Cotrimoxazole preventive therapy (CPT) reduces mortality and morbidity in HIV-infected children [1–3]. Cotrimoxazole (CTX) is widely available, inexpensive, well tolerated and can be given to infants; therefore, it has potential for widespread global use. A randomized, controlled study  of daily CPT in HIV-infected Zambian children reported a reduction in mortality by 43%, and hospitalization by 23% . The benefits of daily CPT continued after the trial with further reductions in mortality and hospitalizations occurring with the addition of HAART . The impact of CPT suggests that in resource-limited settings, CTX is not only effective in preventing Pneumocystis pneumonia (PCP) but may also reduce the incidence of bacterial infections [3–7].
Revised guidelines for CPT issued by the WHO promote more liberal and widespread use of daily prophylaxis for HIV-infected children or exposed infants . However, USA guidelines recommend CPT either daily or 3 days per week [9,10]. Data from adult studies suggest that CPT given three times per week is as effective as daily, but with reduced adverse events and increased tolerability . The WHO recommends that all HIV-infected children who begin CPT should continue until at least 5 years of age . Although CTX is generally well tolerated, a schedule with fewer side effects is preferable given the need for prolonged dosing. Moreover, as HAART becomes more widely available, the pill burden and the potential for drug interactions or adverse events in HIV-infected children have increased, with possible advantages for intermittent over daily CPT. However, the optimal frequency for CPT has not been studied in children. There is no information on the efficacy or tolerability of different prophylactic regimens, especially in resource-limited settings. Furthermore, there is little published data describing survival and morbidity among African infants receiving CPT. We, therefore, investigated the efficacy and tolerability of three times a week compared with daily CPT in South African HIV-infected infants and children.
A prospective study comparing intermittent CPT with daily CPT in HIV-infected children was performed in Cape Town, South Africa. The study had a factorial design, with simultaneous comparison of three times weekly vs. daily CPT, and isoniazid (INH) vs. placebo. Children were randomized to one of four study arms: daily CPT with INH, or daily CPT with placebo; thrice weekly CPT with INH, or thrice weekly CPT with placebo. The study commenced in December 2002. The placebo arm was terminated on 17 May 2004 on the recommendation of the data safety monitoring board (DSMB) due to significantly higher survival in the INH groups . Thereafter, all children receiving placebo were switched to INH. We continued to investigate the efficacy and tolerability of thrice-weekly CPT and INH compared with daily CPT and INH. On the advice of the DSMB, INH was discontinued in December 2007, as most children had then commenced HAART. Results are reported from January 2003 through December 2007.
The study population comprised HIV-infected children aged 8 weeks or older, attending either Red Cross War Memorial Children's Hospital at the University of Cape Town or Tygerberg Hospital at Stellenbosch University. Additional inclusion criteria were weight of 2.5 kg or more, access to transport and informed consent from a parent or legal guardian. Exclusion criteria were chronic diarrhoea, prior hypersensitivity reaction to INH or to sulphur drugs, severe anaemia, neutropaenia, thrombocytopaenia or nonreversible renal failure . Children who required or were already receiving INH preventive therapy as per national tuberculosis guidelines were also excluded. Children on HAART were eligible for enrolment if they had been stable on treatment for 2–3 months. The study was approved by the Research and Ethics Committees of the Faculty of Health Sciences of the Universities of Cape Town and Stellenbosch, South Africa. The trial was conducted in accordance with the Declaration of Helsinki and is registered as Clinical Trials NCT00330304.
A history and physical examination were performed, and sociodemographic, clinical and laboratory data collected at enrolment. Children were seen 4 weekly for the first 6 months, then 6 weekly for the next 6 months and thereafter every 2–3 months, depending on medical and social circumstances.
Randomization was stratified by site to accommodate potential differences between study site populations and for the benefit of precision. Study pharmacists labelled the trial drugs with sequential study numbers according to variable blocked randomization lists prepared by the trial statistician. At enrolment, children were allocated a sequential study number and randomly assigned to one of the four treatment combinations by the pharmacist, according to the randomization list. Study investigators and caregivers were blinded to the INH/placebo allocation; drugs were dispensed by the study pharmacist.
Children were randomized to receive CTX either daily or thrice weekly on Monday, Wednesday and Friday. CTX (5 mg/kg per dose of the trimethoprim component) was given until 12 months of age, when HIV clinical and immune staging was reassessed; it was then continued in those with Centers for Disease Control and Prevention (CDC) clinical category B or C disease, in those with severe immunological impairment and in those with a prior episode of PCP. The INH dosage was 10 mg/kg per day with a variability of 8–12 mg/kg depending on whether half or quarter tablets were required. Placebo had an identical appearance to INH tablets (BE-tabs Pharmaceuticals, Johannesburg, South Africa). INH or placebo was given according to the frequency of the CPT schedule.
Multivitamin supplementation and immunizations were given according to a standard protocol. HAART was not widely available at the start of the study, but became increasingly accessible as the antiretroviral rollout proceeded in South Africa. Local guidelines for paediatric HAART were developed during the study period, enabling children to access such therapy based on medical and social criteria.
HIV status was assessed at enrolment by ELISA (Abbott AxSYM HIV antibody/antigen ELISA; Abbott Laboratories, Abbott Park, Illinois, USA) in those older than 15 months of age and by PCR (Amplicor HIV-1; Roche Diagnostic Systems, Indianapolis, Indiana, USA) in younger children. Blood was taken at enrolment and 6 monthly for a full blood count (FBC), renal function, alanine transaminase (ALT), CD4 cell count and percentage. A chest radiograph was performed at baseline and 6 monthly. A screening tuberculin skin test (2 TU RT23; Staten Serum Institute, Copenhagen, Denmark) was done on enrolment and 6 monthly if negative. Tests were performed more frequently if clinically indicated. In addition, for patients on HAART, a FBC and ALT were performed 1 and 3 months after randomization. A detailed history, clinical examination and laboratory tests as indicated were done if an intercurrent hospitalization or illness occurred. Blood cultures were obtained when indicated by hospital staff, independent of the study investigators and according to standard, written hospital protocols. Antibiotic sensitivity testing was done in accordance with standard local laboratory procedures. Hospital or clinic records were obtained wherever possible for children who died; in the absence of these, a verbal autopsy was performed when feasible.
Children were hospitalized at one of the two study hospitals, or at affiliated secondary level hospitals in the same academic complex. Decisions regarding admission, clinical in-patient management and discharge were taken by hospital staff independent of study investigators. Study staff documented clinical information and results of all in-hospital investigations during hospitalization, and at discharge.
Clinical or laboratory events were graded 1 to 4 according to the Division of AIDS toxicity criteria . A grade 3 or 4 reaction was considered a significant adverse event and managed according to a standard protocol.
The primary outcome was mortality; secondary outcomes included bacteraemia, hospitalizations and Grade 3 or 4 serious adverse events. All analyses were by intention to treat. The Kaplan–Meier method was used to analyse time to death and comparisons were made with the log-rank test. Cox proportional hazards regression was used to estimate hazard ratios . Proportional hazards assumptions were checked with the Grambsch–Therneau test . As randomization was stratified by site, all survival analyses included site as a main effect. Subgroups comparisons for CDC disease severity, age, site and presence of HAART at baseline were done to assess the consistency of the intervention effect . Cause-specific hazard ratios (CHRs) were calculated within a competing risks framework . The incidences of bacteraemia, hospitalization and serious adverse events were compared between groups with incidence rate ratios (IRRs). Incidence of death was compared between infants and older children based on age at randomization. Person-time was calculated in years from enrolment to last known time alive. For hospitalization rates, follow-up time was limited to days not spent in hospital. Prolonged admissions were defined as longer than 3 days. The Wilcoxon rank-sum test was used to compare median follow-up time of the two groups . Anthropometric measurements were standardized with reference to the National Center for Health Statistics standards; z-scores were calculated using Epi Info version 6.0 (CDC, Druid Hills, Georgia, USA) . Data were analysed using Stata statistical software, version 10 (StataCorp LP, College Station, Texas, USA) . All P values are two-tailed.
The primary outcome was mortality. We aimed to demonstrate noninferiority comparing the survival proportions of thrice-weekly CPT (test treatment) to daily CPT (active control). The choice of inferiority margin (Δ = −0.1) was based on expert opinion. We aimed for a sample size of 400 patients, primarily based on power considerations for the INH/placebo comparison ; thus with 200 patients in each group, the upper limit of the observed one-sided 95% confidence interval (CI) was expected to exceed −0.100 with 95% power given a standard survival proportion of 0.900 and test expected survival proportion of 0.900. However, given the significant survival benefit for the INH compared with the placebo group, study enrolment discontinued earlier than planned. Our total sample size is, therefore, 324 patients. Using a one-sided 95% CI approach, three times weekly CPT would be considered noninferior to daily CPT if the CI for difference in survival proportions included zero and exceeded the predefined delta of −0.1 estimated at 1 year of follow-up .
Role of the funding source
The study sponsors had no role in the study design, collection, analysis or interpretation of data, writing of report or in the decision to submit the paper for publication.
Three hundred and thirty-nine children were enrolled in the study; 15 children were excluded from analysis (10 tested HIV-negative, five were lost to follow-up within a month after randomization), Appendix 1, http://links.lww.com/QAD/A63. Three hundred and twenty-four children were followed for 672.1 child-years.
Baseline characteristics of the children stratified by thrice-weekly CPT compared with daily CPT were similar (Table 1). Out of the 324 children, 165 (50.9%) were assigned to intermittent therapy and 159 to daily therapy. The median age of children at enrolment was 23 months [interquartile range (IQR) 9.5–48.6 months]; almost one-third were younger than 12 months. Most children (287, 88.6%) were symptomatic, either CDC clinical category B or C. The median CD4 cell percentage was 20%; the proportion of moderately or severely immunosuppressed children was similar in both groups. Malnutrition was common in both groups [median weight for age z-score equal to −1.5 (IQR −1.8 to −1.2) and median weight for height z-score equal to −0.2 (IQR −0.4 to 0.0)]. At enrolment, 28 (9%) children were on HAART, whereas 179 (55.3%) commenced HAART during the study. The number of children who received HAART during the trial was similar in the intermittent (104, 63.0%) and daily (103, 64.8%) groups.
Follow-up and adherence
Twenty-eight (9%) children were lost to follow-up; 16 out of 28 (57%) were in the daily group. Similar numbers of children withdrew from the study [18 (11%) of the intermittent vs. 20 (13%) of the daily group]. The most commonly cited reason for withdrawal was logistical difficulty for the caregiver (28, 9%), in particular relocation (18, 6%). The median follow-up time was similar in the intermittent and daily groups [1.97 years (IQR 1.3–3.3) vs. 1.92 years (IQR 0.5–3.29), P = 0.37]. Adherence to study medication was excellent and similar in the two dosing schedules .
Effect on mortality
The mortality rate (53 deaths, 16%) was similar in the intermittent group and the daily group [24 deaths in 165 children (14%) vs. 29 deaths in 159 children (18%), hazard ratio 0.75 (95% CI 0.44–1.29)] (Fig. 1). The difference in the cumulative survival proportions estimated at 1 year was 0.04 (90% CI −0.03 to 0.10), and thus noninferiority of the intermittent to the daily dosage can be concluded for mortality. There was no evidence of heterogeneity in the group mortality difference across categories of CDC clinical staging (P = 0.91), age (P = 0.42), CD4 category (P = 0.54), site (P = 0.51) or use of HAART at enrolment (Table 2).
Infants had a six-fold higher incidence of death compared with those older than 1 year [20 deaths per 100 child-years vs. 3.6 deaths per 100 child-years, IRR 5.91 (95% CI 3.3–11.2) P < 0.0001]. Infant death rates were similar comparing intermittent to daily CPT (IRR 0.84, 95% CI 0.41–1.73).
In most children (46/53, 87%), the cause of death could be reliably determined. Causes of death were similar in the two groups. Overall, 17 (32%) deaths were ascribed to clinical sepsis [seven deaths in the intermittent group vs. 10 deaths in the daily group, CHR 0.69 (95% CI 0.26–1.83)]; 13 (25%) died from pneumonia (nine vs. four deaths, CHR 2.16 (95% CI 0.67–6.89)] and eight (15%) died from gastroenteritis [three vs. five deaths, CHR 0.56 (95% CI 0.13–2.38)]. None had confirmed PCP. Four (8%) deaths were due to cardiac failure. Chronic renal failure, HIV encephalopathy with respiratory depression and Burkitt's lymphoma occurred in a single case each. One child with depression committed suicide at home. In seven (13%) children, the cause of death could not be ascertained; five demised at home, whereas two occurred en route to a local day health facility.
Forty-seven incidents of bacteraemia occurred in 30 children; eight children had more than one episode. Streptococcus pneumoniae was the predominant Gram-positive pathogen (15 infections, 32%); just over half of the infections were due to Gram-negative organisms, mostly Enterobacteriaceae (20, 42%), especially Klebsiella sp. (7, 15%). Overall, children in the intermittent group had a two-fold increased risk of bacteraemia compared with those in the daily group [Incidence rate (IR) 9.6 per 100 child-years vs. IR 4.07 per 100 child-years; incidence rate ratio (IRR) 2.36 (95% CI 1.21–4.87), P = 0.006]. This increased risk was mostly due to a higher rate of Gram-negative bacteraemia [20 infections vs. five infections, IRR 3.61 (95% CI 1.31–12.3)].
Thirty bacterial isolates were tested for sensitivity to CTX, of which 25 (83%) were CTX resistant. Overall, there was a marginally higher proportion of CTX-resistant organisms in the intermittent group than in the daily group [19/21 (90%) vs. 6/9 (66.7%)]. Only two of 15 S. pneumoniae isolates were tested against CTX; both were resistant. Of the 20 Enterobacteriaceae, 18 were tested against CTX, of which 13 (72%) were resistant. There were markedly more CTX-resistant Enterobacteriaceae isolates from the intermittent group than the daily group [12/14 (86%) vs.1/4 (25%)]. Seven of these (39%) were extended spectrum B-lactamase producers (six in the intermittent group, one in the daily group). Of the S. aureus isolates, all five had antimicrobial sensitivities done, all of which were resistant to CTX. Four (80%) S. aureus isolates were methicillin resistant (two in each group).
Over the 5-year period, there were 448 hospitalizations (Table 3). Approximately half, 186 (57%) of the children had one or more hospitalization. The admission rate (number of admissions) was similar comparing the intermittent group to the daily group [IRR 0.97 (95% CI 0.81–1.18)]. The most common diagnosis at hospitalization was pneumonia, followed by diarrhoea or clinical sepsis. Children taking intermittent CPT spent significantly more days in hospital than those on daily therapy [228.5 days per 100 child-years vs. 198.5 days per 100 child-years, IRR 1.15 (95% CI 1.04–1.28), P = 0.004].
The overall incidence of grade 3 or 4 toxicity was 6.8 events per 100 child-years (46 events). These occurred with equal frequency in the intermittent group compared to the daily group [25 (54%) vs. 21 (46%), IRR 1.07 (95% CI 0.58–2.02)]. Haematological events, including neutropaenia, thrombocytopaenia or anaemia, occurred most commonly (Table 4). Alternative causes of haematological events included infections, other drugs and HIV disease. A single child in the daily group developed Stevens–Johnson syndrome, requiring discontinuation of CTX.
In this study, intermittent and daily CPT had a similar effect on mortality and tolerability in young HIV-infected children. However, intermittent CPT was associated with more bacteraemia and longer hospitalization. The effect of daily CPT in reducing mortality has been clearly shown in HIV-infected Zambian children not on HAART . This effect was not explained by the prevention of PCP. In the context of high exposure to bacterial pathogens, prevention of bacterial infections could also account for such protection . Consistent with this, a cohort study  in Cote d'Ivoire reported a reduction in severe illness in HIV-infected infants taking daily CPT. The benefits of daily CPT are sustained with the addition of HAART, reducing hospitalization and mortality among HIV-infected children [3,22]. In our cohort of children, mortality and morbidity was significantly higher among infants, demonstrating the need for early HAART in addition to CPT, as has been recently recommended .
Although the intermittent group experienced more bacteraemia, mortality was similar to the daily group. Good access to care, with timely hospitalization and prompt treatment with antibiotics, may have affected outcome, preventing death in children with bacteraemia. Nevertheless, in areas where access to care is suboptimal, bacteraemia may be associated with more severe illness and death . Therefore, a regimen reducing the incidence of bacteraemia is preferable.
The area in which this study was conducted has high levels of CTX resistance. As evidence of this, nasopharyngeal cultures obtained at study enrolment showed CTX resistance in more than 80% of colonizing bacteria, as reported elsewhere . Moreover, almost 90% of blood culture isolates were CTX resistant. Therefore, isolates are not routinely tested for CTX sensitivity nor is CTX used as a first-line antibiotic except for PCP. Nevertheless, daily CPT was associated with a significant reduction in bacteraemia including that due to Gram-negative, CTX-resistant pathogens. This apparent discrepancy between high in-vitro resistance of colonizing or pathogenic organisms, yet prevention of severe disease, has been observed elsewhere [1,2,4,5]. Possible mechanisms include differences between in-vivo and in-vitro bacterial susceptibility, or complex interplay of other organisms that may affect colonization or bacteraemia. Recent studies support a role for CPT in preventing nasopharyngeal colonization with CTX-sensitive pneumococci . In addition, CPT may be effective against preventing invasive infection with methicillin-resistant S. aureus , a pathogen commonly found in HIV-infected children .
Investigators and clinicians were not blinded to the dosing schedule allocation, a limitation of the study. However, the main outcome of mortality is an objective measure, not subject to investigator bias. Furthermore, decisions regarding admission, discharge and the indication for blood cultures were taken by independent hospital clinicians not involved in the study. The higher rates of bacteraemia and hospitalization seen in children taking intermittent CPT are thus likely to be a reflection of more severe illness in this group.
Our realized sample size is smaller than the a priori specified sample size; although this decreased precision, we achieved adequate power to declare noninferiority for the primary endpoint of mortality. In addition, the group effect on hospitalization and bacteraemia incidence was highly significant, indicating adequate power for these comparisons. It is possible that we are unable to demonstrate small differences in toxicity; however, given the increased risk of bacteraemia and hospitalization associated with intermittent CPT, and the overall low rate of serious adverse events, such differences are unlikely to affect clinical guidelines.
This study was done in a hospital setting, with meticulous follow-up of children and good access to care. Adherence to study medication in both groups was excellent, as reported . Our results may not be generalizable to settings where adherence to CPT and follow-up could be lower. However, this strengthens the recommendation for daily CPT, as even if adherence to this regimen is suboptimal, the impact on survival occurred when CPT was taken intermittently. The tolerability of both regimens was comparable and excellent.
Our results might not apply equally to settings with different burdens of invasive bacterial disease. Children with limited exposure to bacterial disease could possibly benefit equally from daily and intermittent therapy. However, as the predominant burden of paediatric HIV is in resource-limited, high bacterial burden countries, our results remain globally relevant. Moreover, our findings may especially be applicable in subtropical Africa, where CPT has the potential to further reduce morbidity and mortality by preventing nontyphoid Salmonella or malaria infections .
On balance, these results support current WHO recommendations for daily CPT in HIV-infected infants and children. Daily CPT is an effective, well tolerated, available and cost-effective intervention for reducing morbidity and mortality in HIV-infected children . Widespread implementation of CPT is needed in areas of sub-Saharan Africa where this intervention is still not available.
Funding for the study was provided by the Rockefeller Foundation, USA, the MRC South Africa, the National Research Foundation, South Africa, the South African Thoracic Society and the Department of Health, South Africa. We thank the children and their caregivers for participating.
H.Z. and M.C. conceived the study, wrote the protocol and grant for funding and supervised the study. L.W. was responsible for data management, maintained the database and assisted with data analysis. S.M. le Roux was a trial physician and assisted with data analysis. T.J., C.M. and N.J. were part of the clinical study team and contributed to data collection. S.S. assisted with design, supervision and coordination of study. A.B.-L. was the trial pharmacist and did adherence measures. S.M. le Roux assisted with data collection and analysis. C.L. contributed to study design, was responsible for the randomization list and did statistical analysis. H.Z. was primarily responsible for drafting the paper and did the initial data analysis. All authors contributed to and approved the final manuscript.
The INH study team: H. Bezuidenhout, P. Brink, L. Frigati, D. Gray, G. Hussey, H. Rabie, R. Streicher and E. Walters.
Data and Safety Monitoring Committee: Dr J. Kaplan (chair), Dr W. El Sadr, Professor P. Donald and Professor N. Beyers; local DSMB: Professor P. Donald (chair), Professor N. Beyers and Professor M. Klein.
There are no conflicts of interest.
1. Chintu C, Bhat GJ, Walker AS, Mulenga V, Sinyinza F, Lishimpi K, et al. Co-trimoxazole as prophylaxis against opportunistic infections in HIV-infected Zambian children (CHAP): a double-blind randomised placebo-controlled trial. Lancet 2004; 364:1865–1871.
2. 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. AIDS 2007; 21:77–84.
3. Walker AS, Mulenga V, Ford D, Kabamba D, Sinyinza F, Kankasa C, et al. The impact of daily cotrimoxazole prophylaxis and antiretroviral therapy on mortality and hospital admissions in HIV-infected Zambian children. Clin Infect Dis 2007; 44:1361–1367.
4. Anglaret X, Chene G, Attia A, Toure S, Lafont S, Combe P, et al. Early chemoprophylaxis with trimethoprim–sulphamethoxazole for HIV-1-infected adults in Abidjan, Cote d'Ivoire: a randomised trial. Cotrimo-CI Study Group. Lancet 1999; 353:1463–1468.
5. Mermin J, Lule J, Ekwaru JP, Malamba S, Downing R, Ransom R, et al. Effect of co-trimoxazole prophylaxis on morbidity, mortality, CD4-cell count, and viral load in HIV infection in rural Uganda. Lancet 2004; 364:1428–1434.
6. Ruffini DD, Madhi SA. The high burden of Pneumocystis carinii pneumonia in African HIV-1-infected children hospitalized for severe pneumonia. AIDS 2002; 16:105–112.
7. Zar HJ, Dechaboon A, Hanslo D, Apolles P, Magnus KG, Hussey G. Pneumocystis carinii pneumonia in South African children infected with human immunodeficiency virus. Pediatr Infect Dis J 2000; 19:603–607.
8. World Health Organization. Guidelines for cotrimoxazole prophylaxis for HIV-related infection among children, adolescents and adults in resource-limited settings. Recommendations for a public health approach
. Geneva, Switzerland: WHO; 2006. http://www.who.int/hiv/pub/guidelines/ctx/en/index.html
. [Accessed 24 November 2009]
9. American Academy of Pediatrics. Antimicrobial prophylaxis.
In: Pickering LK, Baker CJ, Kimberlin DW, Long SS, editors. Red Book: 2009 report of the Committee on Infectious Diseases
, 28th ed. Elk Grove Village, Illinois, USA: American Academy of Pediatrics; 2009. pp. 536–540. http://aapredbook.aappublications.org
. [Accessed 24 November 2009].
10. Centers for Disease Control and Prevention. Guidelines for the Prevention and Treatment of Opportunistic Infections Among HIV-Exposed and HIV-Infected Children. Recommendations from CDC, the National Institutes of Health, the HIV Medicine Association of the Infectious Diseases Society of America, the Pediatric Infectious Diseases Society, and the American Academy of Pediatrics. MMWR Recomm Rep 2009; 45–50.
11. El-Sadr WM, Luskin-Hawk R, Yurik TM, Walker J, Abrams D, John SL, et al. A randomized trial of daily and thrice-weekly trimethoprim–sulfamethoxazole for the prevention of Pneumocystis carinii pneumonia in human immunodeficiency virus-infected persons. Terry Beirn Community Programs for Clinical Research on AIDS (CPCRA). Clin Infect Dis 1999; 29:775–783.
12. Zar HJ, Cotton MF, Strauss S, Karpakis J, Hussey G, Schaaf HS, et al. Effect of isoniazid prophylaxis on mortality and incidence of tuberculosis in children with HIV: randomised controlled trial. BMJ 2007; 334:136.
14. Cook T, DeMets D. Introduction to statistical methods for clinical trials. Chapman & Hall/CRC; 2008.
15. Grambsch P, Therneau T. Proportional hazards tests and diagnostics based on weighted residuals. Biometrika 1994; 81:515–526.
16. Fine JP, Gray RJ. A proportional hazards model for the subdistribution of a competing risk. J Am Stat Assoc 1999; 94:496–509.
17. Epi Info version 6. Database and statistical software for public health professionals. Centers for Disease Control and Prevention (CDC); 2008.
18. Stata statistical software: release 10.0. College Station, Texas: Stata Corporation; 2007.
19. Piaggio G, Elbourne DR, Altman DG, Pocock SJ, Evans SJ. Reporting of noninferiority and equivalence randomized trials: an extension of the CONSORT statement. JAMA 2006; 295:1152–1160.
20. le Roux SM, Cotton MF, Golub JE, le Roux DM, Workman L, Zar HJ. Adherence to isoniazid prophylaxis among HIV-infected children: a randomized controlled trial comparing two dosing schedules. BMC Med 2009; 7:67.
21. Harambat J, Fassinou P, Becquet R, Toure P, Rouet F, Dabis F, et al. 18-month occurrence of severe events among early diagnosed HIV-infected children before antiretroviral therapy in Abidjan, Cote d'Ivoire: a cohort study. BMC Public Health 2008; 8:169.
22. Fassinou P, Elenga N, Rouet F, Laguide R, Kouakoussui KA, Timite M, et al. Highly active antiretroviral therapies among HIV-1-infected children in Abidjan, Cote d'Ivoire. AIDS 2004; 18:1905–1913.
23. Violari A, Cotton MF, Gibb DM, Babiker AG, Steyn J, Madhi SA, et al. Early antiretroviral therapy and mortality among HIV-infected infants. N Engl J Med 2008; 359:2233–2244.
24. Berkley JA, Lowe BS, Mwangi I, Williams T, Bauni E, Mwarumba S, et al. Bacteremia among children admitted to a rural hospital in Kenya. N Engl J Med 2005; 352:39–47.
25. Cotton MF, Wasserman E, Smit J, Whitelaw A, Zar HJ. High incidence of antimicrobial resistant organisms including extended spectrum beta-lactamase producing Enterobacteriaceae and methicillin-resistant Staphylococcus aureus in nasopharyngeal and blood isolates of HIV-infected children from Cape Town, South Africa. BMC Infect Dis 2008; 8:40.
26. Gill CJ, Mwanakasale V, Fox MP, Chilengi R, Tembo M, Nsofwa M, et al. Effect of presumptive co-trimoxazole prophylaxis on pneumococcal colonization rates, seroepidemiology and antibiotic resistance in Zambian infants: a longitudinal cohort study. Bull World Health Organ 2008; 86:929–938.
27. Pappas G, Athanasoulia AP, Matthaiou DK, Falagas ME. Trimethoprim–sulfamethoxazole for methicillin-resistant Staphylococcus aureus: a forgotten alternative? J Chemother 2009; 21:115–126.
28. Ryan M, Griffin S, Chitah B, Walker AS, Mulenga V, Kalolo D, et al. The cost-effectiveness of cotrimoxazole prophylaxis in HIV-infected children in Zambia. AIDS 2008; 22:749–757.
© 2010 Lippincott Williams & Wilkins, Inc.