Variability in HIV viral load quantification in real-world service delivery settings: findings from an observational...

Area of Science:

• Virology
• Clinical Diagnostics
• Public Health

Background:

• Advancements in HIV viral load (VL) testing improve low-level viremia detection.
• Significant variability exists in quantitative results from polymerase chain reaction (PCR) tests, particularly near assay lower limits of quantification (LLOQ).
• Discrepancies in VL quantification can impact patient management and treatment monitoring.

Purpose of the Study:

• To compare VL testing results from two laboratories in Rwanda using different assays.
• To characterize discordancy in VL quantification between laboratories.
• To identify factors associated with quantifiable VL results at a threshold of 40 copies/mL.

Main Methods:

• An observational cohort study involved people living with HIV (PWH) aged ≥40 years.
• Paired research specimens were processed by two laboratories using different VL assays (Lab A: 20 copies/mL LLOQ; Lab B: 40 copies/mL LLOQ).
• Kappa statistics and Poisson regression were used to assess between-lab agreement and identify factors associated with unsuppressed VLs (≥200 copies/mL).

Main Results:

• 593 results from 572 participants were analyzed; mean age was 54.4 years, 58% female.
• Quantifiable VLs at 20 copies/mL (Lab A) were 29.2%, compared to 4.7% at 40 copies/mL (Lab B).
• Poor to moderate agreement was observed below 200 copies/mL, with high agreement at 1000 and 2000 copies/mL; low CD4 counts (<200 cells/mm³) were linked to unsuppressed VLs.

Conclusions:

• Substantial differences in VL quantification exist between assays and laboratories in real-world settings.
• High agreement was found at higher viral load cut-offs commonly used in clinical decision-making.
• Findings support using higher viral suppression thresholds, aligning with UNAIDS 95-95-95 targets, to avoid misclassification at lower levels.