Background: HIV infection has become a common risk factor for hospital admission and a major contributor to childhood morbidity in South Africa. There remains a paucity of data describing the cost of hospitalisation of HIV-infected children in South Africa. The aim of this study was to describe basic demographics and clinical patterns as well as cost implications of the hospitalisation of HIV-infected children in the Karl Bremer Hospital, Cape Town, South Africa.

Methods: A prospective descriptive longitudinal study of HIV-positive paediatric admissions, matched with HIV-negative controls, was conducted. Patients were matched according to age, socio-demographic area and presenting symptoms. Questionnaires were used to elicit demographic and clinical information. Worksheets were used to record any costs incurred, which were calculated at rates applicable to 2001. This was done daily during admission. Data was statistically analysed in MS Excel and MS Access. Thirty HIV-positive children were identified, of which 23 could be matched with 23 HIV-negative children. HIV-positive children had a higher admission rate (2.09 versus 0.26 previous admissions, p = 0.000) and were also younger at the time of first admission to hospital (7.52 versus 13.78 months, p = 0.005). There is a statistically significant difference in duration of hospitalisation in the HIV-positive group when compared to the control group—duration of hospitalisation being longer in the HIV-positive group (7.91 versus 4.96 days, p = 0.005). Despite being treated for the same condition, there is a statistically significant difference in the cost incurred by children in the HIV-positive group (R6 203.16) when compared to the HIV-negative group (R3 901.96); p = 0.000.

Conclusion: This study shows a clear and statistically significant difference between the HIV-positive group and HIV-negative control group of children with regard to admission rate, age at first admission, duration of hospitalisation and cost incurred during hospitalisation. HIV-infected children in the pre-HAART (highly active anti-retroviral therapy) era were hospitalised more frequently and for longer periods than their HIV-uninfected counterparts. These findings seem to suggest that the cost of hospitalising HIV-positive children is significantly more than HIV-negative controls, which will increase the financial burden on already restricted health resources.