The past decade has seen unprecedented increases in access to and delivery of HIV treatment and care. Affordable and effective first-line antiretroviral regimens are now widely available, and an estimated 3 million people have started antiretroviral therapy (ART) in resource-limited settings.1 For those who need it, second-line ART is becoming increasingly affordable and accessible.1,2
Although access to and delivery of HIV treatment have improved, resource constraints have curbed use of laboratory-based diagnostic tests in many developing countries and led to shifting attitudes toward what should be recommended in terms of patient monitoring. In 2006, the World Health Organization's (WHO's) public health response to HIV led to guidelines emphasizing a tiered patient monitoring structure, with CD4 count at the district level and CD4 count and HIV RNA quantification at the regional level but with neither considered compulsory for patient management.3,4 In 2008, based on results from a study by Phillips et al5 suggesting only modest clinical benefit from CD4 count or HIV RNA monitoring to guide switching to second-line ART, WHO emphasized the use of clinical monitoring.6 Whereas HIV RNA testing has been available in only limited settings, the decreasing costs and simplification of CD4 count technologies have allowed scale-up of immunologic monitoring in many developing countries.6 This led the WHO to reaffirm the importance of CD4 counts and to recommend in 2009 that clinical failure should be confirmed at least by immunological criteria when HIV RNA is not available.7
In keeping with this recent recommendation, most current national guidelines consider CD4 counts along with clinical criteria as the standard of care for monitoring patients receiving ART, although virologic monitoring in these settings is still generally considered optional.8-15 In the context of a country like Côte d'Ivoire, a low-income West African country with adult HIV prevalence of approximately 3.9%,16,17 our objective was to examine the clinical benefits, costs, and cost-effectiveness (CE) of CD4 count and/or HIV RNA monitoring in guiding switching to second-line ART.
We utilized a previously published simulation model of the natural history and treatment of chronic HIV disease.18-20 Clinical and cost data were derived from clinical trials and cohort studies conducted in Côte d'Ivoire, as well as publicly available fee schedules and cost databases.2,21-24 Monitoring strategies to guide switching to second-line ART were based on different criteria (clinical, immunologic, or virologic) for detecting first-line antiretroviral failure. The performance of alternative strategies was evaluated using the incremental CE ratio (ICER), expressed as 2006 US dollars per year of life gained (YLS) and defined as the additional cost of a specific strategy, divided by its additional clinical benefit, compared with the next less-expensive strategy.25 We adopted a modified societal perspective (meaning that patient time and transportation costs were not included), with future benefits and costs discounted 3% annually.25-28 Sensitivity analyses were conducted to evaluate the impact of uncertain parameters and assumptions on the results. Additional information on the methods is available in the Technical Appendix (see Supplemental Digital Content 1, http://links.lww.com/QAI/A35) and in previous publications.18-20,29-31
To quantify the benefit from the availability of second-line therapy, we included 2 relevant comparators among the base case strategies: cotrimoxazole prophylaxis only and first-line ART only plus cotrimoxazole prophylaxis. In the base case, we evaluated 3 main monitoring approaches to diagnose first-line ART failure, thereby prompting a switch to second-line ART. These included (1) clinical monitoring, with failure defined as a single WHO stage III-IV event other than tuberculosis (TB) and invasive bacterial diseases; (2) immunologic monitoring, with failure defined as a 50% decrease from peak regimen-specific CD4 count (consistent with WHO recommendations); and (3) virologic monitoring, with failure defined as a 1-log10 increase in HIV RNA and/or return to pretreatment HIV RNA level (Table 1).
In a secondary analysis, we assessed variations of the 3 main monitoring strategies, including (1) alternative clinical criteria for first-line failure (eg, WHO stage III-IV event, including TB and/or invasive bacterial diseases); (2) alternative immunologic criteria for first-line failure (eg, 25% decrease in peak CD4 count); (3) combined clinical and immunologic/virologic monitoring (eg, WHO stage III-IV event or a 1-log10 increase in HIV RNA and/or return to pretreatment HIV RNA level); and (4) a 6-month delay in initiation of second-line ART after observation of virologic failure.
We employed the Cost-Effectiveness of Preventing AIDS Complications (CEPAC) International model, an individual-level Monte Carlo simulation model of HIV disease progression and treatment.18-20,29-31 Drawing from an initial distribution of country-specific demographic (age and sex) and clinical characteristics (CD4 count, HIV RNA level, and history of opportunistic infection), the model draws on monthly transition probabilities to simulate a cohort of individual patients whose clinical course is tracked from model entry until death. The model projects intermediate outcomes (eg, longitudinal CD4 count, number and type of opportunistic infection) and long-term aggregate outcomes (eg, life expectancy and lifetime costs).
Disease Progression in the Absence of ART
HIV disease progression is modeled as a function of HIV RNA level and CD4 count.32 In the model, virologic and immunologic status are represented using 6 HIV RNA strata and 6 CD4 count strata.18,19 Progression in the absence of therapy is based on a patient's true HIV RNA level, which determines the rate of CD4 count decline, and in turn, the risk of specific opportunistic infections and death.32,33 The model tracks this information throughout an individual's lifetime and distinguishes between underlying disease progression and observed measures of clinical, immunologic, and virologic status. Opportunistic infection rates are based on primary data from Côte d'Ivoire and are classified broadly as HIV-related severe events (severe malaria, TB, invasive bacterial diseases, and other WHO stage III-IV events), mild events (noninvasive bacterial diseases and WHO stage II events), and unexplained severe events (acute unexplained fever or acute unexplained diarrhea with hospitalization).18,19 Cotrimoxazole prophylaxis-administered to all patients upon entry into care-results in a reduced risk of bacterial infections, malaria, other WHO stage III-IV events (toxoplasmosis, isosporiasis, pneumocystis, and nocardiosis), and unexplained severe events.19,21,34
Disease Progression in the Presence of ART
Simulated patients can either achieve HIV RNA suppression or not achieve HIV RNA suppression on a particular ART regimen. For patients who achieve HIV RNA suppression, disease progression is modeled based on decreases in HIV RNA level and increases in CD4 count. Virologically suppressed patients face a monthly risk of “late” failure, defined as a 0.5-log10 increase in HIV RNA over at least 2 consecutive months.32 Late virologic failure is followed by a 12-month delay before CD4 count decline.35 Patients who either do not achieve HIV RNA suppression or experience late virologic failure have CD4 declines similar to those not receiving ART. Regardless of an individual's virologic status, patients receiving ART experience an independent reduction in incidence of opportunistic infections and AIDS-related mortality.36,37
Patients experiencing undetected virologic failure and who continue on ART accumulate resistance.38 Rather than model the accumulation of individual resistance mutations, we modeled the effects of incremental resistance accumulation on the efficacy of subsequent ART regimens. This was accomplished by specifying a relative percent decrease in the baseline 24-week virologic suppression of subsequent ART regimens for each month that a patient remains on a virologically failed regimen.39-41 This relative percent decrease is hereafter referred to as the “resistance penalty.”
We specify the resistance penalty in Equation 1, below:
In this equation, “pnetSuccessi” represents the efficacy of current ART regimen i after virologic failure due to accumulated resistance from prior regimens; “pSuccessi” represents the efficacy of ART regimen i after virologic failure in the absence of resistance from prior ART regimens; “respensuccessi” represents the fractional reduction in the initial probability of success of regimen i, per month spent on prior regimens; “cumVLtime” represents the cumulative number of months spent, across all ART regimens, after having virologically failed ART; and i represents the current ART regimen i, in which i is an integer value beginning with 1 and continuing toward the maximum number of sequential lines of ART available.
We assume that the resistance penalty does not affect CD4 response to subsequent regimens, conditional upon virologic response. We also assume that the penalty applies only to the initial 24-week efficacy of a subsequent regimen and not to an individual's probability of virologic failure at later time points. Derivation of the baseline value for the resistance penalty is shown in the Data section.
Patient-Level Monitoring of Clinical, Immunologic, and Virologic Status
Patient-level HIV disease progression and treatment efficacy are monitored through clinical, immunologic (via CD4 counts), and/or virologic (via HIV RNA tests) assessments. Clinical assessments occur upon entry into care, presentation with any acute event, and at 3-month intervals. CD4 and HIV RNA tests, if available, occur upon entry into care and at 6-month intervals thereafter.4 Treatment-related decisions (ie, starting, switching, or stopping ART regimens) are made based on information from clinical assessments and, if available, CD4 counts and/or HIV RNA tests.
Clinical and Cost Data
Cohort Characteristics and Natural History
Data were derived mainly from trials and cohort studies conducted in Côte d'Ivoire by the Programme PAC-CI. Initial distributions of age, sex, and CD4 count were derived from the ACONDA cohort, an observational cohort of HIV-infected adults and a continuation of the ANRS 1203 Cotrame cohort study in Abidjan, Côte d'Ivoire (Table 2).42,43 Incidence of opportunistic infections (a function of CD4 count), HIV-related mortality (a function of both CD4 count and history of opportunistic infection), and efficacy and toxicity of cotrimoxazole prophylaxis were estimated from ANRS 059 trial data, as well as data from the ANRS 1203 and 1220 study cohorts.22,23,48 Risk of non-HIV-related mortality was derived from country-specific life tables for Côte d'Ivoire.49
Effectiveness of nonnucleoside reverse transcriptase inhibitor-based first-line ART was derived from a prospective cohort study of treatment-naive patients in Abidjan.42 At 24 weeks, 80.2% of patients experienced HIV RNA suppression to ≤300 copies per milliliter and a median CD4 count increase of 127 cells per microliter (interquartile range: 64, 201).42 We assumed that, in the absence of resistance, the effectiveness of protease inhibitor (PI)- based second-line ART was similar to that of first-line ART (at 24 weeks, 77.0% suppressed to <400 copies/mL and a mean CD4 count increase of 186 cells/μL).44 Incidence of ART-related severe adverse events led to a switch in drug of similar cost, effectiveness, and drug class.50
To derive the baseline value for the resistance penalty, we used 3 pieces of information: PI-based ART efficacy in the absence of resistance, PI-based ART efficacy in the presence of resistance (ie, thymidine analogue mutations resulting from failure of first-line nucleoside reverse transcriptase inhibitors), and time spent on virologically failed ART. For PI-based ART efficacy in the absence of resistance, 24-week virologic suppression (<400 copies/mL) was 77.0%.44 In the presence of resistance, 24-week virologic suppression (<400 copies/mL) was 73.3% for patients on a second-line PI-based regimen.45 For time spent on virologically failed ART, we estimated that mean time spent on a virologically failed first-line ART regimen was 10.8 months.46 We assumed that this estimate reflected the difference between higher HIV RNA suppression (ie, 77.0%) in the absence of resistance and lower HIV RNA suppression (ie, 73.3%) in the presence of resistance. We substituted these 3 estimates into Equation 1 to solve for the baseline value for the resistance penalty:
These data resulted in an overall estimate of a 0.45% relative decrease in baseline 24-week HIV RNA suppression per month on virologically failed first-line ART. We evaluated a wide range (0%-1.63%) of values for the resistance penalty, reflecting overlap in both the 95% confidence intervals of the ART efficacy data and uncertainty in the variables informing the resistance penalty.
We estimated direct medical costs for HIV-related care (ie, inpatient care, outpatient visits, treatment of acute clinical events, other routine care, medications, and laboratory costs) from the placebo arm of ANRS 059, a trial evaluating cotrimoxazole prophylaxis in Abidjan, as well as from the literature.19,21,24,47 These costs were adjusted to 2006 price levels and converted, when necessary, from local currency to US dollars using official exchange rates.51 Costs of ART came from a publicly available pricing guide for developing countries.2 Reflecting first-line ART in the ACONDA cohort,42 we calculated a cost of first-line ART of $121 per year. This was the weighted average of several regimens-52% stavudine, lamivudine, and nevirapine ($100 annually); 22% stavudine, lamivudine, and efavirenz ($120 annually); 20% zidovudine, lamivudine, and efavirenz ($177 annually); and 6% other ($120 annually). We assumed that second-line ART costs included tenofovir/emtricitabine ($199 annually) plus lopinavir/ritonavir ($550 annually), for a total cost of $749 per year. We assumed that all drug costs reflected a WHO-recommended dosing schedule. For each cost estimate, we established plausible upper and lower bounds to evaluate real-world cost differences (ie, due to projected decreases or realistic variation in second-line ART costs2,52) as well as to assess uncertainty.
Base Case Analysis
Undiscounted life expectancy was 2.2 years for cotrimoxazole prophylaxis only and 12.0 years for first-line ART only plus cotrimoxazole prophylaxis, with projected undiscounted life expectancies ranging from 14.9 years for clinical monitoring to 17.5 years for biannual CD4 monitoring to 19.3 years for biannual HIV RNA monitoring to guide switching to second-line ART. Compared with only one line of ART, the incremental benefits from the availability of second-line ART ranged from a 24.3% increase in undiscounted life expectancy with clinical monitoring to a 46.4% increase with CD4-based monitoring to a 61.3% increase with HIV RNA monitoring (Table 3). Given the availability of second-line ART, CD4-based monitoring increased undiscounted life expectancy by 17.6% compared with clinical monitoring; HIV RNA monitoring resulted in a further 10.2% increase in undiscounted life expectancy compared with CD4-based monitoring. Mean CD4 counts at first-line observed failure ranged from 129 to 467 cells per microliter, with earlier detection of failure (as occurred with the HIV RNA monitoring strategy) associated with a higher CD4 count at time of failure detection and switching.
Table 3 also shows the ICERs for each strategy assuming 3 potential costs for HIV RNA monitoring ($87, $50, and $25). Compared with clinical monitoring, CD4-based monitoring (switching to second-line ART when a 50% decrease in peak CD4 count is observed on first-line ART) had an ICER of $2120 per YLS. In comparison, virologic monitoring (with a failure criterion of 1-log10 increase in HIV RNA or return to pretreatment HIV RNA level) had an ICER ranging from $2920 ($87 per HIV RNA test) to $1990 ($25 per HIV RNA test) per YLS. Complete results for the base case analysis are shown in the Technical Appendix (see Supplemental Digital Content 1, http://links.lww.com/QAI/A35).
Although no consensus exists on a universal threshold below which an intervention would be considered “cost-effective,” benchmarks can be useful to compare the relative value provided by different interventions to improve health. For example, the Commission on Macroeconomics and Health has suggested that an ICER of less than 3 times a country's annual per capita gross domestic product represents a cost-effective intervention.53,54 Based on 3 sources for the annual gross domestic product (GDP) per capita at nominal values for Côte d'Ivoire (US $1016 to US $1178), an approximate CE threshold of 3 times this country's GDP per capita would range from approximately US $3000 to US $3500.55-57
Figure 1 shows the results of sensitivity analyses in which parameters were varied to assess their impact on the ICER of virologic monitoring to detect first-line ART failure. The analysis was conducted using baseline HIV RNA test costs of both $50 (Fig. 1A) and $87 (Fig. 1B). In addition to HIV RNA test costs, CE results were most sensitive to the cost of second-line ART and the relative decrease in second-line ART efficacy per month on virologically failed first-line ART, that is, the “resistance penalty.” CE results were less sensitive to chronic care costs, the cost of CD4 count tests, the efficacy of second-line ART, the discount rate, and the probability of late failure.
As the cost of second-line therapy was reduced, the CE of virologic monitoring became more attractive. If the cost of second-line ART approximated that of first-line ART (approximately $121 per year), the ICER of virologic monitoring improved to less than $1000 per YLS (HIV RNA test cost of $50) and to less than $1200 per YLS (HIV RNA test cost of $87). Decreasing the “resistance penalty” influenced both life expectancy and CE for the HIV RNA monitoring strategy. If the baseline resistance penalty was increased from 0.45% (base case) per month to 1.5% per month, as might occur in third-line and subsequent ART regimens, HIV RNA monitoring was both more effective and more cost-effective than all other strategies.
An influential assumption on the CE of virologic monitoring was the duration of second-line ART following detected failure. Consistent with clinical care in many countries, we assumed that second-line ART was continued until death, despite clinical, immunologic, and/or virologic failure;4 the increase in life expectancy associated with this base case assumption was approximately 8 months, compared with stopping second-line ART after virologic failure. Given the disproportionate increase in costs across these 2 extreme assumptions, reduction in the cost of second-line ART and/or stopping second-line ART at some point after virologic failure had a major influence on the CE of HIV RNA monitoring.
Alternative Criteria for Detecting First-Line ART Failure
Figure 2 shows the relationship between lifetime costs and life expectancy for the 3 main base case monitoring strategies and variations of these strategies that alter the criteria for first-line ART failure. Strategies that rely on laboratory monitoring, either CD4 count or HIV RNA, to guide switching always resulted in higher life expectancy than strategies relying on clinical monitoring alone. Expanding the clinical failure criterion to include both TB and invasive bacterial diseases was more costly and less cost-effective than the base case criterion of WHO stage III-IV events, excluding TB and invasive bacterial diseases, or WHO stage III-IV, including TB but not invasive bacterial diseases. Changing the CD4 monitoring criterion for first-line ART failure from a 50% to a 25% decline in peak CD4 also was an efficient strategy; it was more costly and more effective than the base case strategy, although it was less effective than HIV RNA monitoring. None of the strategies that combined clinical and immunologic or virologic monitoring was more effective, less costly, or more cost-effective, than the base case strategies.
We also assessed the impact of a 6-month delay in initiation of second-line ART following observed failure of first-line ART for all HIV RNA monitoring strategies. This strategy generally was not more effective or more cost-effective than virologic monitoring strategies that did not employ a delay. Results for these sensitivity analyses are shown in the Technical Appendix (see Supplemental Digital Content 1, http://links.lww.com/QAI/A35).
In resource-limited settings, the role of CD4 count and HIV RNA monitoring in ART management remains an area of widespread debate, despite being standard clinical practice in most developed countries.5,58-65 We addressed the impact of using CD4 count and/or HIV RNA monitoring to guide switching to second-line ART in Côte d'Ivoire. We found that earlier detection of first-line ART failure (via immunologic or virologic monitoring) resulted in higher CD4 counts upon observed first-line ART failure, shorter duration on virologically failed first-line ART, and earlier switching to second-line ART. Accordingly, while the incremental life expectancy gains associated with providing second-line ART (compared with first-line ART only) was 24.3% with clinical monitoring, laboratory monitoring substantially increased these survival gains: 46.4% with CD4 count monitoring and 61.3% with HIV RNA monitoring.
This analysis suggests that the CE of laboratory monitoring is influenced most by the cost of second-line ART, the impact of resistance on second-line ART efficacy, and the duration of second-line ART following virologic failure. We found that virologic monitoring would be cost-effective, according to the criteria of the Commission on Macroeconomics and Health, in all of the following instances: if the decrease in second-line ART efficacy due to time spent on virologically failed first-line ART is greater than 0.4% per month, if the cost of second-line ART is less than $750 annually, or if the HIV RNA test cost is less than $90 per test. With the exception of situations where virologic monitoring dominates (ie, is more effective and less costly, or more effective and more cost-effective) CD4-based monitoring, for example, at very low HIV RNA test costs of $25, switching to second-line ART based on a 50% decline in CD4 cell counts is consistently cost-effective. At very low HIV RNA test costs, we found that if CD4 count test costs were decreased to less than $15 per test, CD4-based monitoring was no longer dominated. In this case, HIV RNA testing had a CE ratio of $2030 per life year gained compared with CD4 monitoring, and CD4 count monitoring had a CE ratio of $1970 per YLS compared with clinical monitoring.
Due to the high cost of second-line ART, one of the most influential assumptions on the CE of virologic monitoring was the duration of time that second-line ART was continued following failure. Importantly, there is a substantial life expectancy gain associated with continuing ART after virologic failure; however, this clinical benefit is accompanied by an even greater increase in costs. We found that a reduction in second-line ART costs or discontinuing ART after second-line failure, at some point before the end of life, improved the CE of HIV RNA monitoring. It is likely that as HIV-infected individuals in Côte d'Ivoire and other resource-limited settings begin to fail second-line therapy, further downward pressure on drug prices will also provide the opportunity for third-line therapy, including newer drugs. If that is the case, and there is ongoing development of resistance with continuation of second-line therapy after virologic failure of those regimens, then it will also improve long-term outcomes to switch from second- to third-line therapy at the time of virologic failure. This would further support the use of HIV RNA monitoring.
We also assessed several variations of the 3 main monitoring strategies using alternative clinical criteria for first-line failure. Expanding the clinical failure criterion to include TB or invasive bacterial diseases was more costly and less cost-effective than a criterion of WHO stage III-IV events, excluding TB and invasive bacterial diseases (base case), or WHO stage III-IV events, including TB but not invasive bacterial diseases. Changing the CD4 monitoring criterion for first-line ART failure to a 25% decline in peak CD4 count was an efficient strategy. Although less effective than HIV RNA monitoring, it was both more effective (and more costly) than a criterion of a 50% decline in peak CD4 count (base case). None of the strategies that combined clinical and immunologic or virologic monitoring was more effective or less costly, or more cost-effective, than the base case strategies.
These results have important implications for clinical and budgetary planning. Although the results provide information about the value of different laboratory monitoring techniques to detect the timing of first-line antiretroviral failure, they can also inform issues related to affordability and budget planning. For example, for a clinic of 10,000 patients, similar to the CePRef clinic in Abidjan,42 the percent of patients alive at 10 years with CD4-guided switching of ART would be 66% compared with 60% without CD4 monitoring. Total per person costs of care would increase to $6120 from $5040, with the additional costs attributable primarily to earlier switching from first-line to more expensive second-line ART.
Several modeling studies have sought to address the debate on laboratory monitoring to guide HIV treatment management in resource-limited settings.5,18,61,64,66,67 Although results from this analysis correspond with those found in some, but not all, existing studies,61,64 the current analysis differs from the literature in several ways, including variation in model structure across studies and differences in data to inform the models. For example, Bishai et al accounted for resistance but used US natural history data in their model. Bendavid et al did not model resistance and used data from Southern Africa. Both of these studies examined initiation and switching of ART, not just switching as in the current study.
The different findings in the current study compared with that by Phillips et al are primarily due to changes in model input parameters. The Phillips study assumed that all patients had a WHO stage 4 event before starting ART, and the mean CD4 count of the cohort at the time of ART initiation was 66 cells per microliter. In the current study, only 0.5%-20% of the cohort had a prior stage III-IV event, and the mean CD4 count was 140 cells per microliter. Because the cohort had less severe illness at initiation and experienced lower mortality, the potential benefits of earlier switching using CD4 or HIV RNA testing were greater, leading to lower CE ratios for all monitoring tests.
This study has several limitations. First, we assumed that results from CD4 count and HIV RNA tests accurately reflect underlying patient-level disease status and that laboratories used to analyze test measurements yield accurate and consistent results.68-70 Second, we did not assess different CD4 count and HIV RNA test technologies, nor did we consider alternative delivery mechanisms. Third, limited data were available to inform the “resistance penalty.” Although an emerging literature exists on drug- and/or regimen-specific mutation accumulation,39 little information exists on the impact of regimen-specific mutations on subsequent antiretroviral efficacy. Therefore, we believe the method used to characterize the impact of resistance as a function of time on virologically failed ART most accurately reflects the evidence base.40,41 To that end, we used the most current data available to inform the resistance penalty and performed extensive sensitivity analysis on this parameter. Fourth, we did not account for the fact that CD4 count decline may be discordant with virologic failure; therefore, without HIV RNA monitoring some people will be unnecessarily switched to second-line therapy.71,72 Accounting for this discordant response would make HIV RNA monitoring even more favorable. Finally, we did not factor into the analysis the impact of resistance on HIV transmission dynamics. Specifically, we did not account for any population-level benefit of decreased HIV transmission of wild type or resistant virus due to earlier switching from a virologically failed first-line regimen to an effective second-line regimen. Inclusion of these effects likely would result in HIV RNA monitoring appearing even more favorable.
As ART becomes increasingly available for HIV-infected individuals in Côte d'Ivoire and in other resource-limited settings, it is critical to understand both the clinical and economic value of laboratory monitoring for HIV management. This analysis suggests that CD4 count and HIV RNA monitoring to guide switching to second-line ART in resource-limited settings improves survival and, under most conditions, is cost-effective. These results support the value of investing in low-cost HIV RNA tests, reducing prices for second-line ART, and developing a better understanding of the relationships among delayed switching, development of resistance mutations, and subsequent antiretroviral efficacy.
We are grateful to technical assistance provided by Brandon Morris, Lauren Uhler, Caroline Sloan, and Sarah Bancroft Lorenzana at the US study site. We also extend thanks to the Côte d'Ivoire ANRS research site study team (Programme PAC-CI, Abidjan, Côte d'Ivoire), the Association ACONDA study team (Abidjan, Côte d'Ivoire), the CeDReS laboratory team (CHU de Treichville, Abidjan, Côte d'Ivoire) and the INSERM U897 research team (University of Bordeaux 2, France). Special thanks are extended to the patients who have been participating in the ANRS clinical studies in Abidjan since 1996 and who have greatly contributed to increasing knowledge of HIV natural history and treatment efficacy in Côte d'Ivoire.
1. UNAIDS. 2008 Report on the global AIDS epidemic
. Geneva, Switzerland: UNAIDS; 2008.
4. World Health Organization. Antiretroviral Therapy for HIV Infection in Adults and Adolescents in Resource-Limited Settings: Towards Universal Access. Recommendations for a Public Health Approach. 2006 Revision
. Geneva, Switzerland: World Health Organization; 2006. Available at http://www.who.int/hiv/pub/guidelines/artadultguidelines.pdf
. Accessed March 24, 2010.
5. Phillips AN, Pillay D, Miners AH, et al. Outcomes from monitoring of patients on antiretroviral therapy in resource-limited settings with viral load, CD4 cell count, or clinical observation alone: a computer simulation model. Lancet
9. Government of India, Ministry of Health & Family Welfare, National AIDS Control Organization (NACO). Antiretroviral Therapy Guidelines for HIV-Infected Adults and Adolescents Including Post-exposure Prophylaxis
. 2007. Available at: http://upaidscontrol.up.nic.in/Guidelines/Antiretroviral
%20Therapy%20Guidelines%20for%20HIV%20infected%20Adults%20an.pdf. Accessed October 27, 2009.
11. Botswana, Ministry of Health, Department of HIV/AIDS Prevention and Care. Botswana National HIV/AIDS Treatment Guidelines: 2008 Version
. 2008. Available at: http://www.moh.gov.bw/templates/moh/File/BOTSWANA
%20HIVAIDS%20TREATMENT%20%20GUIDELINES%20(November%201%202008).pdf. Accessed October 27, 2009.
13. Republic of Uganda, Katabira ET, Kamya MR, et al, eds. National Antiretroviral Treatment Guidelines for Adults, Adolescents, and Children
. 2009. Available at: http://idi.mak.ac.ug/docs/mohartguidelines
. Accessed March 22, 2010.
14. The United Republic of Tanzania, Ministry of Health, NACP. National Guidelines for the Clinical Management of HIV and AIDS
. 2005. Available at: http://www.tanzaniahivaids.info/documents/Other
%20Resource%20Documents/Tz%20Clinical%20Guidelines%20for%20ART-2005.pdf. Accessed October 27, 2009.
18. Goldie SJ, Yazdanpanah Y, Losina E, et al. Cost-effectiveness of HIV treatment in resource-poor settings-the case of Côte d'Ivoire. N Engl J Med
19. Yazdanpanah Y, Losina E, Anglaret X, et al. Clinical impact and cost-effectiveness of co-trimoxazole prophylaxis in patients with HIV/AIDS in Côte d'Ivoire: a trial-based analysis. AIDS
20. Walensky RP, Weinstein MC, Yazdanpanah Y, et al. HIV drug resistance surveillance for prioritizing treatment in resource-limited settings. AIDS
21. Anglaret X, Chene G, Attia A, et al. Early chemoprophylaxis with trimethoprim-sulphamethoxazole for HIV-1-infected adults in Abidjan, Côte d'Ivoire: a randomised trial. Cotrimo-CI Study Group. Lancet
22. Minga A, Danel C, Abo Y, et al. Progression to WHO criteria for antiretroviral therapy in a 7-year cohort of adult HIV-1 seroconverters in Abidjan, Côte d'Ivoire. Bull World Health Organ
23. Seyler C, Messou E, Gabillard D, et al. Morbidity before and after HAART initiation in sub-Saharan African HIV-infected adults: a recurrent event analysis. AIDS Res Hum Retroviruses
25. Gold MR. Cost-Effectiveness in Health and Medicine
. New York, NY: Oxford University Press; 1996.
26. Russell LB, Gold MR, Siegel JE, et al. The role of cost-effectiveness analysis in health and medicine. Panel on Cost-Effectiveness in Health and Medicine. JAMA
27. Siegel JE, Weinstein MC, Russell LB, et al. Recommendations for reporting cost-effectiveness analyses. Panel on cost-effectiveness in health and medicine. JAMA
28. Weinstein MC, Siegel JE, Gold MR, et al. Recommendations of the panel on cost-effectiveness in health and medicine. JAMA
29. Freedberg KA, Kumarasamy N, Losina E, et al. Clinical impact and cost-effectiveness of antiretroviral therapy in India: starting criteria and second-line therapy. AIDS
. 2007;21(Suppl 4):S117-S128.
30. Walensky RP, Wood R, Weinstein MC, et al. Scaling up antiretroviral therapy in South Africa: the impact of speed on survival. J Infect Dis
31. Wolf LL, Ricketts P, Freedberg KA, et al. The cost-effectiveness of antiretroviral therapy for treating HIV disease in the Caribbean. J Acquir Immune Defic Syndr
32. Mellors JW, Muñoz A, Giorgi JV, et al. Plasma viral load and CD4+ lymphocytes as prognostic markers of HIV-1 infection. Ann Intern Med
33. Multicenter AIDS Cohort Study. Public use dataset: Release P04. National Technical Information Service; 1995.
34. World Health Organization. Guidelines on co-trimoxazole prophylaxis for HIV-related infections among children, adolescents and adults: recommendations for a public health approach. 2006. Available at: http://www.who.int/hiv/pub/guidelines/ctxguidelines.pdf
. Accessed March 22, 2010.
35. Deeks SG, Barbour JD, Martin JN, et al. Sustained CD4+ T cell response after virologic failure of protease inhibitor-based regimens in patients with human immunodeficiency virus infection. J Infect Dis
36. Cole SR, Hernan MA, Robins JM, et al. Effect of highly active antiretroviral therapy on time to acquired immunodeficiency syndrome or death using marginal structural models. Am J Epidemiol
37. Losina E, Yazdanpanah Y, Deuffic-Burban S, et al. The independent effect of highly active antiretroviral therapy on severe opportunistic disease incidence and mortality in HIV-infected adults in Côte d'Ivoire. Antivir Ther
38. Kuritzkes DR. Preventing and managing antiretroviral drug resistance. AIDS Patient Care STDS
39. Goetz MB, Ferguson MR, Han X, et al. Evolution of HIV resistance mutations in patients maintained on a stable treatment regimen after virologic failure. J Acquir Immune Defic Syndr
40. Lee KJ, Dunn D, Porter K, et al. Treatment switches after viral rebound in HIV-infected adults starting antiretroviral therapy: multicentre cohort study. AIDS
41. Petersen ML, van der Laan MJ, Napravnik S, et al. Long-term consequences of the delay between virologic failure of highly active antiretroviral therapy and regimen modification. AIDS
42. Toure S, Kouadio B, Seyler C, et al. Rapid scaling-up of antiretroviral therapy in 10,000 adults in Côte d'Ivoire: 2-year outcomes and determinants. AIDS
43. Seyler C, Anglaret X, Dakoury-Dogbo N, et al. Medium-term survival, morbidity and immunovirological evolution in HIV-infected adults receiving antiretroviral therapy, Abidjan, Côte d'Ivoire. Antivir Ther
44. Delfraissy JF, Flandre P, Delaugerre C, et al. Lopinavir/ritonavir monotherapy or plus zidovudine and lamivudine in antiretroviral-naive HIV-infected patients. AIDS
45. Johnson M, Grinsztejn B, Rodriguez C, et al. Atazanavir plus ritonavir or saquinavir, and lopinavir/ritonavir in patients experiencing multiple virological failures. AIDS
46. Marconi VC, Sunpath H, Lu Z, et al. Prevalence of HIV-1 drug resistance after failure of a first highly active antiretroviral therapy regimen in KwaZulu Natal, South Africa. Clin Infect Dis
48. El-Sadr WM, Lundgren JD, Neaton JD, et al. CD4+ count-guided interruption of antiretroviral treatment. N Engl J Med
49. Lopez AD, World Health Organization. World Mortality in 2000: Life Tables for 191 Countries
. Geneva, Switzerland: World Health Organization; 2002.
50. Danel C, Moh R, Anzian A, et al. Tolerance and acceptability of an efavirenz-based regimen in 740 adults (predominantly women) in West Africa. J Acquir Immune Defic Syndr
51. Edejer TT-T, World Health Organization. Making Choices in Health: WHO Guide to Cost-Effectiveness Analysis
. Geneva, Switzerland: World Health Organization; 2003.
53. Commission on Macroeconomics and Health. Macroeconomics and Health: Investing in Health for Economic Development
. Geneva, Switzerland: World Health Organization; 2001.
54. Murray CJ, Lauer JA, Hutubessy RC, et al. Effectiveness and costs of interventions to lower systolic blood pressure and cholesterol: a global and regional analysis on reduction of cardiovascular-disease risk. Lancet
56. International Monetary Fund. The World Economic Outlook Database, October 2008. Available at: http://imf.org
. Accessed March 26, 2009.
58. Schooley RT. Viral load testing in resource-limited settings. Clin Infect Dis
59. Calmy A, Ford N, Hirschel B, et al. HIV viral load monitoring in resource-limited regions: optional or necessary? Clin Infect Dis
60. Bendavid E. Monitoring of antiretroviral therapy in low-resource settings. Lancet
. 2008;372:288-289; author reply 289.
61. Bishai D, Colchero A, Durack DT. The cost effectiveness of antiretroviral treatment strategies in resource-limited settings. AIDS
62. Lawn SD, Bekker LG, Calmy A, et al. Monitoring of antiretroviral therapy in low-resource settings. Lancet
. 2008;372:287-288; author reply 289.
63. Walensky RP, Freedberg KA, Weinstein MC. Monitoring of antiretroviral therapy in low-resource settings. Lancet
. 2008;372:288; author reply 289.
64. Bendavid E, Young SD, Katzenstein DA, et al. Cost-effectiveness of HIV monitoring strategies in resource-limited settings: a southern African analysis. Arch Intern Med
65. Moore DM, Mermin J. Monitoring antiretroviral failure in resource-poor settings. Lancet
66. Vijayaraghavan A, Efrusy MB, Mazonson PD, et al. Cost-effectiveness of alternative strategies for initiating and monitoring highly active antiretroviral therapy in the developing world. J Acquir Immune Defic Syndr
67. Hallett TB, Gregson S, Dube S, et al. The impact of monitoring HIV patients prior to treatment in resource-poor settings: insights from mathematical modelling. PLoS Med
68. Bartlett JA, DeMasi R, Dawson D, et al. Variability in repeated consecutive measurements of plasma human immunodeficiency virus RNA in persons receiving stable nucleoside reverse transcriptase inhibitor therapy or no treatment. J Infect Dis
69. Lathey JL, Hughes MD, Fiscus SA, et al. Variability and prognostic values of virologic and CD4 cell measures in human immunodeficiency virus type 1-infected patients with 200-500 CD4 cells/mm3
(ACTG 175). AIDS Clinical Trials Group Protocol 175 Team. J Infect Dis
70. Glencross DK, Aggett HM, Stevens WS, et al. African regional external quality assessment for CD4 T-cell enumeration: development, outcomes, and performance of laboratories. Cytometry B Clin Cytom
. 2008;74(Suppl 1):S69-S79.
71. Reynolds SJ, Nakigozi G, Newell K, et al. Failure of immunologic criteria to appropriately identify antiretroviral treatment failure in Uganda. AIDS
72. Moore DM, Awor A, Downing R, et al. CD4+ T-cell count monitoring does not accurately identify HIV-infected adults with virologic failure receiving antiretroviral therapy. J Acquir Immune Defic Syndr
We are indebted to the entire CEPAC-International team for their contributions, including Christine Danel, Raoul Moh, Eric Ouattara, Eugène Messou, Catherine Seyler, and Siaka Toure` (Programme PAC-CI, Abidjan, Côte d'Ivoire); Yazdan Yazdanpanah (Service Universitaire des Maladies Infectieuses et du Voyageur, Centre Hospitalier de Tourcoing; EA 2694, Faculte` de Me`decine de Lille; and Laboratoire de Recherches Économiques et Sociales, Centre National de la Recherche Scientifique Unite` de Recherche Associe`e 362, Lille, France); Xavier Anglaret, Delphine Gabillard, Hapsatou Touré (INSERM U897, Unversité Bordeaux 2, Bordeaux, France; Nagalingeswaran Kumarasamy and J. Ganesh and A. K. Ganesh (Y.R. Gaitonde Centre for AIDS Research and Education, Chennai, India); Catherine Orrell and Robin Wood (University of Cape Town, Cape Town, South Africa); Neil Martinson and Lerato Mohapi (Perinatal HIV Research Unit, WITS Health Consortium, Johannesburg, South Africa); Kara Cotich, Sue J. Goldie, Marc Lipsitch, Alethea McCormick, Chara Rydzak, George R. Seage III, and Milton C. Weinstein (Harvard School of Public Health, Boston, MA); C. Robert Horsburgh (Boston University School of Public Health, Boston, MA); Elena Losina (Brigham and Women's Hospital, Boston, MA); Heather E. Hsu (Harvard Medical School, Boston, MA); Timothy Flanigan and Kenneth Mayer (Miriam Hospital, Providence, RI); A. David Paltiel (Yale University, New Haven, CT); April D. Kimmel (Weill Cornell Medical College, New York, NY); Aima Ahonkhai, Ingrid V. Bassett, Jessica Becker, Melissa A. Bender, John Chiosi, Andrea L. Ciaranello, Jennifer Chu, Kenneth A. Freedberg, Julie Levison, Benjamin P. Linas, Zhigang Lu, Sarah Lorenzana, Bethany Morris, Ji-Eun Park, Mai Pho, Erin Rhode, Callie A. Scott, Caroline Sloan, Adam Stoler, Lauren Uhler, Rochelle P. Walensky, Bingxia Wang, and Angela Wong (Massachusetts General Hospital, Boston, MA). We are also indebted to the CEPAC-International Scientific Advisory Board, including Richard Chaisson (Johns Hopkins University, Baltimore, MD); Victor De Gruttola (Harvard School of Public Health, Boston, MA); Joseph Eron (University of North Carolina, Chapel Hill, North Carolina); R. R. Gangakhedkar (National AIDS Research Institute, Pune, India); Jonathan Kaplan (Centers for Disease Control and Prevention, Atlanta, Georgia); Salim Karim (University of KwaZulu-Natal, Durban, South Africa); Thérèse N'Dri-Yoman (University of Cocody-Abidjan, Abidjan, Côte d'Ivoire); Douglas Owens (Stanford University, Palo Alto, CA); and John Wong (Tufts-New England Medical Center, Boston, MA). Cited Here...