Comparison of CBNAAT and conventional real time RT PCR for HIV 1 viral load testing

Abstract

Purpose: HIV viral load testing is now recommended for monitoring of anti-retroviral treatment failure in PLHIV. Xpert® HIV-1 Viral Load is a fully automated CB-NAAT. A reduced turnaround time leads to prompt clinical management. Hence the current study was undertaken to compare Xpert® HIV-1 Viral Load with the routinely used conventional real time RT PCR.

Methods: The study was conducted in the Department of Microbiology of a tertiary care medical college after ethics committee approval. 100 HIV positive samples were tested by both CB-NAAT and conventional real time RT PCR for HIV 1 viral load. Results were analyzed using Spearman's correlation co-efficient and Bland Altman plot for agreement. The number of samples with inter assay differences in viral loads exceeding 0.5 log copies/ml was also recorded. The sensitivity, specificity, PPV and NPV as well as the possible misclassification were calculated at the clinically significant value of 1000 copies/ml.

Results: 25 samples in each of the four groups with log 10 value of <3, 3 to <4, 4 to <5 and ≥5 respectively were included. The log difference between the groups varied from 0 to 1.54. CB-NAAT has shown a statistically significant correlation with conventional real time RT PCR by Spearman's rank correlation (R = 0.972) (P < 0.01) and acceptable level of agreement with Bland Altman plot. The sensitivity, specificity, PPV, NPV and diagnostic accuracy was 80%, 100%, 100%, 93.75% and 95% respectively. The overall concordance was 95% with an upward misclassification of 6.25% and downward misclassification of 0%.

Conclusions: Point of care technology with sample in/answer out approach makes it an excellent choice especially in resource constrained and remote settings.

Keywords: ART monitoring; CB-NAAT; HIV 1 viral load; RT PCR.