Paediatric HIV in central Sudan: High sero-prevalence and poor performance of clinical case definitions

Background: Paediatric clinical case definitions (CCDs) for the human immunodeficiency virus (HIV) have been proposed as screening tools in resource-limited countries. Objectives: We assessed the performance of the World Health Organisation CCD (WHO-CCD), the Bloemfontein CCD (B-CCD) and a locally modified version of the Bloemfontein CCD (MB-CCD) in comparison with HIV serology in acutely hospitalised children aged 1.5-14 years. We also determined the HIV sero-prevalence among this group of children. Study design: A cohort of 106 consecutive acute paediatric admissions to a major teaching hospital in central Sudan was recruited over a 3-month period. Results: The WHO-CCD, B-CCD, and MB-CCD were relatively specific with estimates of 96.0% (95% confidence interval [CI] 90.1-98.9), 88.0% (95% CI 80.0-93.6), and 74.0% (95% CI 64.3-82.3), respectively. However, corresponding sensitivities were poor with estimates of 16.7% (95% CI 0.4-64.1), 33.3% (95% CI 4.3-77.7), and 66.7% (95% CI 22.3-95.7), respectively. The HIV sero-prevalence was high at 5.7% (95% CI 2.1-11.9). Conclusions: CCDs performed poorly against HIV serology in acutely hospitalised children aged 1.5-14 years in central Sudan and, therefore, we advocate improving access to serological diagnostic tools. The high HIV sero-prevalence rate among this group of children poses serious challenges to policy makers and warrants further research.