Pooled nucleic acid testing to detect antiretroviral treatment failure in Mexico

Similar to other resource-limited settings, cost restricts availability of viral load monitoring for most patients receiving antiretroviral therapy in Tijuana, Mexico. We evaluated if a pooling method could improve efficiency and reduce costs while maintaining accuracy.

We evaluated 700 patient blood plasma specimens at a reference laboratory in Tijuana for detectable viremia, individually and in 10 × 10 matrix pools. Thresholds for virologic failure were set at ≥500, ≥1000 and ≥1500 HIV RNA copies per milliliter. Detectable pools were deconvoluted using pre-set algorithms. Accuracy and efficiency of the pooling method were compared with individual testing. Quality assurance (QA) measures were evaluated after 1 matrix demonstrated low efficiency relative to individual testing.

Twenty-two percent of the cohort had detectable HIV RNA (≥50 copies/mL). Pooling methods saved approximately one third of viral load assays over individual testing, while maintaining negative predictive values of >90% to detect samples with virologic failure (≥50 copies/mL). One matrix with low relative efficiency would have been detected earlier using the developed QA measures, but its exclusion would have only increased relative efficiency from 39% to 42%. These methods would have saved between $13,223 and $14,308 for monitoring this cohort.

Despite limited clinical data, high prevalence of detectable viral loads and a contaminated matrix, pooling greatly improved efficiency of virologic monitoring while maintaining accuracy. By improving cost-effectiveness, these methods could provide sustainability of virologic monitoring in resource-limited settings, and incorporation of developed QA measures will most likely maximize pooling efficiency in future uses.