mean hourly wage to be $1.00. We further estimated the travel and miscellaneous costs of obtaining ART to be $0.65 per visit. With quarterly visits, this adds $36.04 (($8.36 + 0.65) × 4) to the program's cost. This increased the ICER for 5 years to $730 and $802 for 30-years.
In accordance with the US Panel's guidance, the discount rate was analyzed at 0%, 5%, and 7% with a base-case value of 3%. At 0%, earlier initiation cost was $700/YLS at 5 years, and at 30 years, the ICER was $711. At 5%, the 5-year ICER was $692/YLS, and it was $767 at 30 years. At 7%, the 5-year ICER for earlier initiation was $688/YLS, and it was $764 at 30 years (see Tables 2, 3).
Our results indicate that expansion of ART in Uganda is cost-effective by the benchmark definition proposed by WHO-CHOICE. Our baseline ICERs represent conservative estimates, as we expect that the earlier initiation threshold will generate more averted infections.16 By increasing the rate of averted infections by the new guidelines from 86.3 to 172.6, the ICER decreases to $668 in the 5-year analysis and $710 at 30 years. It is possible that the prophylactic effect of ART will be the source of substantial societal benefit, and evidence of this effect is growing. Additionally, we have limited the prevention effect to only the first 2 years of analysis, although its benefit may be persistent. It should be noted that tuberculosis does not represent the only cause of disease, and the protective effect of ART could be extended to other common OIs, such as bacterial or fungal infections or malignancies. Our findings are consistent with recent studies demonstrating consistently lower costs, both inpatient and outpatient, associated with ART initiation at higher CD4+ cell counts.34,35 In addition, the greater tolerability associated with the use of tenofovir in first-line therapy—the other major recommendation made by the 2010 guidelines—would result in fewer patients switching lines of therapy, and likely lower costs, especially if the price of tenofovir decreases.36
A limitation of the study is that the survival curves from our cohort project an enduring benefit from ART that has yet to be demonstrated in lengthy follow-up.14 The data are based on 10 years of reporting and are the basis for 30-year survival curves, and so greater uncertainty is attached to all reported 30-year figures. In addition, considering survival as the only clinical end point, while straightforward and comprehensible, limits the potential benefits under consideration. Our study would be enhanced by a full cost-utility analysis, if detailed information on costing and quality of life was available, and a model of the natural progression of the disease in the TASO cohort was used. We would expect earlier initiation to remain cost-effective if such data were analyzed because substantial quality-of-life gains could be made in the avoidance of OIs. Critically, in cohort studies, earlier initiation has demonstrated greater impact on progression to AIDS than on mortality.8 The inclusion of productivity costs such as absenteeism from work, the value of volunteer caregivers' time, lost economic productivity due to morbidity and mortality, and lost leisure time would provide a truly societal perspective but fall outside the scope of our analysis.
It is conceivable that the scaling up of ART programs and increased output may reduce marginal cost through better negotiated pricing for drugs and more efficient use of existing facilities.37 However, there is a very real risk that the existing shortage of trained staff will prove to be the bottleneck in the system, limiting the potential savings made due to economies of scale. Sub-Saharan Africa currently has a shortage of trained health personnel, estimated at 5 million workers.38 A recent analysis of PEPFAR-sponsored HIV treatment centers across several countries identified falling costs associated ART recipients on therapy for longer than 6 months.39 We have not assumed falling costs, as further study is required to determine the mechanism by which such results are realized.
Beyond the cost implications of program expansion in Uganda, critical social and systemic factors impede its implementation. Weaknesses in the health system itself are manifested in stock-outs of drugs for patients on treatment regimens and increased waiting time for newly accepted patients. Other systemic problems include poor patient confidentiality and poor preantiretroviral care, that is, monitoring of high CD4+ patients until initiation criteria are met, discouraging subsequent program participation. Stigma associated with HIV/AIDS may lead to the delay in HIV testing. Most important, without an active program aimed at identifying eligible patients, patients will simply continue to present late in their disease and the benefits will be limited.
The current international economic crisis has already had a large impact on the perception and funding of AIDS programming through both the US Global Health Initiative and the Global Fund to Fight AIDS, TB, and Malaria, resulting in a shortage and flatlining of new funds.20 Therefore, whether there will be additional funds to provide for new patients is a matter of current debate. New patients on treatment will result in more drug expenses. However, less sick patients will result in less complex care and less patients admitted to inpatient or advanced services. Organizations, such as TASO, have overcome many of the financial challenges by simplifying care and decentralizing services.40 ART care that was once provided by physicians at health centers can now be provided by lower cadre health workers at community distribution points. There is no current evidence to suggest that this importantly compromises care.40
There is now debate about the merits of CD4 thresholds as an indication for the initiation of ART as early starting of ART,41 for example, at the time of diagnosis and readiness will result in decreased long-term immune damage, reduce the likelihood of immune reconstitution inflammatory syndrome, and decrease the potential for transmitting the virus to others. However, the debate is polarized by the focus on saving more lives (among those already infected) or preventing new infections. Targeting universal access (a target of greater than 80%) to care assumes a rationing of care while targeting immediate seek and treat assumes that rationing is less important. This debate is not yet resolved within the international funding community.
Our results indicate that the implementation of the new CD4+ cell count threshold for cART initiation in Uganda is cost-effective, according to the definition suggested by WHO-CHOICE. It should be noted that such thresholds are moving targets, in that they reflect both the current state of clinical evidence and the fiscal realities faced by governments. The earlier initiation of ART in developed countries represents evidence to this effect. In the Ugandan context, increasing access to ART must compete for funding with other public health initiatives already in use, such as childhood immunization, cotrimoxazole prophylaxis, and adult male circumcision programs, which have demonstrated significant cost-effectiveness. ART represents an opportunity to reduce both new infections and mortality in specific patients, and the adoption of the new guidelines expands this benefit to more Ugandans.
1. Ford N, Calmy A, Mills EJ. The first decade of antiretroviral therapy in Africa. Global Health. 2011;7:33.
3. Braitstein P, Brinkhof MW, Dabis F, et al.. Mortality of HIV-1-infected patients in the first year of antiretroviral therapy: comparison between low-income and high-income countries. Lancet. 2006;367:817–824.
4. Boulle A, Van Cutsem G, Hilderbrand K, et al.. Seven-year experience of a primary care antiretroviral treatment programme in Khayelitsha, South Africa. AIDS. 2010;24:563–572.
6. Hammer SM, Eron JJ Jr, Reiss P, et al.. Antiretroviral treatment of adult HIV infection: 2008 recommendations of the International AIDS Society-USA panel. JAMA. 2008;300:555–570.
7. Severe P, Juste MA, Ambroise A, et al.. Early versus standard antiretroviral therapy for HIV-infected adults in Haiti. N Engl J Med. 2010;363:257–265.
8. Sterne JA, May M, Costagliola D, et al.. Timing of initiation of antiretroviral therapy in AIDS-free HIV-1-infected patients: a collaborative analysis of 18 HIV cohort studies. Lancet. 2009;373:1352–1363.
9. Ford N, Kranzer K, Hilderbrand K, et al.. Early initiation of antiretroviral therapy and associated reduction in mortality, morbidity and defaulting in a nurse-managed, community cohort in Lesotho. AIDS. 2010;24:2645–2650.
10. Kitahata MM, Gange SJ, Abraham AG, et al.. Effect of early versus deferred antiretroviral therapy for HIV on survival. N Engl J Med. 2009;360:1815–1826.
12. Hontelez JA, de Vlas SJ, Tanser F, et al.. The impact of the new WHO antiretroviral treatment guidelines on HIV epidemic dynamics and cost in South Africa. PLoS One. 2011;6:e21919.
13. Mills EJ, Ford N. Home-based HIV Counseling and testing as a gateway to earlier initiation of antiretroviral therapy. Clin Infect Dis. 2012;54:282–284.
14. Mills EJ, Bakanda C, Birungi J, et al.. Life expectancy of persons receiving combination antiretroviral therapy in low-income countries: a cohort analysis from Uganda. Ann Intern Med. 2011;155:209–216.
15. Eshleman SH, Hudelson SE, Redd AD, et al.. Analysis of genetic linkage of HIV from couples enrolled in the HIV Prevention Trials Network 052 trial. J Infect Dis. 2011;204:1918–1926.
16. Cohen MS, Chen YQ, McCauley M, et al.. Prevention of HIV-1 infection with early antiretroviral therapy. N Engl J Med. 2011;365:493–505.
17. Shelton JD. HIV/AIDS. ARVs as HIV prevention: a tough road to wide impact. Science. 2011;334:1645–1646.
18. Moszynski P. Global Fund suspends new projects until 2014 because of lack of funding. BMJ. 2011;343:d7755.
19. Leach-Kemon K, Chou DP, Schneider MT, et al.. The global financial crisis has led to a slowdown in growth of funding to improve health in many developing countries. Health Affairs. 2012;31:228–235.
20. Geng EH, Bwana MB, Kabakyenga J, et al.. Diminishing availability of publicly funded slots for antiretroviral initiation among HIV-infected ART-eligible patients in Uganda. PloS One. 2010;5:e14098.
21. Bakanda C, Birungi J, Nkoyooyo A, et al.. Cohort profile: the TASO-CAN Cohort Collaboration. Int J Epidemiol. 2012;41:946–950.
22. Mills EJ, Bakanda C, Birungi J, et al.. Mortality by baseline CD4 cell count among HIV patients initiating antiretroviral therapy: evidence from a large cohort in Uganda. AIDS. 2011;25:851–855.
23. Bakanda C, Birungi J, Mwesigwa R, et al.. Density of healthcare providers and patient outcomes: evidence from a nationally representative multi-site HIV treatment program in Uganda. PloS One. 2011;6:e16279.
24. Bakanda C, Birungi J, Mwesigwa R, et al.. Association of aging and survival in a large HIV-infected cohort on antiretroviral therapy. AIDS. 2011;25:701–705.
25. Sculpher M. The role and estimation of productivity costs in economic evaluation. In: Drummond M, McGuire A, eds. Economic Evaluation in Health Care. Oxford, United Kingdom: Oxford University Press; 2001.
26. Kunihara NR, Nuwaha F, Mayanja R, et al.. Barriers to the use of antiretroviral drugs in Rakai district of Uganda. Afr Health Sci. 2010;10:120–129.
27. Gold MR, Siegel JE, Russell LB, et al., eds. Cost-Effectiveness in Health and Medicine. New York, NY: Oxford University Press; 1996.
29. Fleurence RL, Hollenbeak CS. Rates and probabilities in economic modelling: transformation, translation and appropriate application. Pharmacoeconomics. 2007;25:3–6.
30. Marseille E, Kahn JG, Pitter C, et al.. The cost effectiveness of home-based provision of antiretroviral therapy in rural Uganda. Appl Health Econ Health policy. 2009;7:229–243.
31. Filler SJ, Berruti AA, Menzies N, et al.. Characteristics of HIV care and treatment in PEPFAR-supported sites. J Acquir Immune Defic Syndr. 2011;57:e1–e6.
33. Marseille E, Saba J, Muyingo S, et al.. The costs and benefits of private sector provision of treatment to HIV-infected employees in Kampala, Uganda. AIDS. 2006;20:907–914.
34. Koenig SP, Bang H, Severe P, et al.. Cost-effectiveness of early versus standard antiretroviral therapy in HIV-infected adults in Haiti. PLoS Med. 2011;8:e1001095.
35. Walensky RP, Wolf LL, Wood R, et al.. When to start antiretroviral therapy in resource-limited settings. Ann Intern Med. 2009;151:157–166.
36. Bender MA, Kumarasamy N, Mayer KH, et al.. Cost-effectiveness of tenofovir as first-line antiretroviral therapy in India. Clin Infect Dis. 2010;50:416–425.
37. Marseille E, et al.. HIV Prevention costs and program scale: data from the PANCEA project in five low and middle-income countries. BMC Health Serv Res. 2007;7:108–117.
38. Zachariah R, Ford N, Philips M, et al.. Task shifting in HIV/AIDS: opportunities, challenges and proposed actions for sub-Saharan Africa. Trans R Soc Trop Med Hyg. 2009;103:549–558.
39. Menzies NA, Berrutti AA, Berzon R, et al.. The cost of providing comprehensive HIV treatment in PEPFAR-supported programs. AIDS. 2011;25:1753–1760.
40. Jaffar S, Amuron B, Foster S, et al.. Rates of virological failure in patients treated in a home-based versus a facility-based HIV-care model in Jinja, southeast Uganda: a cluster-randomized equivalence trial. Lancet. 2009;374:2080–2089.
41. Cohen MS, Dye C, Fraser C, et al.. HIV treatment as prevention: debate and commentary-will early infection compromise treatment-as-prevention strategies? PLoS Med. 2012;9:e1001232.