Abstract
Plasma HIV viral load testing is the preferred means of monitoring antiretroviral treatment response. Dried blood spots (DBSs) hold considerable logistical advantages over EDTA samples, but they more frequently misclassify virological failure and have higher limits of detection (LoD). Plasma separation cards (PSCs) may overcome these limitations. Health workers collected EDTA whole blood by venipuncture and 140 μl of finger-prick blood by capillary tube from 53 HIV-infected adults. Capillary blood was immediately transferred to PSCs. Additionally, 432 EDTA samples from HIV-infected adults were spotted onto PSCs and analyzed together with the finger-prick samples. Specificity and sensitivity of PSC with paired EDTA-PSC samples tested on a cobas 6800/8800 system with the cobas HIV-1 test (cobas HIV) was determined. LoD (3rd HIV-1 WHO International Standard) and stability at a range of temperatures and storage durations was determined using cobas HIV and cobas AmpliPrep/cobas TaqMan HIV-1 test v2.0 (CAP/CTM). Of 132 specimens with quantitative values for paired EDTA-PSC samples, the mean log(10) difference between samples was 0.05 copies/ml (95% confidence interval [CI], -0.01 to 0.11). The LoD for cobas HIV was 790.2 copies/ml and for CAP/CTM was 737.9 copies/ml. At 1,000 copies/ml, PSC sensitivity was 97.0% (128/132) and specificity was 97.2% (343/353). Results correlated well with those from EDTA samples (Deming R(2) = 0.90). PSC results were unaffected by temperature and storage conditions. PSC samples correlate well with plasma viral load and have adequate sensitivity and specificity. The improved performance may be as a result of a reduction in contribution from cell-associated viral nucleic acids. The card provides an alternative sample collection technology to DBSs.