RESUMEN
BACKGROUND: Limited access to HIV testing of children impedes early diagnosis and access to antiretroviral therapy. Our objective was to evaluate the feasibility and acceptability of routine pediatric HIV testing in an urban, fee-for-service, outpatient clinic in Durban, South Africa. METHODS: We assessed the number of patients (0-15 years) who underwent HIV testing upon physician referral during a baseline period. We then established a routine, voluntary HIV testing study for pediatric patients, regardless of symptoms. Parents/caretakers were offered free rapid fingerstick HIV testing of their child. For patients <18 months, the biological mother was offered HIV testing and HIV DNA polymerase chain reaction was used to confirm the infant's status. The primary outcome was the HIV testing yield, defined as the average number of positive tests per month during the routine compared with the baseline period. RESULTS: Over a 5-month baseline testing period, 931 pediatric patients registered for outpatient care. Of the 124 (13%) patients who underwent testing on physician referral, 21 (17%, 95% confidence interval: 11-25%) were HIV infected. During a 13-month routine testing period, 2790 patients registered for care and 2106 (75%) were approached for participation. Of these, 1234 were eligible and 771 (62%) enrolled. Among those eligible, 637 (52%, 95% confidence interval: 49-54%) accepted testing of their child or themselves (biological mothers of infants <18 months). There was an increase in the average number of HIV tests during the routine compared with the baseline HIV testing periods (49 versus 25 tests/month, P = 0.001) but no difference in the HIV testing yield during the testing periods (3 versus 4 positive HIV tests/month, P = 0.06). However, during the routine testing period, HIV prevalence remains extraordinarily high with 39 (6%, 95% confidence interval: 4-8%) newly diagnosed HIV-infected children (median 7 years, 56% female). CONCLUSIONS: Targeted and symptom-based testing referral identifies an equivalent number of HIV-infected children as routine HIV testing. Routine HIV testing identifies a high burden of HIV and is a feasible and moderately acceptable strategy in an outpatient clinic in a high prevalence area.