Objectives: In settings where CD4 testing is not available, alternative markers to start paediatric anti-retroviral therapy (ART) could be used. A comprehensive evaluation of these markers has not been performed.
Methods: Prospective cross-sectional study of HIV-infected Malawian children not eligible for ART based on clinical criteria. Associations between CD4 and alternative markers [haemoglobin, total lymphocyte count (TLC), serum albumin, thrombocytes and growth parameters] were analysed, and accuracy of existing and new cut-offs were evaluated.
Results: In all, 417 children were enrolled. Of 261 children aged ≥5 years, 155 (59%) qualified to start ART using CD4. In this group, only TLC was associated with CD4 ( p < 0.001). Sensitivity for TLC was 21% (95% CI: 15–29%), using World Health Organization cut-offs. Improved cut-offs increased sensitivity to 73% (95% CI: 65–80%), specificity 62% (95% CI: 52–72%).
Conclusion: Clinical staging alone is an unreliable strategy to start ART in children. TLC is the only alternative marker for CD4, cut-offs need to be revised though.