Description of viral load measurements (N = 250).

Background
The standard of care for HIV-infected patients on antiretroviral therapy (ART) relies on regular monitoring of HIV viral load. Access to viral load monitoring has improved in recent years, but remains scarce in some settings. Alternatives to plasma, such as dried specimens, can help bring routine viral load testing to remote settings. This study aimed to evaluate performances of the cobas® plasma separation card (PSC) to detect virological failure at the threshold of 1,000 copies/mL in real-life conditions in Vietnam.
Setting
Patients were enrolled in two hospitals in Hanoi (Vietnam).
Methods
This cross-sectional evaluation enrolled 250 HIV-infected adults using convenience sampling. All provided plasma and PSC samples. The sensitivity and specificity of PSC, as compared to plasma, to identify patients in virological failure were estimated. Convenience sampling was used to reach the target numbers of 150 and 100 patients with plasma viral load ≥ and <1,000 copies/mL, respectively.
Results
Overall, 250 patients were enrolled. Of the 127 (50.8%) patients with plasma viral load ≥1,000 copies/mL, the PSC viral load was also ≥1,000 copies/mL in 126 (sensitivity (95% confidence interval (CI)): 99.2% (95.7–99.9)). Of the 123 (49.2%) patients with plasma viral load <1,000 copies/mL, the PSC viral load was also <1,000 copies/mL in 108 (specificity (95% CI): 87.8% (80.7–93.0)). Overall, the concordance rate was 93.6%.
Conclusions
This field evaluation of the cobas® PSC showed its high effectiveness in identifying patients in virological failure. The specificity was 87.8%, lower than in previous reports. However, in cases of discrepancy, the PSC viral load level was never far from the 1,000 copies/mL threshold, which may be due to measurement error inherent to the technique.