Background:  Focusing on HIV treatment, African countries are rolling out national programs for early infant HIV diagnosis, targeting both perinatal follow-up clinics and maternal and child health immunization clinics. Regionalized polymerase chain reaction (PCR) testing is planned but shipment of blood poses logistic challenges. The use of dried blood spots (DBS) simplifies shipment but has not been validated in Uganda.

Methods:  Whole blood specimens were collected from HIV-exposed children (mean age 22.6 weeks; range 6 to 84 weeks) at a perinatal follow-up clinic, a hospital ward, and a maternal and child health clinic. Cell pellets (CP) were prepared for Amplicor DNA (AMPD), DBS for Amplicor DNA, and total nucleic acid (TNA) and plasma for ultra-sensitive HIV p24 antigen (UP24Ag) assay. Results were compared against a “gold-standard.”

Results:  Results with the AMPD and TNA assays on DBS were in complete agreement and had good sensitivity (96.0%) and specificity (94.9%) compared to AMPD on CP; UP24Ag had slightly lower values (90.7% and 94.2%). Between CP and DBS on AMPD, 11 specimens gave discordant results; repeat testing, UP24Ag and TNA results, when available, was in complete agreement with the DBS result on AMPD, suggesting that AMPD on DBS is a better gold-standard than AMPD on CP. Re-analysis using AMPD on DBS as the gold-standard gave much improved values for TNA and UP24Ag.

Conclusions:  In this study, DBS out-performed CP as the specimen of choice for HIV PCR-based testing in terms of both accuracy and convenience; UP24Ag on plasma proved to be an acceptable alternative. The use of DBS PCR or plasma UP24Ag for testing should expedite rapid scale-up of early infant diagnosis in resource-limited settings.