Table 2. Area under the curve (AUC) of a receiver-operator characteristics (ROC) graph comparing the accuracy of the PwD, BED, and LAg assays in identifying HIV infection recency.

Assay	AUC	SE	95% CI
130-day cut-off
PWD	0.83	0.03	0.78–0.88
BED	0.78	0.02	0.74–0.82
LAg	0.81	0.02	0.78–0.85
180-day cut-off
PWD	0.82	0.03	0.76–0.88
BED	0.75	0.02	0.71–0.79
LAg	0.79	0.02	0.75–0.82
360-day cut-off
PWD	0.78	0.05	0.68–0.89
BED	0.74	0.03	0.69–0.79
LAg	0.72	0.02	0.68–0.77

The table shows the results for the area under the curve (AUC) of a receiver operating characteristics (ROC) graph. The AUC is an objective measure of the accuracy of a classification schema. The best possible value is 1.0, which represents a 100% sensitivity and 100% specificity of the assay to correctly distinguish recent from established HIV infections. The results show that the PwD assay has the best predictive performance for the three window periods. CI: confidence interval.