In 2006, 2.3 million children under the age of 15 years were living with HIV/AIDS worldwide and 380 000 died . Without antiretroviral therapy (ART), HIV-infected African children have a median life expectancy of only 2 years , compared with 9–10 years in industrialized countries .
Cotrimoxazole prophylaxis substantially reduces HIV-related mortality, morbidity and hospital admission rates in adults and children in low-income countries without access to ART [4–7]. Recently updated World Health Organization (WHO) guidelines  advocate cotrimoxazole prophylaxis in all symptomatic adults and children, and asymptomatic adults and children with mild to moderate immune suppression, as a life-saving, simple, well-tolerated and inexpensive intervention for people living with HIV/AIDS, which should be implemented as part of the HIV chronic care package and as a key element of pre-ART care (Table 1). Cotrimoxazole prophylaxis has, however, yet to be adopted as a policy at a national level in many low-income countries. UNAIDS estimate that 4 million children who could benefit from cotrimoxazole are currently not receiving it, either because it is not universally available or the children have not yet been diagnosed with HIV .
The disparity between WHO guidance and uptake at a national level in many countries can be explained to some extent by the heavy burden of competing priorities. The objective of this analysis is to estimate the cost-effectiveness of adopting cotrimoxazole prophylaxis as a standard of care for HIV-infected children in Zambia using data from the Children with HIV Antibiotic Prophylaxis (CHAP) trial, thus providing evidence to inform allocative decisions by policy makers at a national government and global donor level .
The CHAP trial was a randomized double-blind trial of daily cotrimoxazole prophylaxis in 534 HIV-infected children after infancy in Zambia (median age at trial entry 4.5 years; range 1–14 years), which showed a 43% relative risk reduction in mortality over 19 months follow-up and a 23% reduction in hospital admissions . This occurred despite high levels of background in-vitro resistance to cotrimoxazole. The effect was observed across all ages and levels of CD4 cell percentage (CD4%), although the number of children with a high CD4% and no symptoms was relatively small. The main protective effect appeared to be a reduction in mortality and hospital admissions as a result of respiratory infections (presumed bacterial) .
A cost-effectiveness analysis, from the health service perspective, of cotrimoxazole prophylaxis in HIV-infected children older than 12 months was conducted expressing outcomes in terms of mean survival duration, quality adjusted life-years (QALYs) and disability adjusted life-years (DALYs). A decision-analytical model structured around CD4% as an indicator of HIV disease progression allowed extrapolation of both costs and survival beyond trial follow-up over the children's lifetime . A lifetime treatment effect was assumed on the basis of results from the CHAP trial suggesting no diminution of cotrimoxazole efficacy over the trial period (mean follow-up 19 months) despite considerable background resistance (∼60–80%) of common bacteria to cotrimoxazole [11,12]. The model was populated using data from the CHAP trial. Probabilistic sensitivity analysis allowed long-term estimates of the probability that prophylactic cotrimoxazole is cost-effective conditional on a threshold cost-effectiveness value of US$1019 for an additional (quality/disability adjusted) life-year in HIV-infected children after infancy, i.e. gross domestic product (GDP) per capita in Zambia (US$ in 2006) .
The model consisted of three live states described by CD4% and death (Fig. 1). CD4% is preferred over CD4 cell count as a predictive marker because of the rapid natural reduction in CD4 cell count with increasing age in uninfected young children. As the patient population of interest was not receiving ART, the model allowed only forward transitions (decline in CD4%). Transition probabilities between states in the model were calculated using patient-level data from the CHAP trial. Progression between CD4% states depended both on time in state and age at state entry, so Weibull models were used with fractional polynomial transformations of age at entry as covariates, whereas transitions to the death state only depended on age at state entry so an exponential model was used . A semi-Markov model is similar to a Markov cohort model, but can reflect transitions between health states that depend on elements of patient history, such as time spent in state. A semi-Markov model structure was thus constructed using the statistical programming language R in order to describe the decline in CD4% and mortality of the patient cohort . The model was run for a period of 20 years to ensure that the entire cohort had moved to the dead state.
Cost analysis was conducted in accordance with standard guidelines [16,17]. All costs are expressed in 2006 US$. Patient-specific direct medical costs including dose and number of days of drug therapy, numbers of inpatient days in the University Teaching Hospital, Lusaka, Zambia (UTH) and local health centres, outpatient visits and investigations were estimated from resource use data collected during the trial; unit costs were measured as described below.
Drug cost data in 2006 US$ were obtained from the Ministry of Health for all items included in the basic healthcare package. Prices for all other items were supplied by the hospital pharmacy at UTH. After confirmation of HIV status, children in the CHAP trial started cotrimoxazole or matched placebo and were seen 4-weekly for 16 weeks and then 8-weekly. Families were encouraged to bring children to the clinic for all intercurrent illnesses. Inpatient stays at UTH and at local health centres were recorded. All protocol-driven laboratory tests were excluded from cost estimations. Routine haematology and CD4% monitoring was excluded in the base case. Other routine laboratory and diagnostic investigations for the management of intercurrent illnesses were included. Unit costs of paediatric inpatient stays and outpatient visits to UTH were estimated by hospital finance staff using 2004 expenditure and a top-down approach. These costs in Kwacha (Zambian currency) were inflated to 2006 prices using the consumer price index  and converted to US$ at the rate K3307.97 = US$1 (June 2006).
The cost per outpatient visit at health centres and the costs per bed day in the local health centres, second and third level hospitals were derived from estimations of the cost of the basic healthcare package allowing the cost-effectiveness of adopting cotrimoxazole prophylaxis as part of the basic healthcare package to be estimated (Central Board of Health, Zambia, Costs of a basic healthcare package for 1st, 2nd and 3rd level referral in Zambia, 2004). These 2003 costs were inflated to 2006 as before .
CD4% was measured at 24-week intervals in the trial. To estimate the costs for each health state as defined by CD4%, resource use data incurred in the 12 weeks before and after each measurement of CD4% were assigned to the measured CD4% category. The cost per cycle for each CD4%-defined health state was estimated using a generalized linear model regression that assumed a gamma family distribution and a log link, adjusted for clustering by patient and including age as a covariate in the model. To assess the sensitivity of the cost estimates to any missing CD4 cell count measurements, two approaches were taken: (i) interpolation of missing CD4% values; and (ii) imputation of any measurements still missing using a simple linear regression that included age, WHO stage at baseline and available CD4% data.
Utility and disability weights
Health-related quality of life weights (in terms of utilities on a 0 (equivalent to death) to 1 (equivalent to full health) scale) are required in order to estimate the incremental cost-effectiveness in terms of cost per QALY. There are no published utility weights for HIV-infected children. Therefore, we used estimates from a meta-analysis of utility values from 25 adult HIV/AIDS studies from 1985–2000 in high-income countries . To investigate the impact of the change in definition of AIDS in 1993 and the improved treatment options available over time, the year of publication was included in the model but was found not to predict utility. The utility value for AIDS from the meta-analysis (0.702) was almost identical to the utility value for AIDS patients in a township clinic in South Africa before initiating ART (0.70) . We thus made the assumption that these values could also be applied to HIV-infected African children not receiving ART.
The cost-effectiveness was also expressed in terms of DALYs calculated in accordance with standard methods and using disability weights for HIV and AIDS and an average life expectancy in Zambia of 39.7 years [21–24]. Standard age weighting of DALYs, which increases the value of life up to the age of 25 years and then decreases, was omitted in a sensitivity analysis.
The utility values and the disability weights associated with clinical descriptors were assigned to model states defined by the proportion of patients in each CD4% category with AIDS (WHO stage 4) or with HIV (WHO stage < 4) in the baseline cohort for the CHAP trial.
The characteristics of patients in the base case were similar to those of the CHAP children at trial entry (Table 2). A discount rate of 3% was applied to both costs and outcomes . Routine follow-up haematology was not included in the base case in accordance with WHO guidance .
Inputs were entered into the model as appropriate probability distributions, e.g. the results of the regression equations used to estimate the transition probabilities and health state costs were used to define multivariate normal distributions for the estimated coefficients, with correlation incorporated using the Cholesky decomposition of the covariance matrix . Having specified distributions for all parameters to reflect uncertainty, Monte Carlo simulation was used to propagate the uncertainty through the model to be jointly reflected in a distribution of mean costs and outcomes per patient, i.e. the expected costs and health benefits were calculated on the basis of 50 000 repeated random draws from the specified input distributions, so that the set of results reflect the range and likelihood of possible values for each input parameter. The uncertainty around these estimates of cost-effectiveness was addressed using cost-effectiveness acceptability curves, which describe the probability that each treatment is cost-effective . In addition, sensitivity analyses were undertaken to explore the robustness of the model, and uncertainty from sources other than imprecision of the input parameters.
A technical appendix providing greater detail on methods is available from the corresponding author on request.
Internal validity of the model
The projected survival at different starting ages was within 10% of survival measured in the CHAP trial (median starting age of 4.4 years).
Unit costs are summarized in Table 3. Although children in CHAP received syrup as part of the trial, cotrimoxazole tablets are cheaper than suspension. A survey of formulation acceptability to patients and carers revealed that most families preferred the tablets (D.M. Gibb, personal communication). Therefore, we used the tablet price in the base case and included the cost of children under 5 years of age receiving suspension in a sensitivity analysis.
The equality-of-distributions test indicated that the average cost for each CD4% category did not differ significantly according to which CD4% profile was used (observed, interpolated or interpolated and imputed), and so the observed data were used in the base case analysis. Children with CD4% of 16% or higher had lower inpatient and outpatient costs (P < 0.05), and those with CD4% of 8–15% also had lower inpatient costs than those with CD4% of 7% or lower, although their outpatient costs were similar (Table 4). Older children had lower outpatient costs, and children on cotrimoxazole had lower inpatient costs.
The incremental cost per life-year saved was US$72. The incremental cost per QALY was US$94. The incremental cost per DALY averted was US$53 (Table 5). The cost-effectiveness acceptability curves for each outcome show a 100% probability that cotrimoxazole is cost-effective at a threshold willingness to pay of US$1019, i.e. the mean GDP per capita in Zambia in 2006  (Fig. 2).
We conducted a scenario analysis that incorporated costs should cotrimoxazole prophylaxis be implemented at a local health centre rather than a tertiary care level. Outpatient visit costs at the local health centres were approximately US$2 less than at UTH, whereas the weighted inpatient cost across all levels of inpatient care was similar to the cost at UTH. Cotrimoxazole prophylaxis as part of the basic healthcare package implemented at a local health centre level yielded incremental cost-effectiveness ratios (ICERs) of less than or equal to US$5 per outcome.
The ICERs for all outcome measures remained below US$250 for a range of sensitivity analyses including bi-monthly haematology and 6-monthly CD4 cell counts as standard practice while on cotrimoxazole prophylaxis, varying the proportions with AIDS in each CD4 cell category, allowing children under 5 years to receive cotrimoxazole suspension, decreasing the duration of treatment effect to trial follow-up, omitting age weighting of DALYs and varying the discount rate to 0% and 5% (Table 5). Varying the age of the starting cohort to greater than one year but less than 8 years yields ICERs of less than US$300 per outcome. At starting cohort aged 9 years or greater, cotrimoxazole dominates, i.e. its positive treatment effect is cost saving.
This study demonstrates that cotrimoxazole prophylaxis is a highly cost-effective intervention for HIV-infected children in a high HIV prevalence low-income country. The ICERs for cotrimoxazole prophylaxis are US$72 per life-year gained, US$94 per QALY and €53 per DALY in the base case.
This is the first estimation of the cost-effectiveness of cotrimoxazole prophylaxis in HIV-infected African children after infancy. There have been a number of cost-effectiveness studies, which although based on different modelling approaches, have yielded similar cost-effectiveness ratios for cotrimoxazole prophylaxis in HIV-infected African adults [27–30]. Goldie et al.  used a decision model including healthcare costs only, which incorporated CD4 cell count, HIV viral load and history of opportunistic infections, and reported an ICER of US$240 per life-year saved (2002 US$) over lifetime follow-up.
Marseille et al.  estimated the cost-effectiveness of cotrimoxazole prophylaxis from the perspective of a company healthcare scheme in a cohort of relatively healthy Ugandan workers (90% WHO stage 1, 10% WHO stage 2). Cotrimoxazole prophylaxis was initiated at the onset of WHO stage 2 with a 5-year time horizon. The model incorporated direct medical costs and indirect costs to the company, including absenteeism, funeral expenses and death benefit payments, with the company bearing healthcare costs only until disability. The base case analysis found cotrimoxazole prophylaxis to be cost saving for skilled and unskilled workers over a 5-year time horizon. In a 10-year time frame the ICER were US$110 per DALY and US$225 per DALY (2005 US$) for skilled and unskilled workers, respectively. Pitter et al.  found cotrimoxazole prophylaxis as a component of home-based care in rural Uganda to be cost-saving over a one-year time frame if administered to all HIV-infected adults and children over 5 years regardless of clinical or immunological criteria. Low unit costs for a home-based care model in a rural setting enhance the cost-effectiveness of the intervention.
There is no definitive interpretation of a threshold for cost-effectiveness in resource-limited countries. WHO has suggested that interventions with ICERs less than the GDP per capita may be considered highly cost-effective . In this study, in all sensitivity and scenario analyses the ICERs remain below US$300, substantially less than the GDP per capita for Zambia in 2006 of US$1019 . Therefore, cotrimoxazole prophylaxis in HIV-infected children may be considered highly cost-effective for the Zambian healthcare system using WHO criteria. It has been argued that the cost-effectiveness threshold should reflect the health gains that are displaced when a health system adopts a new intervention that increases costs . No data on these ‘opportunity costs’ are available for Zambia.
Of note is the fact that the cost-effectiveness ratios were highly sensitive to the cost of outpatient visits. Providing cotrimoxazole prophylaxis as part of the basic healthcare package yielded ICERs less than or equal to US$5 per outcome, emphasizing that cotrimoxazole is an extremely cost-effective intervention if employed at the local health centre level. Although our effectiveness data were taken from a trial based in a tertiary care setting, extrapolation to the primary healthcare level is supported by very similar efficacy (46% relative risk reduction in mortality) demonstrated in a trial of cotrimoxazole prophylaxis in a rural setting in Uganda .
Alternatively, we can compare the ICERs to those of interventions that are already considered cost-effective in the African setting, e.g. prevention of mother-to-child transmission US$1–731 per DALY, preventive therapy for tuberculosis US$169–288 per DALY, malaria control US$1–121 per DALY [33,34]. The cost-effectiveness of ART in HIV-infected African adults has been estimated at US$984 per life-year and US$1102 per QALY . The cost-effectiveness of ART in children has yet to be estimated. Although having a markedly different impact on life expectancy, however, ART delivery for children is relatively complex and is currently limited to tertiary or secondary care levels in most African countries. It is clear that cotrimoxazole prophylaxis falls well within the range of interventions currently funded by African healthcare systems, and is a low-cost low-technology intervention that can be rolled out everywhere now as a prelude to ART.
WHO guidelines advocate clinical monitoring only of HIV-infected children on cotrimoxazole prophylaxis with no requirement for haematology and CD4 cell count tests unless already part of routine care . In the base case analysis, we assumed clinical follow-up alone, reflecting actual practice should cotrimoxazole be implemented at the local health centre level, where laboratory tests are not always routinely available. In a sensitivity analysis, routine blood monitoring doubled the cost-effectiveness ratios but the results were still well below the cost-effectiveness threshold. Cotrimoxazole prophylaxis thus remains highly cost-effective even if routine blood monitoring tests become more readily available.
Cost-effectiveness was also sensitive to age at initiation of cotrimoxazole, increasing up to 8 years and becoming cost saving for patients 9 years or older. This may be explained by the fact that vertically infected children are at higher risk of disease progression as they grow older. The higher absolute risk of older children is therefore associated with a greater absolute benefit from cotrimoxazole for the same relative risk reduction and thus increased cost-effectiveness.
In sensitivity analysis investigating the impact of reducing the duration of efficacy to trial follow-up (19 months) versus patient lifetime in the base case, cotrimoxazole was slightly less cost-effective. The small difference in ICERs may be attributed to the relatively short mean survival (3.26 years for placebo, 5.01 years for cotrimoxazole) of patients entered in to the model.
A number of limitations apply to our model. Many recent models of HIV disease utilize virological as well as immunological data to predict disease progression [36,37]. HIV viral load data are not routinely available in Africa and our model is based on the CD4% data available from the CHAP trial. There is no valid and reliable measure to calculate QALYs in children . Adult utility weights from a high-income setting were used in the model . We were, however, able to validate the utilities with one study in South African township HIV-infected adults before accessing ART .
Of note is the fact that the analysis presented here focused on definitively HIV-infected children after infancy who are at high risk of recurrent bacterial infections; reducing this risk is the most likely mode of action of cotrimoxazole prophylaxis . Here, we do not consider the cost-effectiveness of the provision of cotrimoxazole prophylaxis to all HIV-exposed infants born to known HIV-positive women. The latter is a different scenario that would need to balance the benefits of preventing early Pneumocystis jiroveci pneumonia in those infants who are actually HIV infected, with the costs of treating many babies who may later be found not to have HIV infection, particularly when prevention of mother-to-child transmission programmes are in place. In addition, the cost-effectiveness of cotrimoxazole prophylaxis in children who have access to ART cannot be estimated from our data but is likely to remain as or more cost-effective, as has been demonstrated in adults with access to ART .
The national implementation of cotrimoxazole prophylaxis lags behind WHO guidance in many countries. Policy decisions take into account not only cost-effectiveness but also sustainability, infrastructure, healthcare worker capacity, social justice and equity considerations as well as competing priorities. Cotrimoxazole prophylaxis is a low-cost, low-technology intervention capable of delivery at the local health centre level, whose clinical efficacy has been proved in a low-income country with high background cotrimoxazole resistance. Zambia is implementing a national programme in 2007. One novel approach to the scarcity of doctors in rural clinics may be the deregulation of prescribing of cotrimoxazole prophylaxis to other cadres, e.g. nurses and healthcare assistants (C. Chintu, personal communication). Some countries, e.g. Malawi, have addressed problems with regular cotrimoxazole supply by linking it to special supply chains established for ART (R. Weigel, personal communication). The evidence of cost-effectiveness presented here may be used to inform global donors of the value of investing in upfront costs including national training and public awareness programmes as well as in acquisition costs of cotrimoxazole.
In conclusion, this study demonstrates the cost-effectiveness of cotrimoxazole prophylaxis in HIV-infected children after infancy, which together with data from adult cost-effectiveness analyses, adds weight to WHO recommendations and strongly supports the adoption of cotrimoxazole prophylaxis as part of the basic healthcare package in Zambia and similar resource-constrained settings.
M.R., M.J.S., B.C., M.G.B., C.M., C.C., V.M. and D.M.G. applied for funding. M.R., S.G., M.J.S., B.C., A.S.W. and D.M.G. designed the study. M.R., B.C., V.M. and D.K. collected data. M.R., S.G. and A.S.W. analysed data. S.G., N.H. and M.R. constructed the decision analytical model. All authors contributed to interpretation of the data. M.R. and S.G. wrote the manuscript with editing by all authors. All authors were involved in the decision to submit the manuscript for publication.
The authors would like to thank the children and families in enrolled in the CHAP trial, the study team, and other staff at University Teaching Hospital, Lusaka and the School of Medicine, University of Zambia. They especially thank the finance and administration staff involved in cost data collection, and would like to thank Dr Margaret Thomason, Clinical Trials Unit, Medical Research Council, UK, for her help with data collection. The authors acknowledge Professor G. Bhat and Professor A. Nunn who helped with the application for funding. Finally they acknowledge the very helpful comments provided by two unknown referees.
Sponsorship: The economic analysis was sponsored by an unrestricted grant from the Advisory Board of Irish Aid, Department of Foreign Affairs, Ireland. The funding source had no role in study design, data analysis, data interpretation or writing of the report.
Conflicts of interest: None.
2. Newell ML, Coovadia H, Cortina-Borja M, Rollins N, Gaillard P, Dabis F, et al. Mortality of infected and un-infected infants born to HIV-infected mothers in Africa; a pooled analysis. Lancet 2004; 364:1236–1243.
3. European Collaborative Study. Fluctuations in symptoms in HIV-infected children: the first 10 years of life. Pediatrics 2001; 108:116–122.
4. Chintu C, Bhat GJ, Walker AS, Mulenga V, Sinyinza F, Lishimpi K, et al. Cotrimoxazole as prophylaxis against opportunistic infections in HIV-infected Zambian children (CHAP): a double-blind randomized placebo-controlled trial. Lancet 2004; 364:1865–1871.
5. Wiktor SZ, Sassan-Morroko M, Grant AD, Abouya L, Karon JM, Maurice C, 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.
6. Anglaret X, Chene G, Attia A, Toure S, Lafont S, Coombe P, et al. Early chemoprophylaxis with trimethoprim-sulphamethoxazole for HIV-1 infected adults in Abidjan, Cote d'Ivoire: a randomised trial. Lancet 1999; 353:1463–1468.
7. Mermin J, Lule J, Ekwaru JP, Malamba S, Downing R, Ransom R, et al. Effect of cotrimoxazole prophylaxis on morbidity, mortality, CD4-cell count, and viral load in HIV infection in rural Uganda. Lancet 2004; 364:1428–1434.
8. World Health Organization. HIV/AIDS Program. Guidelines for co-trimoxazole prophylaxis for HIV-related infections among children, adolescents and adults in resource limited settings, 2006. Available at: www.who.int/hiv/pub/guidelines/ctx
. Accessed: 12 January 2007
9. Mulenga V, Ford D, Walker AS, Mwenya D, Mwansa J, Sinyinza F, et al. Effect of cotrimoxazole on causes of death, hospital admissions and antibiotic use in HIV-infected children in the CHAP trial. AIDS 2007; 21:77–84.
10. Walker S, Mulenga V, Sinyinza F, Lishimpi K, Nunn A, Chintu C, et al. Determinants of survival without antiretroviral therapy after infancy in HIV-1 infected Zambian children in the CHAP trial. J Acquir Immune Defic Synd 2006; 42:637–645.
11. Mwansa J, Mutela K, Zulu I, Amadi B, Kelly P. Antimicrobial sensitivity in Enterobacteria in AIDS patients, Zambia. Emerg Infect Dis J 2002; 8:92–93.
12. Mwenya DM, Charalambous BM, Gibb DM, Nunn A, Mwansa JCL, Gillespie SH. Impact of co-trimoxazole on carriage and antibiotic resistance of Streptococcus pneumoniae and Haemophilus influenzae in HIV infection children in Zambia. In: Fifth International Symposium on Pneumococci and Pneumococcal Diseases. Alice Springs, Central Australia; April 2006. Abstract PO4.12.
14. Collett D. Modeling survival data in medical research. London, UK: Chapman and Hall/CRC; 1994.
15. Hawkins N, Sculpher MJ, Epstein D. Cost-effectiveness analysis of treatments for chronic disease: using R to incorporate time dependency of treatment response. Med Decis Making 2005; 25:511–519.
16. Drummond M, Sculpher M, Torrance G, O'Brien B, Stoddart G. (eds.) Methods for the economic evaluation of healthcare programmes, 3rd Edn. Oxford, UK: Oxford Medical Publications; 2005.
17. Kumaranayake L, Pepperall J, Goodman H, Mills A, Walker D. Costing guidelines for HIV prevention strategies 2000. Health Economics and Financing Program, Health Policy Unit, London School of Hygiene and Tropical Medicine. Available at: www.hivtools.lshtm.ac.uk/downloads/costings/costgui.pdf
. Accessed: January 2007
18. Central Statistics Office of Zambia. Available at: www.zamstats.gov.zm
. Accessed: January 2007
19. Tengs T, Lin T. A meta-analysis of utility estimates for HIV/AIDS. Med Decis Making 2002; 22:475–481.
20. Hughes J, Jelsma J, Maclean E, Darder M, Tinise X. The health-related quality of life of people living with HIV/AIDS. Disabil Rehabil 2004; 26:371–376.
21. Murray CJ. Quantifying the burden of disease: the technical basis for disability-adjusted life years. Bull WHO 1994; 72:429–445.
22. Fox-Rushby J, Hanson K. Calculating and presenting disability adjusted life years (DALYs) in cost-effectiveness analysis. Health Policy Plan 2001; 16:326–331.
23. Sassi F. Calculating QALYs, comparing QALY and DALY calculations. Health Policy Plan 2006; 21:402–408.
25. Evans D, Tan-Torres Edejer T, Adam T, Lim SS. Achieving the millennium development goals for health: methods to assess the costs and health effects of interventions for improving health in developing countries. BMJ 2005; 331:1137–1140.
26. Briggs A, Sculpher M, Claxton K. Decision modelling for health economic evaluation. Oxford: Oxford University Press; 2006.
27. Yazdanpanah Y, Losina E, Anglaret X, Goldie S, Walensky R, Weinstein M, et al. Clinical impact and cost-effectiveness of cotrimoxazole prophylaxis in patients with HIV/AIDS in Côte d'Ivoire: a trial based analysis. AIDS 2005; 19:1299–1308.
28. Goldie S, Yazdanpanah Y, Losina E, Weinstein M, Anglaret X, Walensky R, et al. Cost-effectiveness of HIV treatment in resource-poor settings – the case of Côte d'Ivoire. N Engl J Med 2006; 355:1141–1153.
29. Marseille E, Saba J, Muyingo S, Kahn J. The costs and benefits of private sector provision of treatment to HIV-infected employees in Kampala, Uganda. AIDS 2006; 20:907–914.
30. Pitter C, Kahn J, Marseille E, Lule JR, McFarland DA, Ekwaru JP, et al. Cost-effectiveness of cotrimoxazole prophylaxis among persons with HIV in Uganda. J Acquir Immune Defic Syndr 2007; 44:336–344.
32. Culyer A, McCabe C, Briggs A, Klaxton K, Buxton M, Akehurst R, et al. Searching for a threshold, not setting one: the role of the National Institute for Health and Clinical Excellence. J Health Serv Res Policy 2007; 12:56–58.
33. Creese A, Floyd K, Alban A, Guinness L. Cost-effectiveness of HIV/AIDS interventions in Africa: a systematic review of the evidence. Lancet 2002; 359:1635–1642.
34. Goldman CA, Coleman PG, Mills AJ. Cost-effectiveness of malaria control in sub-Saharan Africa. Lancet 1999; 354:378–385.
35. Cleary SM, McIntyre D, Boulle AM. The cost-effectiveness of antiretroviral treatment in Khayelitsha, South Africa – a primary data analysis. Cost Effect Resource Alloc 2006; 4:20, E-pub.
36. Sanders GD, Bayoumi AM, Sundaram V, Bilir SP, Neukermans CP, Rydzak CE, et al. Cost-effectiveness of screening for HIV in the era of highly active antiretroviral therapy. N Engl J Med 2005; 352:570–585.
37. Sendi P, Gunthard HF, Simcock M, Ledergerber B, Schüpbach J, Battegay M, et al. Cost-effectiveness of genotypic antiretroviral resistance testing in HIV-infected patients with treatment failure. PLoS ONE 2007; 2:e173.
38. Ungar WJ. Paediatric health economic evaluations: a world view. Healthcare Q 2007; 10:134–140, 142–5; discussion 145–6.
© 2008 Lippincott Williams & Wilkins, Inc.