Anaemia and zidovudine-containing antiretroviral therapy in paediatric antiretroviral programmes in the IeDEA Paediatric West African Database to evaluate AIDS.

Introduction: There is a risk of anaemia among HIV-infected children on antiretroviral therapy (ART) containing zidovudine (ZDV) recommended in first-line regimens in the WHO guidelines. We estimated the risk of severe anaemia after initiation of a ZDV-containing regimen in HIV-infected children included in the IeDEA West African database.

Methods: Standardized collection of data from HIV-infected children (positive PCR<18 months or positive serology ≥ 18 months) followed up in HIV programmes was included in the regional IeDEA West Africa collaboration. Ten clinical centres from seven countries contributed (Benin, Burkina Faso, Côte d'Ivoire, Gambia, Ghana, Mali and Senegal) to this collection. Inclusion criteria were age <16 years and starting ART. We explored the data quality of haemoglobin documentation over time and the incidence and predictors of severe anaemia (Hb<7 g/dL) per 100 child-years of follow-up over the duration of first-line antiretroviral therapy.

Results: As of December 2009, among the 2933 children included in the collaboration, 45% were girls, median age was five years; median CD4 cell percentage was 13%; median weight-for-age z-score was-2.7; and 1772 (60.4%) had a first-line ZDV-containing regimen. At baseline, 70% of the children with a first-line ZDV-containing regimen had a haemoglobin measure available versus 76% in those not on ZDV (p ≤ 0.01): the prevalence of severe anaemia was 3.0% (n=38) in the ZDV group versus 10.2% (n=89) in those without (p<0. 01). Over the first-line follow-up, 58.9% of the children had ≥ 1 measure of haemoglobin available in those exposed to ZDV versus 60.4% of those not (p=0.45). Severe anaemia occurred in 92 children with an incidence of 2.47 per 100 child-years of follow-up in those on a ZDV-containing regimen versus 4.25 in those not (p ≤ 0.01). Adjusted for age at ART initiation and first-line regimen, a weight-for-age z-score ≤-3 was a strong predictor associated with a 5.59 times risk of severe anaemia (p<0.01).

Conclusions: Severe anaemia is frequent at baseline and guides the first-line ART prescription, but its incidence seems rare among children on ART. Severe malnutrition at baseline is a strong predictor for development of severe anaemia, and interventions to address this should form an integral component of clinical care.