20th International AIDS Conference - Melbourne, Australia


TUPE373 - Poster Exhibition

Trends in pediatric characteristics at antiretroviral therapy (ART) initiation, and retention on ART in Swaziland, 2004-2010

C. Azih1, H. Nuwagaba-Biribonwoha2, H. Kamiru3, A.F. Auld4, D. Baughman4, L. Gonzalez3, S. Agolory4, G. Bicego4, P. Ehrenkranz5, V. Okello1

1Swaziland National AIDS Program, Ministry of Health, Government of the Kingdom of Swaziland, Mbabane, Swaziland, 2ICAP, Columbia University, New York, United States, 3ICAP, Columbia University, Mailman School of Public Health, New York, United States, 4United States Centers for Disease Control, Atlanta, United States, 5United States Centers for Disease Control, Mbabane, Swaziland

Background: Swaziland''s national antiretroviral therapy (ART) program was initiated in 2004. Adults and children initiated ART according to World Health Organization (WHO) guidelines, and, when stable (ART adherent and clinically well), were down-referred to lower-level health facilities for follow-up care.
Methods: A retrospective cohort study to assess attrition (documented death, stopping ART, loss-to-follow-up, LTFU defined as no clinic visit in >90 days) was conducted among a nationally representative sample of children initiating ART during 2004-2010. 12/28 clinics initiating children < 15 years(y) on ART were selected using probability-proportional-to-size sampling; medical records were randomly selected for abstraction. Changes in characteristics at ART initiation were assessed using trend tests. Adjusted hazard ratios, AHRs (95% Confidence Intervals) for attrition were estimated using Cox proportional hazards regression models adjusted for multistage sampling and study design.
Results: Among 2,008 pediatric ART enrollees, 30% were < 2y, 20% 2-4y, 32% 5-9y and 18% 10-14y. During 2004-2010, the proportion of ART enrollees < 2y increased from 8% to 34%; the proportion with WHO stage III/IV declined from 90% to 61%; and among children >5y at ART initiation, median CD4+ count increased from 199(IQR 56-273) to 267(IQR 148-369) cells/ µL. Ten percent of the children who initiated ART were down-referred.
At 24 months after ART initiation, 78% remained on ART, 5% had died, and 16% were LTFU, trends over time were not statistically significant.
In a complete-case multivariable analysis, attrition did not change with year of ART initiation. Compared with children 0-2y, children 10-14y had lower attrition: AHR 0.34(0.22-0.53). Compared with children starting ART at WHO stage I/II, children with WHO stage IV had higher attrition: AHR 1.66(1.13-2.45). For every one-unit increase in weight-for-age z-score, risk of attrition decreased 13%: AHR 0.87(0.82-0.92). Down-referral after ART initiation was protective against attrition in univariate analysis: AHR 0.43(0.23-0.82), but not in multivariable analysis: AHR 0.52(0.23-1.20).
Conclusions: Over the evaluation period, more children initiated ART at a younger age and with less advanced HIV disease. The lower attrition observed among down-referred children is probably due to selection of children with fewer risk factors for attrition, but could be a viable strategy for optimizing retention.

Back to the Programme-at-a-Glance