Search tips
Search criteria 


Logo of cidLink to Publisher's site
Clin Infect Dis. 2010 December 1; 51(11): 1325–1333.
PMCID: PMC3058781

Viral Load Predicts New World Health Organization Stage 3 and 4 Events in HIV-Infected Children Receiving Highly Active Antiretroviral Therapy, Independent of CD4 T Lymphocyte Value


Background. Many resource-limited countries rely on clinical and immunological monitoring without routine virological monitoring for human immunodeficiency virus (HIV)-infected children receiving highly active antiretroviral therapy (HAART). We assessed whether HIV load had independent predictive value in the presence of immunological and clinical data for the occurrence of new World Health Organization (WHO) stage 3 or 4 events (hereafter, WHO events) among HIV-infected children receiving HAART in Latin America.

Methods. The NISDI (Eunice Kennedy Shriver National Institute of Child Health and Human Development International Site Development Initiative) Pediatric Protocol is an observational cohort study designed to describe HIV-related outcomes among infected children. Eligibility criteria for this analysis included perinatal infection, age <15 years, and continuous HAART for ≥6 months. Cox proportional hazards modeling was used to assess time to new WHO events as a function of immunological status, viral load, hemoglobin level, and potential confounding variables; laboratory tests repeated during the study were treated as time-varying predictors.

Results. The mean duration of follow-up was 2.5 years; new WHO events occurred in 92 (15.8%) of 584 children. In proportional hazards modeling, most recent viral load >5000 copies/mL was associated with a nearly doubled risk of developing a WHO event (adjusted hazard ratio, 1.81; 95% confidence interval, 1.05–3.11; P = .033), even after adjustment for immunological status defined on the basis of CD4 T lymphocyte value, hemoglobin level, age, and body mass index.

Conclusions. Routine virological monitoring using the WHO virological failure threshold of 5000 copies/mL adds independent predictive value to immunological and clinical assessments for identification of children receiving HAART who are at risk for significant HIV-related illness. To provide optimal care, periodic virological monitoring should be considered for all settings that provide HAART to children.

The importance of quantitative human immunodeficiency virus (HIV) RNA, or viral load, assays for monitoring the treatment of HIV infection is well established for adults and children and is considered standard practice in higher-resource settings [1, 2]. Such monitoring allows for identification of virological treatment failure and use of interventions to achieve virological suppression, prevent immunological deterioration, and avoid the development of drug resistance. In addition, virological monitoring can distinguish virological failure from other causes of apparent treatment failure, thus averting unnecessary changes in antiretroviral therapy [3].

Viral load assays, however, are not widely available in resource-limited settings, because they are costly and technically difficult to implement and maintain. As a result, treatment of HIV-infected children in many regions of the world has often been guided by clinical and immunological assessment alone, consistent with World Health Organization (WHO) guidelines [4].

Although viral load before initiation of highly active antiretroviral therapy (HAART) has been established as an independent predictor of disease progression in children [57], the value of routine viral load measurement in addition to immunological monitoring for prediction of new HIV-related clinical events in children receiving HAART has not been clearly demonstrated.

The objective of the present analysis was to assess the independent contribution of viral load in predicting the occurrence of new WHO stage 3 or 4 events among a cohort of perinatally HIV-infected Latin American children receiving continuous HAART who were enrolled in the NISDI (Eunice Kennedy Shriver National Institute of Child Health and Human Development [NICHD] International Site Development Initiative) study.


The NISDI Pediatric Protocol is an ongoing prospective cohort study of HIV-infected and HIV-exposed uninfected children at multiple clinical sites in Latin America. HIV-infected infants, children, and adolescents (≤21 years old) enrolled from September 2002 until April 2007 were included in this analysis. A more detailed description of the protocol and cohort has been published elsewhere [8].

At enrollment into the NISDI study, demographic, laboratory (CD4 T lymphocyte values, viral load, and results of hematology and chemistry assays), and clinical data were systematically collected. Clinical information included history of antiretroviral and other medications, history of prior infections since birth, and current growth parameters. Standardized criteria developed by the Pediatric AIDS Clinical Trials Group for diagnosing specific diseases were modified as necessary for application to the NISDI protocol ( .html).

All enrolled children were medically evaluated in a standardized fashion at baseline and every 6 months thereafter, including assignment of Centers for Disease Control and Prevention (CDC) HIV disease classification, report of interim infections and diagnoses, laboratory assessments (same as baseline), and growth parameters. Food and Drug Administration- approved assays were used to perform viral load testing at the sites, in accordance with manufacturer instructions; results were made available to the treating clinician.

Perinatally HIV-infected children who had been treated with HAART regimens continuously for at least 6 months were eligible for inclusion in this analysis. HAART was defined as at least 3 different drugs from at least 2 classes: nucleoside reversetranscriptase inhibitors, protease inhibitors, and nonnucleoside reverse-transcriptase inhibitors. Eligibility could occur as early as the day of enrollment into the NISDI study or when a subject reached 6 months of continuous HAART during the study.

Only new WHO stage 3 or 4 events (hereafter, WHO events) [9] occurring after eligibility for the analysis were considered as outcomes. WHO-defined outcomes were used because this is the classification system used in Latin American countries. Because WHO definitions for stage 3 and 4 events change slightly after the age of 15 years, the analyses were restricted to subjects <15 years old at the time of eligibility, and subjects were censored once they reached this age. Because the NISDI protocol used the CDC HIV disease classification system and data that were required for reclassification according to WHO criteria were often missing for events occurring before enrollment, only the CDC classification (instead of the WHO classification) was used for preentry clinical staging.

Immunological status defined on the basis of CD4 T lymphocyte values (hereafter, CD4-defined immunological status)was categorized using age-related WHO immunological classification as severe, advanced, mild, or no (or nonsignificant) immunosuppression ( An algorithm from the WHO (AnthroPlus; http:// was used to convert anthropometric data at eligibility into age-and sex-adjusted z scores: heightfor-age and body mass index (BMI) z scores for all children and weight-for-age z scores for children ≤10 years old (AnthroPlus did not have conversion tables for children 110 years old).

For statistical analysis, the primary outcome measure was the first occurrence of a new WHO event after the time of eligibility for this analysis; events occurring before or at eligibility were not counted, nor were those that were present at eligibility that continued to be reported at subsequent visits. Simple descriptive statistics were used to describe WHO events. Bivariable analyses (the Wilcoxon rank-sum test for continuous variables or the Fisher exact test for categorical variables) were used to examine the association between new WHO events and characteristics assessed at the time of study eligibility.

Cox proportional hazards modeling was used to assess the time to WHO events as a function of CD4-defined immunological status, viral load, and hemoglobin level. Because these laboratory measures were assessed repeatedly over the course of follow-up, they were treated as time-varying predictors; laboratory values had to be obtained at least 14 days before the WHO event (since these measures can be transiently affected by an intercurrent infection) and no more than 183 days before the WHO event (since study visits occurred at 6-month intervals). Covariates associated with the occurrence of WHO events at an a level of ≤0.2 in bivariable analyses were identified as potential confounders and considered as candidates for inclusion in the modeling. Subjects were censored for interruption of HAART of 7 days or more, at the time of the first WHO event, or at the time of their 15th birthday, whichever occurred first. In separate models, viral load was treated as a continuous measure, as a dichotomous indicator measure for detectable versus nondetectable level (≥400 vs <400 copies/mL), and as a dichotomous measure using the most recent WHO virological definition for confirmation of treatment failure (>5000 vs ≤5000 copies/mL) [10]. The possibility of an interaction between CD4-defined immunological status and each coding system for viral load was investigated.

All analyses were conducted using SAS software, version 9.0 (SAS Institute).


Between September 2002 and April 2007, 1629 children were enrolled in the NISDI Pediatric Protocol, including 816 HIV1-infected children, of whom 731 were perinatally infected. Of the perinatally infected children, 147 were excluded (10 who enrolled after their 15th birthday and 137 who had received <6 months of continuous HAART by the time of their 15th birthday), leaving 584 children in the final study population.

Among those eligible for this analysis, 63% were born in Brazil, 46% were male, and 37% were <5 years old at the time of eligibility; 69% had been receiving continuous HAART regimens for at least 6 months at the time of enrollment into the NISDI study and thus were immediately eligible for this analysis (Table 1). The mean duration of follow-up after meeting eligibility for this analysis was 2.5 years (median, 2.7 years). Ninety-two (15.8%) of the 584 eligible subjects experienced a WHO event during follow-up; of the 92 events, 73 (79%) were classified as WHO stage 3. The most commonly reported WHO events were oral candidiasis (n = 17), unexplained diarrhea lasting 14 days or more (n = 17), recurrent pneumonia (n = 14), low weight-for-age z score (n = 14), and unexplained anemia, neutropenia, or thrombocytopenia (n = 8) (Table 2). Weight-for-age z score, CD4-defined immunosuppression, viral load, hemoglobin level, and CDC disease classification at eligibility were significantly associated with the risk of a new WHO event (P<.01) (Table 1). Mean length of continuous HAART at eligibility did not differ between those who later experienced a WHO event and those who did not (2.1 years for both groups; P = .6). Of those not experiencing a new WHO event, 27 were censored at the time of interruption of HAART, and 1 was censored at the time the subject withdrew consent. Overall, new WHO events occurred at a median time from eligibility of 1 year (range, 4 days to 3.9 years); subjects who did not experience a new WHO event were followed up for a median duration of 3.2 years (range, 1 day to 5 years). Twenty-one percent of subjects had at least one regimen change during follow-up; viral load before regimen change was >5000 copies/ mL in 70% and did not differ between those who did and those who did not go on to experience a new WHO event.

Table 1.
Baseline Characteristics according to Occurrence of the First New World Health Organization (WHO) Stage 3 or 4 Event
Table 2.
First Stage 3 or 4 World Health Organization (WHO) Event after Meeting Study Eligibility

Proportional hazard modeling found no evidence of an interaction between CD4-defined immunological status and viral load. Models then examined the independent effect of immunological status, viral load, and hemoglobin level, with adjustment for candidate covariates, including age, BMI z score, CDC classification, and current HAART regimen. Country of birth was not included in the models because the small number of WHO events observed for most countries would prevent the models from running successfully. Because WHO weight-forage z scores are not available beyond 10 years of age, BMI z scores were substituted in the modeling.

In the model using the WHO virological definition of treatment failure (>5000 vs ≤5000 copies/mL), participants with severe (adjusted hazard ratio [AHR], 3.37; 95% confidence interval [CI], 1.66–6.84) or mild/advanced (AHR, 2.39; 95% CI, 1.34–4.25) immunosuppression had a significantly increased risk of developing a WHO event, compared with those with no significant immunosuppression (P = .001) (Table 3). In the presence of CD4-defined immunological status, viral load and hemoglobin level made significant independent contributions to predicting the occurrence of new WHO events. A viral load >5000 copies/mL was associated with a nearly 2-fold increased risk of developing a new WHO event (AHR, 1.81; 95% CI, 1.05–3.11; P = .033). Each 1 g/dL decrease in hemoglobin level was associated with a 32% increased likelihood of developing a new WHO event (AHR, 1.32; 95% CI, 1.11–1.56; P = .002). The model also adjusted for BMI z score and age.

Table 3.
Risk (Adjusted Hazard Ratio) of New World Health Organization (WHO) Stage 3 or 4 Events in Children Receiving Highly Active Antiretroviral Therapy

A similar modeling approach was used with viral load treated as a continuous measure (log10 transformed to address the skewness of the distribution) and with a viral load cut point of 400 copies/mL (ie, detectable vs undetectable) (Table 3). Each 0.5 log10 increase in viral load retained a modest but significant independent effect on predicting WHO events (AHR, 1.14; 95% CI, 1.03–1.27). The AHR for detectable viral load was significantly elevated when hemoglobin level was not included in the model (data not shown) but was no longer independently associated with predicting WHO events in the presence of hemoglobin level (AHR, 1.61; 95% CI, 0.88–2.96; P = .123). A 1 g/dL decrease in hemoglobin level was statistically associated with subsequent new WHO events in all 3 models, even after controlling for CD4-defined immunological status and viral load.


In this large cohort of perinatally HIV-infected Latin American children receiving HAART for at least 6 months, most recent viral load, particularly with a threshold of >5000 copies/mL, was an independent predictor of WHO stage 3 and 4 events, even after adjusting for most recent CD4-defined immunosuppression, hemoglobin level, and other cofactors. This analysis supports the clinical value of routine virological monitoring in children receiving HAART and adds pediatric data to the growing body of evidence from studies in adults for the need to make routine virological monitoring available in all HIV treatment settings [11, 12].

HAART has led to dramatic reductions in mortality and illness in adults and children worldwide [1315], but several factors contribute to the persistent risk of poor outcomes despite HAART. During the first months after HAART initiation, when preexisting immunosuppression and (often) malnutrition have not yet been corrected, illness and death remain major problems [16, 17]. In addition, during the first 3–6 months of therapy immune reconstitution inflammatory syndrome (IRIS), which is often related to mycobacterial infections, are common causes of clinical illness [18]. In this study, all subjects had received at least 6 months of continuous HAART; thus, the observed illness events would not be attributable to the expected persistent risk of elevated morbidity during the early treatment period or misclassification of IRIS events. Instead, the WHO events that occurred among the nearly 16% of children in this Latin American cohort are most likely the result of treatment failure due to the reduction of ART effectiveness by nonadherence, drug resistance, less potent regimens, or a combination of these factors.

The goal of monitoring children receiving HAART should be to detect evidence of treatment failure as early as possible and to identify children at increased risk of HIV-related clinical illnesses and immunological decline before these events occur. However, in resource-limited settings where virological testing is not widely available, prediction of virological failure by clinical and immunological assessments alone lacks sensitivity and specificity [3, 19, 20]. Lack of routine virological monitoring in these settings may thus impair or delay the recognition of treatment failure until untoward clinical events or substantial drug resistance occur [21]. In the present study, routine virological monitoring at 6-month intervals was available through the research protocol. At eligibility for this analysis, 62% of children continued to have detectable viremia (≥400 copies/ mL), and 42% met the current WHO definition of virological treatment failure (>5000 copies/mL) despite receiving at least 6 months of HAART, although successful treatment should have achieved virological suppression. Although 6 months was the minimum duration of HAART required for eligibility, most children had been receiving HAART for substantially longer (mean, 2.1 years); in addition, most children had prior treatment experience, with only 30% receiving their first HAART regimen at eligibility. These factors likely contributed to the relatively high rate of observed virological failure at baseline, compared with that in other pediatric cohort studies [14, 15]. The trend for younger age to be associated with greater risk of WHO events did not persist in multivariable analysis but deserves further investigation in future studies.

As expected, CD4-defined immunosuppression was a strong independent predictor for serious HIV-related illness events for children receiving HAART. However, viral load was also independently associated with serious clinical events. Viral load did not simply predict a decline in CD4 T lymphocyte values in the path to clinical disease; rather, viral load predicted clinical illness, even after taking CD4-defined immunosuppression and other factors into account, indicating that viral load combined with CD4 T lymphocyte values performs better than CD4 T lymphocyte values alone for identifying children receiving HAART who are at higher risk for clinical illness. This observation, similar to findings in adults receiving treatment [22- 25], provides additional justification for promoting the availability of viral load monitoring of children receiving HAART in resource-limited settings: routine viral load monitoring added to immunological monitoring would increase the ability of medical providers to identify children receiving HAART at highest risk of clinical complications, to focus efforts to improve medication adherence in those children with virological failure, and to evaluate the need for a new HAART regimen.

Hemoglobin level at eligibility was associated with subsequent WHO events, and most recent hemoglobin level was an independent predictor of the risk of WHO events. Anemia is a well-described risk factor for increased morbidity and mortality in adults and children initiating HAART [17, 26, 27] and in adults already receiving treatment [24]. However, this analysis offers evidence that hemoglobin level is an important predictor of increased risk of HIV-related illness in children already receiving continuous HAART, despite the very low rate of moderate or severe anemia (1.4% with grade 2 anemia or higher). In fact, each 1 g/dL decrease in hemoglobin level was associated with up to a 33% increased risk of experiencing a WHO event. Additional studies are warranted to explore the potential value of hemoglobin monitoring, given that this widely available test may offer predictive value for children receiving HAART even in the absence of frank anemia.

In resource-rich settings where virological monitoring is routine, virological failure is suspected when patients receiving HAART have repeatedly detectable viral load measurements, often defined in studies as ≥400 copies/mL but more commonly defined in clinical guidelines as ≥50–200 copies/mL [1, 2]. The current WHO guidelines recommend using a viral load of >5000 copies/mL as virological confirmation of treatment failure for patients receiving HAART in resource-limited settings where viral load testing is possible [10]. In the present analysis, having a detectable viral load (≥400 copies/mL) was not independently associated with an increased risk of HIV-related illnesses when CD4-defined immunosuppression and hemoglobin level were taken into account, but viral load at the WHO definition of virological failure was independently predictive for identifying children at increased risk of HIV-related illness, even after adjusting for these factors. These results suggest that routine measurements of CD4 T lymphocytes and hemoglobin, combined with ability to detect a viral load >5000 copies/mL, would constitute an effective approach to monitoring children receiving HAART for clinically relevant treatment failure. The ability to detect viral load at a level of >5000 copies/mL, unlike virological detection at thresholds of 50 or even 400 copies/mL, may be more readily attainable with quantitative RNA assays from dried blood spots, making this approach to virological monitoring for treatment failure more feasible in a broad range of resource-limited settings [28].

There are limitations to this observational study. Protocoldefined laboratory monitoring was limited to every 6 months, which may not allow for detection of potential changes in viral load, CD4 T lymphocyte values, and other parameters closer in time to the WHO event. Most HAART regimen changes were preceded by a viral load >5000 copies/mL, suggesting that treating clinicians may have intervened on the basis of the results of protocol-performed viral load testing, potentially leading to an underestimate of the contribution of viral load observed in our study for predicting significant clinical events. Drug resistance testing was not available in this study to inform reasons for virological failure or guide clinician management of patients with treatment failure. Reported WHO-defined events often rely on clinical diagnoses without laboratory confirmation. Although information on HAART was prospectively collected once children were enrolled in the NISDI study, information on HAART used before enrollment was abstracted from medical records and may have been less reliable.

The findings of this analysis support efforts to include routine virological monitoring with assays capable of detecting HIV RNA levels at the WHO-defined virological failure threshold of >5000 copies/mL for children receiving HAART in all settings. At a minimum, twice-yearly monitoring, as done in the NISDI study, would be recommended. The ability to detect virological failure, combined with immunological and other routine clinical monitoring, will serve to identify those children at increased risk of significant HIV-related illnesses and should assist in determining early and appropriate interventions to improve clinical outcomes in children receiving HAART.

Nisdi Pediatric Study Group 2010

Principal investigators, co-principal investigators, study coordinators, data management center representatives, and NICHD staff are listed below.

Brazil. Belo Horizonte—Jorge Pinto (principal investigator) and Flávia Faleiro, Universidade Federal de Minas Gerais. Caxias do Sul—Ricardo da Silva de Souza (principal investigator), Nicole Golin (co-principal investigator), and Sílvia Mariani Costamilan, Universidade de Caxias do Sul/Serviço Municipal de Infectologia. Nova Iguacu—Jose Pilotto (principal investigator), Beatriz Grinsztejn (co-principal investigator), Valdilea Veloso (co-principal investigator), and Gisely Falco, Hospital Geral Nova de Iguacu-HIV Family Care Clinic. Porto Alegre—Ricardo da Silva de Souza (principal investigator), Breno Riegel Santos (co-principal investigator), and Rita de Cassia Alves Lira, Universidade de Caxias do Sul/Hospital Conceição; Ricardo da Silva de Souza (principal investigator), Mario Ferreira Peixoto (co-principal investigator), and Elizabete Teles, Universidade de Caxias do Sul/Hospital Fêmina; Ricardo da Silva de Souza (principal investigator), Marcelo Goldani (co-principal investigator), and Margery Bohrer Zanetello, Universidade de Caxias do Sul/Hospital de Clínicas de Porto Alegre; Regis Kreitchmann (principal investigator) and Debora Fernandes Coelho, Irmandade da Santa Casa de Misericordia de Porto Alegre. Ribeirão Preto—Marisa M. Mussi-Pinhata (principal investigator), Maria Cälia Cervi (co-principal investigator), Márcia L. Isaac (co-principal investigator), and Bento V. Moura Negrini (co-principal investigator), Hospital das Clínicas da Faculdade de Medicina de Ribeirão Preto da Universidade de São Paulo. Rio de Janeiro—Ricardo Hugo S. Oliveira (principal investigator) and Maria C. Chermont Sapia, Instituto de Puericultura e Pediatria Martagão Gesteira); Esau Custodio Joao (principal investigator), Maria Leticia Cruz (co-principal investigator), Plinio Tostes Berardo, and Ezequias Martins, Hospital dos Servidores do Estado. São Paulo—Regina Celia de Menezes Succi (principal investigator) and Daisy Maria Machado, Universidade Federal de São Paulo; Marinella Della Negra (principal investigator), Wladimir Queiroz (co-principal investigator), and Yu Ching Lian (co-principal investigator), Instituto de Infectologia Emilio Ribas.

Mexico. Mexico City—Noris Pavía-Ruz (principal investigator), Patricia Villalobos-Acosta (co-principal investigator), and Dulce Morales-Pérez (co-principal investigator), Hospital Infantil de México Federico Gómez.

Peru. Lima—Jorge Alarcón Villaverde (principal investigator), Instituto de Medicina Tropical “Daniel Alcides Carrión”—Sección de Epidemiologia, Universidad Nacional Mayor de San Marcos (UNMSM); Maria Castillo Díaz (co-principal investigator), Instituto Nacional de Salud del Nin'98o; and Mary Felissa Reyes Vega, Instituto de Medicina Tropical “Daniel Alcides Carrión”—Sección de Epidemiologia, UNMSM.

Data Management and Statistical Center. Yolanda Bertucci, Laura Freimanis Hance, René Gonin, D. Robert Harris, Roslyn Hennessey, Margot Krauss, James Korelitz, Sharon Sothern de Sanchez, and Sonia K. Stoszek, Westat (Rockville, Maryland).

NICHD. Rohan Hazra (principal investigator), Lynne Mofenson, Jennifer Read, George Siberry, and Carol Worrell (Bethesda, Maryland).


Financial support. NICHD (contract HHSN267200800001C, control N01-HD-8-0001).

Potential conflicts of interest. All authors: no conflicts.


1. Working Group on Antiretroviral Therapy and Medical Management of HIV-Infected Children Guidelines for the use of antiretroviral agents in pediatric HIV infection: February 23, 2009. [Accessed 22 March 2010].
2. Panel on Antiretroviral Guidelines for Adults and Adolescents Guidelines for the use of antiretroviral agents in HIV-1-infected adults and adolescents: December 1, 2009. [Accessed 22 March 2010].
3. Mee P, Fielding KL, Charalambous S, Churchyard GJ, Grant AD. Evaluation of the WHO criteria for antiretroviral treatment failure among adults in South Africa. AIDS. 2008;22:1971–1977. [PubMed]
4. World Health Organization Antiretroviral therapy of HIV infection in infants and children: towards universal access—recommendations for a public health approach. 2006. [Accessed 22 March 2010]. [PubMed]
5. Shearer WT, Quinn TC, LaRussa P, et al. Viral load and disease progression in infants infected with human immunodeficiency virus type 1. N Engl J Med. 1997;336(19):1337–1342. [PubMed]
6. Palumbo PE, Raskino C, Fiscus S, et al. Predictive value of quantitative plasma HIV RNA and CD4+ lymphocyte count in HIV-infected infants and children. JAMA. 1998;279(10):756–761. [PubMed]
7. Mofenson LM, Korelitz J, Meyer WA, 3rd, et al. The relationship between serum human immunodeficiency virus type 1 (HIV-1) RNA level, CD4 lymphocyte percent, and long-term mortality risk in HIV-1-infected children. J Infect Dis. 1997;175(5):1029–1038. [PubMed]
8. Hazra R, Stoszek SK, Freimanis-Hance L, et al. Cohort profile: NICHD International Site Development Initiative (NISDI)-a prospective, observational study of HIV-exposed and HIV-infected children at clinical sites in Latin American and Caribbean countries. Int J Epidemiol. 2009;38(5):1207–1214. [PMC free article] [PubMed]
9. World Health Organization WHO case definitions of HIV for surveillance and revised clinical staging and immunological classification of HIV-related disease in adults and children: August 7, 2006. [Accessed 22 October 2008].
10. World Health Organization Antiretroviral therapy for HIV infection in children: towards universal access—recommendations for a public health approach, 2010 revision. 2010. [Accessed 23 July 2010]. [PubMed]
11. Gupta RK, Hill A, Sawyer AW, et al. Virological monitoring and resistance to first-line highly active antiretroviral therapy in adults infected with HIV-1 treated under WHO guidelines: a systematic review and meta-analysis. Lancet Infect Dis. 2009;9(7):409–417. [PubMed]
12. Calmy A, Ford N, Hirschel B, et al. HIV viral load monitoring in resource-limited regions: optional or necessary? Clin Infect Dis. 2007;44(1):128–134. [PubMed]
13. Brady MT, Oleske JM, Williams PL, et al. Declines in mortality rates and changes in causes of death in HIV-1-infected children during the HAART era. J Acquir Immune Defic Syndr. 2010;53(1):86–94. [PMC free article] [PubMed]
14. Sutcliffe CG, van Dijk JH, Bolton C, Persaud D, Moss WJ. Effectiveness of antiretroviral therapy among HIV-infected children in sub-Saharan Africa. Lancet Infect Dis. 2008;8(8):477–489. [PubMed]
15. Ciaranello AL, Chang Y, Margulis AV, et al. Effectiveness of pediatric antiretroviral therapy in resource-limited settings: a systematic review and meta-analysis. Clin Infect Dis. 2009;49(12):1915–1927. [PMC free article] [PubMed]
16. Callens SF, Shabani N, Lusiama J, et al. Mortality and associated factors after initiation of pediatric antiretroviral treatment in the Democratic Republic of the Congo. Pediatr Infect Dis J. 2009;28(1):35–40. [PubMed]
17. Bolton-Moore C, Mubiana-bewe M, Cantrell RA, et al. Clinical outcomes and CD4 cell response in children receiving antiretroviral therapy at primary health care facilities in Zambia. JAMA. 2007;298(16):1888–1899. [PubMed]
18. Smith K, Kuhn L, Coovadia A, et al. Immune reconstitution inflammatory syndrome among HIV-infected South African infants initiating antiretroviral therapy. AIDS. 2009;23(9):1097–1107. [PMC free article] [PubMed]
19. Bagchi S, Kempf MC, Westfall AO, et al. Can routine clinical markers be used longitudinally to monitor antiretroviral therapy success in resource-limited settings? Clin Infect Dis. 2007;44(1):135–138. [PubMed]
20. 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. 2008;49(5):477–484. [PubMed]
21. Vaz P, Chaix ML, Jani I, et al. Risk of extended viral resistance in human immunodeficiency virus-1-infected Mozambican children after first-line treatment failure. Pediatr Infect Dis J. 2009;28(12):e283–e287. [PubMed]
22. Ferry T, Raffi F, Collin-illeul F, et al. Uncontrolled viral replication as a risk factor for non-AIDS severe clinical events in HIV-infected patients on long-term antiretroviral therapy: APROCO/COPILOTE (ANRS CO8) cohort study. J Acquir Immune Defic Syndr. 2009;51(4)(4):407–415. [PubMed]
23. 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. 2008;22(16):2097–2106. [PMC free article] [PubMed]
24. Lundgren JD, Mocroft A, Gatell JM, et al. A clinically prognostic scoring system for patients receiving highly active antiretroviral therapy: results from the EuroSIDA study. J Infect Dis. 2002;185(2):178–187. [PubMed]
25. Grabar S, Le Moing V, Goujard C, et al. Response to highly active antiretroviral therapy at 6 months and long-term disease progression in HIV-1 infection. J Acquir Immune Defic Syndr. 2005;39(3):284–292. [PubMed]
26. Stringer JS, Zulu I, Levy J, et al. Rapid scale-up of antiretroviral therapy at primary care sites in Zambia: feasibility and early outcomes. JAMA. 2006;296(7):782–793. [PubMed]
27. Cross Continents Collaboration for Kids (3Cs4kids) Analysis andWriting Committee Markers for predicting mortality in untreated HIVinfected children in resource-limited settings: a meta-analysis. AIDS. 2008;22(1):97–105. [PubMed]
28. Johannessen A, Garrido C, Zahonero N, et al. Dried blood spots perform well in viral load monitoring of patients who receive antiretroviral treatment in rural Tanzania. Clin Infect Dis. 2009;49(6):976–981. [PubMed]

Articles from Clinical Infectious Diseases: An Official Publication of the Infectious Diseases Society of America are provided here courtesy of Oxford University Press