Substantial progress has been made in the prevention and treatment of HIV among children through the widespread implementation of interventions to prevent mother-to-child transmission of HIV and with the introduction of highly active antiretroviral therapy (HAART). However, although the rate of new infections in children dropped by almost 25% over the last 5 years, an estimated 2.3 million of children were still living with HIV in 2010, 90% of them in sub-Saharan Africa, where prevention of mother-to-child transmission of HIV coverage remains low.1,2 In high-resource countries, declining incidence rates of AIDS and of specific opportunistic infections, as well as changes in the age pattern and in the nature of the most frequent infections in children after HAART have been documented.3–5 By contrast, there are only limited data on the incidence of clinical conditions in HIV-infected children receiving HAART in resource-limited countries. As the pattern of infectious diseases occurring in tropical areas is likely to be different from that observed in Western countries, additional epidemiologic data are needed from African settings on the morbidity affecting HIV-infected children after HAART initiation.
HAART acts by suppressing HIV replication, thereby allowing CD4 restoration, which in turn results in a restored immune protection against infectious pathogens. Consequently CD4-cell count is 1 of the most important determinants of HIV-related morbidity and mortality, and is used in practice for HIV/AIDS classification, in the clinical decision-making process and as a marker of disease progression.6 In children, age and level of immunosuppression at HAART initiation are associated with both long-term CD4 recovery and morbidity.7–10
Together, these considerations underline the need for evaluating the influence of CD4 level and age on opportunistic and other conditions occurring in African children. This work aims to estimate the incidence of AIDS- and non-AIDS–defining diseases, globally, by age and by CD4-strata, in HIV-infected children receiving HAART in sub-Saharan Africa.
Study Population and Setting
The data used for this analysis were pooled from a clinical trial and an observational cohort (respectively, ANRS 12103/12167 “Burkiname” and ANRS 1244/1278 “Programme Enfant Yopougon”).11,12 The Burkiname study was a 2-year, 1-arm, phase 2 clinical trial conducted in Bobo-Dioulasso, Burkina Faso, to assess a once-daily HAART regimen. From February 2006 to November 2007, 51 children, ages 30 months to 15 years, naïve to antiretroviral therapy (ART) except prophylaxis for prevention of mother-to-child transmission of HIV and eligible for ART were included in the study. In addition, 282 children ages 15 months to 15 years, at any stage of HIV infection and with or without ART, were included and followed in Abidjan, Côte d’Ivoire as part of the Programme Enfant Yopougon (PEY) observational cohort from 2000 to 2004. Detailed descriptions of the populations and procedures for both studies have been published elsewhere.11,12 This analysis focused on the incidence of clinical events by CD4-strata during the first 2 years after HAART initiation. HAART was defined as the association of 2 nucleoside reverse transcriptase inhibitors and either 1 protease inhibitor or 1 non-nucleoside reverse transcriptase inhibitor. Thus, children not infected with HIV-1, not receiving HAART, started on HAART >24 months before inclusion in the cohort or with <2 CD4 measurements were excluded.
Classification of Morbidity Events
The 10th revision of the World Health Organization (WHO) International Classification of Diseases-10 was used to classify and label nonspecific causes of morbidity. AIDS-related events were retrospectively classified using the “WHO Clinical staging of HIV for infants and children with established HIV infection” annex published in the 2010 revision of the Recommendations for Antiretroviral Therapy in Infants and Adolescents.6
Clinical and Follow-up Procedures
Follow-up visits were scheduled every month in the Burkiname study and every trimester in the PEY cohort. In both studies, all clinical events (HIV-related or not and including adverse events) were collected at each scheduled visit using standardized procedures and were retrospectively reviewed by an ad hoc committee on a monthly basis. CD4-cell count and percentage and viral load were measured at baseline in the 2 studies and then every 3 months in the Burkiname trial and every 6 months in the PEY cohort. In both studies, all children received cotrimoxazole prophylaxis on a daily basis during their follow-up.
In the Burkiname trial, absolute and percentage CD4-cell count was determined using the FACSCount flow cytometer (Becton Dickinson, BD Biosciences, San Jose, CA). In the PEY study, an automated blood cell counter was used for absolute CD4-cell count determination (MaxM coulter, Beckman, Paris, France) and flow cytometry (FACScan, Becton-Dickinson, Aalst-Erembodegem, Belgium) for the relative percentage of CD4 cells.
Plasma viral loads were determined using plasma HIV-1 RNA assays (HIV-1 RNA 3.0 assay, Quantiplex, East Walpole, MA) in the PEY study until April 2003 then using a real-time reverse transcriptase polymerase chain in both studies. The detection threshold was set at 300 copies/mL (2.5 log10 copies/mL) using 0.2 mL of plasma.
Incidence Rate Calculation
The clinical events (endpoints) considered in this analysis were WHO stage 2 events, WHO stage 3 or stage 4 events, tuberculosis, oral candidiasis, severe bacterial infections, pneumonia, symptomatic malaria infections, mycosis infections other than AIDS-defining mycosis, viral infections with cutaneous manifestations (including varicella-zoster and herpes simplex viruses), acute and infectious diarrhea, otitis media, upper respiratory tract infection (URI) and conjunctivitis. All recurrences were reviewed and validated by the study committee. For the AIDS-related events, only the first of each event type was considered. For bacterial pneumonia and mycosis, recurrences were considered if at least 1 month elapsed between 2 episodes and if they were separated by a visit without this event.
A mixed-effect log-linear model was used to compute the incidence rates (IRs) of each type of event globally, by current CD4 categories (<15%, 15–<25%, ≥25%), and by baseline age categories (≤5 years versus >5 years), and to estimate the slope of incidence trends.13 A random effect was introduced to account for multiple (correlated) events from a same patient.
For the CD4-stratified analysis, modeled CD4 values (see below) rather than observed values were used to compute the time spent in each CD4 stratum because CD4 were measured with error and intermittently.14 To compute the time spent in a CD4 stratum, the total follow-up period was split into 3 months intervals; if the start and the end of the time interval belonged to the same CD4 category, then the whole interval contributed to the time in this category. If the start and the end of the interval differed, then only half of the interval contributed to the time in the category. The time scale considered in the analysis was the time since HAART initiation.
A simple transitional model with 2 states (viral load <1000 copies/mL or “good virologic response” versus viral load ≥1000 copies/mL or “virologic failure”) was used to model the virologic response using a mixed-effect logistic model. The probability of viral load exceeding 1000 copies/mL was related to the previous state and to the time spent in both states.
The percentage of CD4 cells rather than the CD4-cell count was considered in this analysis because, in children, the former remains relatively constant over age.6 The CD4 dynamic over time was modeled using a nonlinear mixed-effect model based on stochastic differential equations. The determinist trend of the CD4 dynamic was represented as a nonlinear asymptotic function of time with a large initial increase in CD4 declining progressively to reach an asymptotic level.15–17 A more formal and detailed description of the CD4 model is given in the Appendix18 (Supplemental Digital Content 1, http://links.lww.com/INF/B386).
A natural statistical framework for such estimation is the Bayesian approach. More specifically, the model alternated between the following steps: 1) estimating the CD4 dynamic conditionally to the virologic response, 2) imputing the CD4 stratum (<15%, 15–<25%, ≥25%) and the time spent in each stratum at each time interval and 3) estimating the event intensity per CD4 stratum. The open source softwares WinBugs and R were used for the estimation.19,20
Both studies received ethical approval from the national ethics committees. The ANRS 12103 study was registered in the WHO international clinical trial database with the number NCT00122538. Legal representatives of children received comprehensive information before inclusion in the study and during the study, and provided written informed consent.
Of the 333 children enrolled in the PEY cohort (n = 282) or in the Burkiname trial (n = 51), 188 were started on HAART, followed up during the first 24 months of HAART and had at least 2 CD4 measurements, and thus were included in this analysis (Fig., Supplemental Digital Content 2, http://links.lww.com/INF/B486).18 The characteristics of the study population at HAART initiation are given in Table 1; compared with children in the Burkiname trial, those in PEY were more likely to have experienced AIDS-related events (P < 0.001) despite having higher CD4% (P < 0.001). They were also more likely to having missing CD4 measurements (32% had 1 measurement missing and 28% 2 versus 8% and 2% in the Burkiname trial). In both studies, mother-to-child transmission of HIV was the predominant mode of infection. All children received cotrimoxazole prophylaxis during their follow-up. No children from the Burkiname and 6 (4%) children from the PEY were lost to follow-up. Overall follow-up in the study was 355 children-years; the median follow-up duration was 22 months (interquartile range = 18–24).
Seven deaths were reported during the follow-up period and 6 of them occurred in children of the PEY cohort. Causes of death were dominated by infections. Of the 23 children who were followed up in the PEY cohort in the first 24 months after HAART initiation but were not included in the present study because they had <2 CD4 measurements, 5 (22%) died (versus 7 deaths in the remaining 188 children, P = 0.002) and 2 (9%) were lost to follow-up.
Change in CD4 -3Over Time
Whereas the estimated asymptotic CD4% level of a child with persistent good virologic response was around 34% (95% confidence interval [CI]: 30–37), it decreased to an estimated 26% for a child experiencing virologic failure during half of his follow-up period and 17% for a child with persistent virologic failure (Fig. 1). The estimated proportions of children with CD4% <15% and ≥25% were respectively 72% (95% CI: 66–78) and 2% (95% CI: 1–5) at baseline, 39% (95% CI: 34–45) and 15% (95% CI: 11–20) at 6 months, 27% (95% CI: 20–33) and 33% (95% CI: 28–38) at 12 months, 18% (95% CI: 14–23) and 46% (95% CI: 41–52) at 24 months.
The estimated probability of virologic success (viral load <1000 copies/mL) increased from 7% (95% CI: 2–15) at HAART initiation to 61% (95% CI: 55–67) at 6 months and remained stable afterwards.
Global and CD4-specific Incidence Rates of Morbidity
The global and CD4-strata specific IRs of the 13 selected clinical events are given in Table 2. The most common infections were URI (100 infections/100 person-years) and infectious diarrhea (37/100 person-years). A linear decrease in the IRs across all CD4% categories was observed for WHO stage 2 events (relative decrease: 0.59, 95% CI: 0.36–0.84), severe bacterial infections (0.68, 95% CI: 0.48–0.96), infectious diarrhea (0.71, 95% CI: 0.54–0.92) and pneumonia (0.69, 95% CI: 0.48–0.98). On the other hand, the incidence rate of WHO stage 3/4 events was significantly higher in the CD4 stratum <15% compared with the 15–<25% and ≥25% strata but not significantly different between the 2 latter strata (rate ratios for CD4 stratum <15% compared with stratum 15–<25%, stratum <15% compared with stratum ≥25% and stratum 15–<25% compared with stratum ≥25% were 3.83 [95% CI: 1.72–9.92], 4.84 [95% CI: 1.96–20.26] and 1.28 [95% CI: 0.36–5.63], respectively). A similar pattern was observed for viral infections (rate ratios for children in stratum <15% compared with those in strata 15–<25% and ≥25% were 3.68 [95% CI: 1.28–12.99] and 2.51 [95% CI: 0.98–8.15], respectively). Tuberculosis was observed mainly among children with CD4 <15%.
A global change in the IRs within CD4% strata was also compared between children who achieved a viral load <1000 copies/mL before 6 months and those who did not. Higher rate of infectious diarrhea and URI were observed in children with lack of virological response (rate ratios were respectively: 1.19, 95% CI: 1.00–1.39 and 1.16, 95% CI: 1.05–1.27). Due to lack of power, it was not possible to test whether this association was homogeneous across CD4% strata.
Change in Incidence Over Time
Incidence trends over time for the selected events along with comparison of the estimated rate ratios during the first and second year after HAART initiation are depicted in Figure 2. A global decrease in the IRs over the 2 years of follow-up was observed for most morbidity endpoints: the estimated IR (per 100 person-years) of WHO stage 2 events decreased from 56.8 (95% CI: 38.2–81.0) at HAART initiation to 11.9 (95% CI: 7.1–17.6) during the second year; WHO stage 3/4 events decreased from 27.5 (95% CI: 14.7–44.6) to 8.2 (95% CI: 4.8–13.4); severe bacterial infection decreased from 39.9 (95% CI: 25.8–61.6) to 16.7 (95% CI: 1.4–23.6); tuberculosis decreased from 10.2 (95% CI: 3.2–21.4) to 1.01 (95% CI: 0.15–3.3); pneumonia decreased from 37.8 (95% CI: 23.0–56.9) to 15.5 (95% CI: 10.3–22.3); infectious diarrhea decreased from 56.7 (95% CI: 38.83–81.77) to 26.7 (95% CI: 19.7–35.3); viral infection decreased from 13.4 (95% CI: 6.4–26.8) to 4.7 (95% CI: 2.3–8.7); otitis decreased from 40.7 (95% CI: 26.2–61.61) to 18.0 (95% CI: 12.3–25.2) and conjunctivitis from 23.1 (95% CI: 13.3–39.9) to 10.5 (95% CI: 6.3–15.9).
By contrast, the incidence of oral candidiasis, URI and mycosis remained stable over time (Fig. 2). Owing to the limited sample size, it was not possible to carry out this analysis stratified by CD4-strata to determine if time changes were homogeneous across all CD4-strata.
Influence of Age on the Incidence Rates
Globally, children <5 years experienced higher rates of infectious diarrhea, URI and otitis compared with older children (rate ratios were 1.85 [95% CI: 1.32–2.70], 1.41 [95% CI: 1.16–1.72] and 1.85, 95% CI: 1.19–2.94, respectively). The analysis stratified by CD4 and age strata showed that this age difference affected IRs more specifically in the CD4 stratum <15% (the same rate ratios in the stratum <15%: 1.89, 95% CI: 1.09–3.13, 1.47, 95% CI: 0.99–2.17 and 3.03, 95% CI: 1.41–6.67, respectively). The incidence rate of WHO 3/4 events was significantly higher in children older than 5 years (rate ratio: 2.49 [95% CI: 1.26–5.28]); however, this difference was no longer observed when results were adjusted for CD4.
This study demonstrated a significant reduction in the incidence of AIDS-defining opportunistic infections, with the exception of oral candidiasis, in the 2 years after HAART initiation in 2 prospective cohorts of African children. In contrast, the rates of URI and skin mycosis remained stable. Although the burden of WHO stage 3/4 events and of viral infections was concentrated in the group of children with CD4% <15%, the IRs of most of the infections decreased linearly across all CD4% strata. The nature of the most common infections varied also with children’s age; as in general population, children <5 years were at higher risk of infectious diarrhea, URI and otitis compared with older children.
Whereas IR estimates of opportunistic and other infections in HIV-infected children receiving HAART have been made available from high-resource countries3,5,10,21,22 and Latin America23; few studies have reported similar results from African settings. Our findings are consistent with previous studies, with rates of WHO stage 2 and WHO stage 3/4 events decreasing over the 24 months of follow-up, and with severe bacterial infection or pneumonia found as the most frequent AIDS-defining infections after HAART initiation.3,10,21,22–26
Furthermore, the rates of WHO stages 2 and 3/4 events estimated during the first and second years after HAART initiation were 2 to 3 times higher compared with those documented in European or American settings (based on the Centers for Disease Control and Prevention classification and not on the WHO classification).10,21,22,27 Likewise, the incidence rates of pneumonia and infectious diarrhea estimated in this study were in the range of the grade 3/4 events reported in the Children with HIV Early Antiretroviral Therapy (CHER) trial28 but 5 to 10 times higher compared with the results from the Pediatric AIDS Clinical Trials Group (PACTG) cohort3 and 2 to 3 times higher compared with the Latin American Eunice Kennedy Shriver National Institute of Child Health and Human Development (NICHD) International Site Development Initiative cohort (NISDI) estimates.23 It should be emphasized that the children included in this study were at more advanced stages of the disease compared with those from other available studies.3,10,21–23,28,29 Despite the great increase in the CD4%, similar to the results reported in other studies,30–32 <50% of our children reached the threshold of 25% after 2 years of treatment.
The risk of AIDS-related infections decreased with increasing CD4, which has also been reported in the PACTG cohort.3 However, different patterns of association between CD4 and IRs were observed; WHO stage 3/4 events were especially observed in the children with CD4%<15% whereas the decline in the incidence of WHO stage 2 events, severe bacterial infections, infectious diarrhea and pneumonia was almost linear across all CD4% strata. It is not known if further reduction in the infection risk could be achieved at CD4 levels higher than 25%. The time before reaching the CD4 stratum ≥25% could be 1 explanation for the lack of difference in the IRs of WHO stage 3/4 between CD4-strata 15–<25% and ≥25%. However, because of the small number of WHO stage 3/4 events in the CD4 stratum ≥25%, it was not possible to test formally this assumption.
In this study, the IRs of common non-AIDS–defining conditions were also investigated. With decreasing mortality and AIDS-related morbidity resulting from the scale-up of ART, these conditions are becoming a growing part of the morbidity burden in children as it has been observed in high-resource settings.25,28 Whereas there was no significant difference in the rate of otitis between CD4-strata, a decrease over time was observed that could be explained by the children growing older.
It is important to note the various limitations of this study to prevent overinterpretation of the results. First, because of the limited sample size, it was not possible to investigate in detail the change in IRs over time by CD4-strata, to stratify the analysis by other covariates or to analyze more specific conditions. Our study population was comprised of 2 distinct populations; the timing of HAART initiation, the monitoring frequency and the follow-up procedures were different between the 2 cohorts, which might have resulted in different levels of support and of accuracy of reporting of clinical events between participants in the 2 studies. However, this reporting bias is less likely for severe events. Besides, although some children received a protease inhibitor and not a non-nucleoside reverse transcriptase inhibitor, this difference in the regimen is likely to have very minor impact on the children outcomes as previously shown.12 Children were included in this analysis only if they were followed for >6 months with at least 2 CD4 measurements while receiving ART, which may have resulted in an underestimation of the incidence of AIDS-defining diseases. Another important challenge in analyzing morbidity in resource-constrained settings is the limited availability of diagnosis procedures. It is therefore likely that the IRs of some infections, such as pneumocystosis or tuberculosis, may have been underestimated while the incidence of pneumonia may have been overestimated and some misclassification bias introduced. However, our diagnosis procedures did not change over time, protecting from a differential misclassification over time. Despite these limitations, we believe that this study provides trends that are important to take into account when planning any further study of the morbidity in children infected with HIV receiving HAART.
In conclusion, we showed a substantial reduction in AIDS- and non-AIDS–defining morbidity among children after HAART initiation, although we also highlighted that the IRs estimated were significantly higher than those observed in high-resource settings. Children will continue to suffer a high burden of infections unless renewed efforts for increasing HAART scale-up are provided.
We are grateful to Joanna Orna-Gliemann for her assistance in editing this article.
The ANRS 12222 Morbidity/Mortality Study Group
Xavier Anglaret, Robert Colebunders, François Dabis, Joseph Drabo, Serge Eholié, Delphine Gabillard, Pierre-Marie Girard, Karine Lacombe, Christian Laurent, Vincent Le Moing and Charlotte Lewden.
Other Representants of Participating Studies
Gérard Allou, Clarisse Amani-Bossé, Divine Avit, Aida Benalycherif, Pierre de Beaudrap, Charlotte Boullé, Patrick Coffie, Ali Coulibaly, Eric Delaporte, Lise Denoeud, Serge Diagbouga, Didier Koumavi Ekouevi, Jean-François Etard, Sabrina Eymard-Duvernay, Patricia Fassinou, Isabelle Fournier-Nicolle, Hervé Hien, Charlotte Huet, Issouf Konate, Sinata Koulla-Shiro, Valériane Leroy, Olivier Marcy, Pierre Régis Martin, Nicolas Meda, Eugène Messou, Albert Minga, Eitel Mpoudi-Ngolé, Philippe Msellati, Boubacar Nacro, Nicolas Nagot, Ibra Ndoye, Thérèse N’Dri-Yoman, Abdoulaye Ouédraogo, Vara Ouk, Men Pagnaroat, Roger Salamon, Vonthanak Saphonn, Olivier Segeral, Catherine Seyler, Besigin Tonwe-Gold, Moussa Traore, Philippe Van de Perre, Ida Viho and Marcel Zannou.
2. Stringer EM, Ekouevi DK, Coetzee D, et al.PEARL Study Team. Coverage of nevirapine-based services to prevent mother-to-child HIV
transmission in 4 African countries. JAMA. 2010;304:293–302
3. Gona P, Van Dyke RB, Williams PL, et al. Incidence of opportunistic and other infections in HIV
in the HAART era. JAMA. 2006;296:292–300
4. Sánchez JM, Ramos Amador JT, Fernández de Miguel S, et al. Impact of highly active antiretroviral therapy
on the morbidity
and mortality in Spanish human immunodeficiency virus-infected children
. Pediatr Infect Dis J. 2003;22:863–867
5. Nesheim SR, Kapogiannis BG, Soe MM, et al. Trends in opportunistic infections in the pre- and post-highly active antiretroviral therapy
eras among HIV
in the Perinatal AIDS Collaborative Transmission Study, 1986-2004. Pediatrics. 2007;120:100–109
7. Dunn DHIV Paediatric Prognostic Markers Collaborative Study Group. . Short-term risk of disease progression in HIV
receiving no antiretroviral therapy or zidovudine monotherapy: a meta-analysis. Lancet. 2003;362:1605–1611
8. Hainaut M, Ducarme M, Schandene L, et al. Age-related immune reconstitution during highly active antiretroviral therapy
in human immunodeficiency virus type 1-infected children
. Pediatr Infect Dis J. 2003;22:62–69
9. Soh CH, Oleske JM, Brady MT, et al.Pediatric AIDS Clinical Trials Group. Long-term effects of protease-inhibitor-based combination therapy on CD4 T-cell recovery in HIV
and adolescents. Lancet. 2003;362:2045–2051
10. Ylitalo N, Brogly S, Hughes MD, et al.Pediatric AIDS Clincial Trials Group Protocol 219C Team. Risk factors for opportunistic illnesses in children
with human immunodeficiency virus in the era of highly active antiretroviral therapy
. Arch Pediatr Adolesc Med. 2006;160:778–787
11. Nacro B, Zoure E, Hien H, et al. Pharmacology and immuno-virologic efficacy of once-a-day HAART in African HIV
: ANRS 12103 phase II trial. Bull World Health Organ. 2011;89:451–458
12. Rouet F, Fassinou P, Inwoley A, et al.ANRS 1244/1278 Programme Enfants Yopougon. Long-term survival and immuno-virological response of African HIV
to highly active antiretroviral therapy
regimens. AIDS. 2006;20:2315–2319
13. Cook RJ, Erdman JN, Dickens BM. Respecting adolescents’ confidentiality and reproductive and sexual choices. Int J Gynaecol Obstet. 2007;98:182–187
14. Tsiatis AA, DeGruttola V, Wulfsohn MS. Modeling the relationship of survival to longitudinal data measured with error. Applications to survival and CD4 counts in patients with AIDS. J Am Statist Assoc. 1995;90:27–37
15. De Beaudrap P, Etard JF, Diouf A, et al.ANRS 1215/90 Study Group. Modeling CD4+ cell count increase over a six-year period in HIV
-1-infected patients on highly active antiretroviral therapy
in Senegal. Am J Trop Med Hyg. 2009;80:1047–1053
16. De Beaudrap P, Rouet F, Fassinou P, et al. CD4 cell response before and after HAART initiation according to viral load and growth indicators in HIV
in Abidjan, Côte d’Ivoire. J Acquir Immune Defic Syndr. 2008;49:70–76
17. Lewis J, Walker AS, Castro H, et al. Age and CD4 count at initiation of antiretroviral therapy in HIV
: Effects on long-term T-cell reconstitution. J Infect Dis. 2012;205:548–556
18. Oksendal BK Stochastic Differential Equations: An Introduction With Applications. 2003 Berlin Springer
19. Spiegelhalter D, Thomas A, Best N, et al. WinBUGS User Manual. 2003 Cambridge, UK Medical Research Council Biostatistics Unit
20. R Development Core Team.R: A Language and Environment for Statistical Computing. 2008 Vienna R Foundation for Statistical Computing
21. Chiappini E, Galli L, Tovo PA, et al. Changing patterns of clinical events in perinatally HIV
during the era of HAART. AIDS. 2007;21:1607–1615
22. Guillén S, García San Miguel L, Resino S, et al.Madrid Group for Research on Pediatric HIV
Infection. Opportunistic infections and organ-specific diseases in HIV
: a cohort study (1990-2006). HIV
23. Alarcón JO, Freimanis-Hance L, Krauss M, et al.NISDI Pediatric Study Group 2011. Opportunistic and other infections in HIV
in Latin America compared to a similar cohort in the United States. AIDS Res Hum Retroviruses. 2012;28:282–288
24. Mulenga V, Ford D, Walker AS, et al.CHAP Trial Team. Effect of cotrimoxazole on causes of death, hospital admissions and antibiotic use in HIV
. AIDS. 2007;21:77–84
25. Viani RM, Araneta MR, Deville JG, et al. Decrease in hospitalization and mortality rates among children
with perinatally acquired HIV
type 1 infection receiving highly active antiretroviral therapy
. Clin Infect Dis. 2004;39:725–731
26. Kourtis AP, Bansil P, Posner SF, et al. Trends in hospitalizations of HIV
and adolescents in the United States: analysis of data from the 1994-2003 Nationwide Inpatient Sample. Pediatrics. 2007;120:e236–e243
27. Gibb DM, Duong T, Tookey PA, et al.National Study of HIV
in Pregnancy and Childhood Collaborative HIV
Paediatric Study. Decline in mortality, AIDS, and hospital admissions in perinatally HIV
-1 infected children
in the United Kingdom and Ireland. BMJ. 2003;327:1019
28. Violari A, Cotton MF, Gibb DM, et al.CHER Study Team. Early antiretroviral therapy and mortality among HIV
-infected infants. N Engl J Med. 2008;359:2233–2244
29. Nachman S, Gona P, Dankner W, et al. The rate of serious bacterial infections among HIV
with immune reconstitution who have discontinued opportunistic infection prophylaxis. Pediatrics. 2005;115:e488–e494
30. Gibb DM, Newberry A, Klein N, et al. Immune repopulation after HAART in previously untreated HIV
. Paediatric European Network for Treatment of AIDS (PENTA) Steering Committee. Lancet. 2000;355:1331–1332
31. Bolton-Moore C, Mubiana-Mbewe 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;298:1888–1899
32. Puthanakit T, Kerr S, Ananworanich J, et al. Pattern and predictors of immunologic recovery in human immunodeficiency virus-infected children
receiving non-nucleoside reverse transcriptase inhibitor-based highly active antiretroviral therapy
. Pediatr Infect Dis J. 2009;28:488–492