The past decade has yielded remarkable progress in the expansion of access to antiretroviral therapy (ART), with dramatic expansion of access to treatment for persons living in resource-limited settings.1 There is convincing evidence that high rates of viral suppression can be achieved in programs providing standardized antiretroviral (ARV) regimens with limited laboratory monitoring, and successful therapy is beginning to yield lower mortality related to HIV.2 However, persons with HIV continue to present with advanced immune suppression, and a consistent finding has been high rates of early mortality (8%–20%) after initiation of ART among persons in sub-Saharan Africa compared with persons treated in Europe and the United States.3,4 Tuberculosis (TB), cryptococcal meningitis, and a broad variety of infectious processes, including pneumonia and diarrheal illness, have been identified as the most common causes of early mortality in these settings. A variety of interventions, including isoniazid preventive therapy for TB and cotrimoxazole (CTX) prophylaxis, might help reduce this early mortality.5
Cotrimoxazole is a broad-spectrum antimicrobial agent that, when taken regularly by persons with HIV, has been shown to reduce morbidity and mortality in a variety of settings.6–8 In Europe and North America, CTX prophylaxis is recommended for persons with advanced immune suppression, primarily because it reduces the risk of Pneumocystis jiroveci pneumonia (PCP).9 In contrast, in African settings, CTX prevents a variety of different infections and is effective across a wide range of CD4 cell counts—not just among those with advanced immune suppression.6–8 Initial efficacy studies in Africa were done among persons who were not receiving ART, and until recently, there was very little information available about whether CTX was effective among persons on ART. However, several recent studies have evaluated the association between receipt of CTX and mortality among persons initiating ART.10–12
In 2007, a retrospective cohort study of ART clinics in Malawi evaluated the association between CTX administration and mortality. The study compared outcomes among patients who were treated at clinics routinely providing CTX with outcomes among patients with similar demographic and immunologic characteristics treated at clinics not providing CTX and found markedly lower mortality rates after 6 months of ART at clinics providing CTX (risk reduction = 0.407; P = 0.0013).11 Among patients enrolled in the Development of Antiretroviral Therapy in Africa trial, receipt of CTX was associated with reduced mortality during the first year of ART for mortality overall [odds ratio = 0·65, 95% confidence interval (0·50 to 0·85); for persons on ART for 12–48 weeks: 0.59 (0.37 to 0.95); and for persons on ART for 72–96 weeks: 1.28 (0.54 to 0.03).12 In a cohort of more than 14,000 patients initiating ART in South Africa, receipt of CTX was associated with reduced mortality at 12 months (hazard ratio 0.64, P < 0.001).10 Provision of CTX was not randomized in any of these studies, but the reported reductions in mortality were fairly consistent despite the variations in the study settings and anticipated rates of specific opportunistic infections (OIs) that might be prevented by CTX. In the analysis of participants in the Development of Antiretroviral Therapy in Africa trial, the association with reduced mortality diminished over time; there was no further mortality benefit for CTX after 72 weeks of ART.12
Among persons not taking ART, the reported impact of CTX on rates of specific OIs is highly variable13–18; this may be due to regional differences in disease definitions, availability of diagnostic tests, or rates of resistance to CTX. None of the studies assessing the influence of CTX on mortality among persons initiating ART provided detailed data on OI rates. However, there are compelling data that the benefits of CTX and ART are additive with respect to malaria,17 and a study evaluating discontinuation of CTX in persons taking ART was interrupted because of marked increases in rates of malaria and diarrhea after CTX discontinuation.19
Cotrimoxazole is a relatively inexpensive drug, and cost-effectiveness analyses in persons not receiving ART actually demonstrated cost-savings associated with CTX provision.20,21 Other studies have demonstrated the cost-effectiveness of broadly providing CTX to all HIV-infected adults in a treatment setting in which ART was available for those with advanced disease,22 and in children in a setting in which ART became available after trial completion.23 There have not been previous cost-effectiveness analyses of provision of CTX in the setting of ART initiation. Although data regarding rates of CTX administration are very limited, there seems to have been a marked expansion of the use of CTX since comprehensive guidelines were released by World Health Organization (WHO) in 2006.24 Reports have documented the development of national policies related to administration of CTX prophylaxis and rates of CTX administration in specific cohorts.25
The primary objective of this study was to evaluate the cost-effectiveness of providing CTX to reduce early mortality among persons with advanced HIV who are initiating ART in low-income countries. Because CTX prophylaxis is already provided to some patients in many settings, we evaluated the impact of expanding from assumed current average rates of administration to nearly full coverage, excluding only persons who are allergic to or intolerant of CTX. A secondary objective was to estimate possible additional cost savings from reduced morbidity in terms of OI prevention during the first 6 months of ART.
We developed a decision-analytic model to compare costs and outcomes from 2 scenarios for providing CTX to patients initiating ART (see Figures, Supplemental Digital Content 1,https://links.lww.com/QAI/A258). The base case scenario used estimated average rates of CTX use among adults on ART in low-income countries. The comparator was full CTX coverage. The model was built using TreeAge Pro 2011 (TreeAge Software Inc., Williamstown, MA).
Estimation of Current Coverage Levels
We used data from available publications and reports to inform an estimate of current rates of CTX administration for people initiating ART.26,27 Published rates of CTX administration range from 21% to 94% and seem to be increasing over time.10,12,26–28 We estimated current coverage with CTX at 65%. We assumed even costs and benefits over the range of rates of coverage.
Cost-effectiveness of Deaths Averted Due to Increased Coverage
To determine the potential benefits of increased CTX coverage, we estimated the incremental cost-effectiveness ratio (ICER) as the difference between healthcare utilization costs at base case and full coverage divided by the difference in outcomes. The primary outcome of the analysis was deaths averted during the first 6 months of ART. The time frame and analytic horizon of our analysis were the first 6 months after initiation of ART because mortality risk is concentrated in that period,3,5 and we assumed that deaths occurred at the midpoint.
Mortality rates used in the analysis were all-cause mortality for patients receiving CTX during ART initiation and patients on ART alone.10,11,29 We presented the results for the analysis as the net reduction in deaths realized in the full coverage versus the base case coverage. Causes of OI that are common, would be preventable by CTX, and pose a high risk for mortality were also considered for the analysis.
We adopted a health care provider perspective that included relevant healthcare utilization costs for persons accessing ART. These include routine visits (ie, outpatient care) and drugs (ARVs and CTX) (Table 1). We did not include other out-of-pocket expenditures incurred by individuals, such as transportation and caregiver costs. We estimated the total cost of treatment as the sum of drug cost (ARVs and CTX) and 6-month outpatient care assuming one routine visit per month.33 The cost of ARVs was based on 6 months of a commonly used adult regimen in low-income countries. We also considered the costs related to rates of adverse reactions arising from CTX in an HIV-infected person.20,34 Instead of using cost data from a particular low-income country, we used estimates that could be generalized to a variety of low-income settings in sub-Saharan Africa. Drug prices were derived from the WHO global price reporting mechanism (for ARVs) and the international drug price indicator guide (for CTX).30,31 Data on average cost per outpatient visit were based on WHO-CHOICE estimates for AFR-E countries.32 All costs are reported in 2009 US dollars.
Cost-analysis of OI Cases Averted
We also estimated the OI cases averted and potential cost savings that could be realized by averting OI treatment costs. The OIs considered were malaria, severe bacterial infections, and PCP. Because specific data comparing rates of OIs on ART and CTX with rates on ART alone were unavailable, we used estimated rates of OIs among persons receiving ART alone for the base case and estimated the impact of CTX on rates of specific OIs based on available data.6,13–18 The number of averted OI cases was estimated as the difference between the total OI cases determined in the base case scenario and the total predicted with full CTX coverage.
Costs per OI were also estimated based on a health care provider perspective. These include routine visits (ie, inpatient and outpatient care) and drugs. We estimated the total cost of treatment as the sum of drug cost (CTX and OI drugs) and healthcare utilization cost (ie, outpatient and inpatient care). Prices for OI drugs were derived from the international drug price indicator guide.31
We assumed a range of treatment costs for each OI based on severity of illness (mild to severe), given that the course of treatment can vary considerably. We assumed that the cost of one outpatient visit was incurred regardless of the type of OI. For a mild case of malaria, we added the cost for 3 days of artemether plus lumefantrine (artemisinin-containing regimen), whereas for a severe case of malaria, we added the cost for 1 week of intravenous quinine and 1 week of hospitalization. For a mild case of severe bacterial infection, we included the cost for 7 days of amoxicillin, whereas for a severe case of severe bacterial infection, we included the cost for 10 days of intravenous ceftriaxone and 10 days of hospitalization. In the case of mild PCP, we included the costs for 1 week of intravenous CTX, 2 weeks of oral CTX therapy, and 1 week of hospitalization, whereas in the case of severe PCP, we estimated the treatment costs as 21 days of intravenous CTX and 21 days of hospitalization.
We conducted univariate sensitivity analysis and developed a tornado diagram of the estimated ICER when each model parameter is varied independently. A tornado diagram is a graphical presentation of a 1-way sensitivity analysis that shows the changes in outcome or cost of the optimal intervention as given parameters are varied over their ranges; this allows identification of the most influential parameters in a model.35 Ranges and confidence intervals for mortality rates were derived from published literature. Available data on OI rates were limited and varied widely by setting. To account for this considerable variation in the sensitivity analysis, we set lower bounds for these parameters at one tenth of the base values and set upper bounds equal to estimates for the incidence of OIs observed in people not on CTX if these rates were available, or 10 times the base values for the incidence of OIs on ART without CTX (Table 1).
We also conducted multivariate probabilistic sensitivity analysis on model parameters and cost inputs. For costs, we varied inputs by half and twice their base values to derive upper and lower bounds. We based the probabilistic sensitivity analysis on a replicated sample of 10,000 random draws using Monte Carlo simulation sampling in TreeAge Pro. We used results of the probabilistic sensitivity analysis to develop a cost-effectiveness acceptability curve to illustrate the uncertainty surrounding the a priori decision over a range of willingness to pay. The acceptability curve is interpreted as the probability that the a priori decision is optimal, given the threshold willingness to pay.36
Cost-effectiveness of Deaths Averted Due to Increased Coverage
There were 94 deaths per 1000 patients initiating ART in the base case scenario (Table 2). Moving from the base case to full coverage resulted in 22 deaths averted over 6 months. The cost per patient initiating ART and CTX given base case coverage was $160.36, and the cost per patient considering full coverage was $163.65. The incremental cost of moving from base case coverage to full coverage was approximately $3.29 per patient and $146.91 per death averted over 6 months. Because we assumed even costs and benefits over the range of coverage rates, the model yields similar incremental cost-effectiveness estimates regardless of the chosen baseline rate.
The ICER ranges from the univariate sensitivity analysis are presented in a tornado diagram (Fig. 1). The univariate sensitivity analysis results were consistent with the initial model results. Holding other variables constant and doubling the 6-month drug cost for CTX increased the ICER per death averted to approximately $200, whereas doubling the cost of ARVs increased the ICER per death averted to approximately $230. Increasing the all-cause mortality for ART with CTX from 7% to 10% increased the ICER per death averted to approximately $195, whereas increasing the all-cause mortality for ART without CTX from 14% to 18% increased the ICER per death averted to approximately $194.
The probabilistic sensitivity analysis estimates are also consistent with the initial model results. The incremental cost based on the Monte Carlo simulation of 10,000 replicated samples was approximately $4 per death averted (ICER $161). The results also showed that full coverage was the optimal strategy over majority of the acceptability curve if given a willingness-to-pay threshold of $0 to $7000 (Fig. 2).
Cost-analysis of OI Cases Averted
In the full coverage scenario, the estimated number of cases of malaria was 254 per 1000 patients initiating ART; thus, we estimated 45 fewer incident cases in the full coverage scenario compared with the base case (Table 3). With 45 fewer cases of malaria, full coverage could potentially save between $69.95 and $203.32 per patient in OI treatment costs. For severe bacterial infections, full coverage reduced incidence to 72 new cases of severe bacterial infections or 22 fewer cases per 1000 patients initiating ART. With 22 fewer cases of bacterial infection, full coverage could potentially save between $68.62 and $126.71 per patient in OI treatment costs. For PCP, full coverage would avert 4 cases of PCP per 1000 patients initiating ART over 6 months and potentially save between $75.69 and $88.41 per patient in OI treatment costs.
We used sensitivity analysis of the cumulative incidence rates to produce lower and upper bound estimates for cases averted for the 3 types of OIs. The results demonstrate that the effect on OI reduction due to increased CTX coverage persisted as we varied the incidence of OIs to reflect the uncertainty around the assumed base values. Using the broad ranges for the OI incidence produced 4 cases averted as the low estimate and 195 cases averted as the high estimate for malaria. For PCP, we derived 1 case averted as the low estimate and 72 cases averted as the high estimate. Finally, for severe bacterial infections, we derived 2 cases averted as the low estimate and 275 cases averted as the high estimate.
The cost-effectiveness of CTX in the context of reducing early mortality for patients initiating ART had not previously been evaluated. We evaluated the costs, deaths averted, and incremental cost-effectiveness of expanding provision of CTX to persons with advanced HIV during the first 6 months of ART. Our findings suggest that increased CTX coverage would be a cost-effective approach for reducing early mortality among persons initiating ART.
There are no universally accepted criteria for defining an intervention as cost-effective when using deaths averted as an outcome measure; therefore, our findings are not easily adaptable in the context of willingness to pay. However, our results are comparable with those from studies that have used life years saved/gained as the measure of outcome to demonstrate the cost-effectiveness of CTX for patients with HIV.21,22 Among the studies on adult HIV patients, CTX was cost-effective with ICERs ranging from $150 to $1180 per life year gained when compared with no CTX. Other cost-effective interventions to improve outcomes among patients with advanced HIV in sub-Saharan Africa include ART and TB screening and treatment.37–42 These studies demonstrate cost-effectiveness at varying thresholds. Antiretroviral therapy interventions fall in the higher range of costs with ICERs over $1000 per life year saved/gained.22,33,43–46
Reducing mortality associated with ART initiation is a critical goal because it allows for full realization of the impact of the substantial resources that are consumed during ART initiation, for example, in staging and adherence counseling. We focused our evaluation on the impact of expanding CTX coverage on mortality rather than calculating a comprehensive measure, such as disability-adjusted life years, because there are now several available studies that evaluated mortality data specific to the setting of ART initiation, whereas comparative data about OI rates in persons receiving or not receiving CTX in this setting are very limited. Because serious adverse reactions to CTX are uncommon,34 additional benefits related to averting specific OIs should strengthen the acceptability of broad CTX administration; if data on the rates of OIs in persons taking ART with and without CTX become available, it may be possible to model the impact of the intervention more fully. Our results should not be confused with results measuring the cost per disability-adjusted life year gained.
Given that a health care provider perspective was used to determine health care utilization costs, our ICER per death averted may underestimate the additional savings to the health system by increasing CTX coverage to patients who need it. This may be relevant because patients who die early incur substantially higher productivity losses compared with those who survive the first 6 months of ART.45 Our findings are limited to the effect of CTX prophylaxis during the first 6 months of ART. We chose the 6-month horizon because mortality rates after ART initiation are highest during the initial months of treatment and because at least one study of CTX prophylaxis during ART initiation suggests that the effectiveness of CTX may diminish over time.12 Thus, our findings should not be applied to continued provision of CTX to persons who have been maintained on ART for more than 6 months.
We used generic cost information because goods such as ARVs and OI drugs are often procured on the international market, meaning the costs would be fairly consistent across settings. Nontraded goods, such as outpatient and inpatient costs, are more difficult to generalize to a specific setting, and we have chosen region-specific costs reflecting a variety of low-income settings in sub-Saharan Africa.32 Our findings are limited to settings for which these costs are appropriate.
Other limitations to our analysis should be considered. The lack of randomization of the intervention strategies is a key methodological concern for the 3 studies that informed our assumptions about the mortality impact of CTX. For ethical reasons, a randomized trial of CTX prophylaxis among people initiating ART is unlikely to be performed, and we were reassured by the absence of apparent systematic allocation biases and by the fairly consistent reported reductions in mortality. The studies were conducted in varying settings in sub-Saharan Africa, but it may not be appropriate to generalize the findings to other settings, where rates of OIs may differ.
We assumed that the per-patient costs of prophylaxis were constant, regardless of coverage level. The per-patient costs associated with reaching high coverage levels may differ (eg, if patients fail to take CTX because of adherence issues); however, very high levels of coverage have been achieved in several settings and described as feasible,26,28 and the added costs/requirements for ensuring high coverage with CTX are likely to remain low in settings where retention in care and adherence to ART have been achieved.
Finally, we did not specifically address a challenge that is encountered in many HIV treatment settings, which is that commodities made available specifically for treatment of persons with HIV may be used for other conditions. This results in regular stock-outs for CTX, which could increase the costs or ICER from the perspective of an HIV-specific funding initiative. We found that the ICER was favorable in analyses assuming double the base case drug costs, meaning that a substantial diversion of the CTX supply could be tolerated and its use for ART patients would remain cost-effective.
Achieving high CTX coverage during ART initiation will require further advocacy with both providers and service recipients and also continued attention to maintaining an adequate CTX supply. If it is achieved, expansion of access to CTX prophylaxis during ART initiation could have an important impact on the effectiveness of ongoing HIV initiatives.
1. WHO. Towards universal access: scaling up priority HIV/AIDS interventions in the health sector. Progress Report. 2010.
2. 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.
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. Lawn SD, Harries AD, Anglaret X, et al.. Early mortality among adults accessing antiretroviral treatment programmes in sub-Saharan Africa. AIDS. 2008;22:1897–1908.
5. Lawn SD, Harries AD, Wood R. Strategies to reduce early morbidity and mortality in adults receiving antiretroviral therapy in resource-limited settings. Curr Opin HIV AIDS. 2010;5:18–26.
6. Anglaret X, Chene G, Attia A, et al.. Early chemoprophylaxis with trimethoprim-sulphamethoxazole for HIV-1-infected adults in Abidjan, Cote d'Ivoire: a randomised trial. Cotrimo-CI Study Group. Lancet. 1999;353:1463–1468.
7. Mermin J, Lule J, Ekwaru JP, et al.. Effect of co-trimoxazole prophylaxis on morbidity, mortality, CD4-cell count, and viral load in HIV infection in rural Uganda. Lancet. 2004;364:1428–1434.
8. Wiktor SZ, Sassan-Morokro M, Grant AD, et al.. Efficacy of trimethoprim-sulphamethoxazole prophylaxis to decrease morbidity and mortality in HIV-1-infected patients with tuberculosis in Abidjan, Cote d'Ivoire: a randomised controlled trial. Lancet. 1999;353:1469–1475.
9. Kaplan JE, Benson C, Holmes KH, et al.. Guidelines for prevention and treatment of opportunistic infections in HIV-infected adults and adolescents: recommendations from CDC, the National Institutes of Health, and the HIV Medicine Association of the Infectious Diseases Society of America. MMWR Recomm Rep. 2009;58:1–207; quiz CE201–204.
10. Hoffmann CJ, Fielding KL, Charalambous S, et al.. Reducing mortality with cotrimoxazole preventive therapy at initiation of antiretroviral therapy in South Africa. AIDS. 2010;24:1709–1716.
11. Lowrance D, Makombe S, Harries A, et al.. Lower early mortality rates among patients receiving antiretroviral treatment at clinics offering cotrimoxazole prophylaxis in Malawi. J Acquir Immune Defic Syndr. 2007;46:56–61.
12. Walker AS, Ford D, Gilks CF, et al.. Daily co-trimoxazole prophylaxis in severely immunosuppressed HIV-infected adults in Africa started on combination antiretroviral therapy: an observational analysis of the DART cohort. Lancet. 2010;375:1278–1286.
13. Anglaret X, Messou E, Ouassa T, et al.. Pattern of bacterial diseases in a cohort of HIV-1 infected adults receiving cotrimoxazole prophylaxis in Abidjan, Cote d'Ivoire. AIDS. 2003;17:575–584.
14. Badri M, Ehrlich R, Wood R, et al.. Initiating co-trimoxazole prophylaxis in HIV-infected patients in Africa: an evaluation of the provisional WHO/UNAIDS recommendations. AIDS. 2001;15:1143–1148.
15. Fischl MA, Dickinson GM, La Voie L. Safety and efficacy of sulfamethoxazole and trimethoprim chemoprophylaxis for Pneumocystis carinii
pneumonia in AIDS. JAMA. 1988;259:1185–1189.
16. Kaplan JE, Hanson D, Dworkin MS, et al.. Epidemiology of human immunodeficiency virus-associated opportunistic infections in the United States in the era of highly active antiretroviral therapy. Clin Infect Dis. 2000;30(suppl 1):S5–S14.
17. Mermin J, Ekwaru JP, Liechty CA, et al.. Effect of co-trimoxazole prophylaxis, antiretroviral therapy, and insecticide-treated bednets on the frequency of malaria in HIV-1-infected adults in Uganda: a prospective cohort study. Lancet. 2006;367:1256–1261.
18. van Oosterhout JJ, Laufer MK, Perez MA, et al.. Pneumocystis pneumonia in HIV-positive adults, Malawi. Emerg Infect Dis. 2007;13:325–328.
19. Campbell J, Tappero J, Degerman R, et al. HIV-infected Ugandans on HAART with CD4 cell counts over 200 cells/mm3
who discontinue cotrimoxazole have increased risk of malaria and diarrhea. Paper presented at: HIV Implementers' Meeting 2008; Kampala, Uganda. Abstract B7_3085.
20. Pitter C, Kahn JG, Marseille E, et al.. Cost-effectiveness of cotrimoxazole prophylaxis among persons with HIV in Uganda. J Acquir Immune Defic Syndr. 2007;44:336–343.
21. Yazdanpanah Y, Losina E, Anglaret X, et al.. Clinical impact and cost-effectiveness of co-trimoxazole prophylaxis in patients with HIV/AIDS in Cote d'Ivoire: a trial-based analysis. AIDS. 2005;19:1299–1308.
22. Goldie SJ, Yazdanpanah Y, Losina E, et al.. Cost-effectiveness of HIV treatment in resource-poor settings—the case of Cote d'Ivoire. N Engl J Med. 2006;355:1141–1153.
23. Ryan M, Griffin S, Chitah B, et al.. The cost-effectiveness of cotrimoxazole prophylaxis in HIV-infected children in Zambia. AIDS. 2008;22:749–757.
24. WHO. Co-trimoxazole Prophylaxis for HIV-Related Infections Among Children, Adolescents and Adults. Recommendations for a Public Health Approach. Geneva, Switzerland: WHO; 2006.
25. Date AA, Vitoria M, Granich R, et al.. Implementation of co-trimoxazole prophylaxis and isoniazid preventive therapy for people living with HIV. Bull World Health Organ. 2010;88:253–259.
26. HEALTHQUAL. The HIVQUAL International Update. Special Issue: Cotrimoxazole preventive therapy. 2009; Supplemental (Spring).
27. Auld AF, Mbofana F, Shiraishi RW, et al.. Four-year treatment outcomes of adult patients enrolled in Mozambique's rapidly expanding antiretroviral therapy program. PLoS One. 2011;6:e18453.
28. Watera C, Todd J, Muwonge R, et al.. Feasibility and effectiveness of cotrimoxazole prophylaxis for HIV-1-infected adults attending an HIV/AIDS clinic in Uganda. J Acquir Immune Defic Syndr. 2006;42:373–378.
29. Alemu AW, Sebastian MS. Determinants of survival in adult HIV patients on antiretroviral therapy in Oromiyaa, Ethiopia. Glob Health Action. 2010;3.
30. WHO. Global Price Reporting Mechanism: Transaction Prices for Antiretroviral Medicines and HIV Diagnostics from 2008 to March 2010.
31. International Drug Price Indicator Guide. Geneva, Switzerland: WHO; 2009. Available at: http://erc.msh.org
32. WHO. CHOosing Interventions that are Cost Effective (WHO-CHOICE), 2005: Results of Unit Costs for patient services for the 14 GBD Regions. Geneva, Switzerland: WHO; 2005. Available at: http://www.who.int/choice/costs/unit_regions/en/index.html
. Accessed March 29, 2011.
33. Rosen S, Long L, Fox M, et al.. Cost and cost-effectiveness of switching from stavudine to tenofovir in first-line antiretroviral regimens in South Africa. J Acquir Immune Defic Syndr. 2008;48:334–344.
34. Grimwade K, Swingler G. Cotrimoxazole prophylaxis for opportunistic infections in adults with HIV (Review). Cochrane Database Syst Rev. 2003 CD003508.
35. Haddix AC, Teutsch SM, Corso PS. Prevention Effectiveness: A Guide to Decision Analysis and Economic Evaluation. New York, NY: Oxford University Press; 2003.
36. Fenwick E, Claxton K, Sculpher M. Representing uncertainty: the role of cost-effectiveness acceptability curves. Health Econ. 2001;10:779–787.
37. Andrews JR, Lawn SD, Rusu C, et al. Cost-effectiveness of Xpert MTB/RIF for screening for tuberculosis prior to antiretroviral therapy in South Africa. Paper presented at: 6th IAS Conference on HIV Pathogenesis Treatment and Prevention; 2011; Rome, Italy. Abstract MOAD0102.
38. Bachmann MO. Effectiveness and cost effectiveness of early and late prevention of HIV/AIDS progression with antiretrovirals or antibiotics in Southern African adults. AIDS Care. 2006;18:109–120.
39. Creese A, Floyd K, Alban A, et al.. Cost-effectiveness of HIV/AIDS interventions in Africa: a systematic review of the evidence. Lancet. 2002;359:1635–1643.
40. Currie CS, Floyd K, Williams BG, et al.. Cost, affordability and cost-effectiveness of strategies to control tuberculosis in countries with high HIV prevalence. BMC Public Health. 2005;5:130.
41. Loubiere S, Meiners C, Sloan C, et al.. Economic evaluation of ART in resource-limited countries. Curr Opin HIV AIDS. 2010;5:225–231.
42. Shrestha RK, Mugisha B, Bunnell R, et al.. Cost-utility of tuberculosis prevention among HIV-infected adults in Kampala, Uganda. Int J Tuberc Lung Dis. 2007;11:747–754.
43. Badri M, Cleary S, Maartens G, et al.. When to initiate highly active antiretroviral therapy in sub-Saharan Africa? A South African cost-effectiveness study. Antivir Ther. 2006;11:63–72.
44. Bendavid E, Wood R, Katzenstein DA, et al.. Expanding antiretroviral options in resource-limited settings—a cost-effectiveness analysis. J Acquir Immune Defic Syndr. 2009;52:106–113.
45. Cleary SM, McIntyre D, Boulle AM. The cost-effectiveness of antiretroviral treatment in Khayelitsha, South Africa—a primary data analysis. Cost Eff Resour Alloc. 2006;4:20.
46. Walensky RP, Wolf LL, Wood R, et al.. When to start antiretroviral therapy in resource-limited settings. Ann Intern Med. 2009;151:157–166.