Table 2 Test performance characteristics of the regression-based and WHO-based monitoring algorithms to determine viral failure (>1000 copies/ml) in 496 Ugandans on ART at the Infectious Diseases Institute in Kampala, Uganda
| Â | Sensitivity (95% CI) | Specificity (95% CI) | PPV (95% CI) | NPV (95% CI) | % Failures missed (1-sensitivity) | % Switched unnecessarily** | % Patients tested |
|---|---|---|---|---|---|---|---|
Regression-based variables (30% CD4 fall or ever missed >2 days) with viral load testing [see Figure 1] | 67% (63–71%) | 100% | 100% | 97% (96–99%) | 33% | 0% | 22% |
Regression-based variables (30% CD4 fall or ever missed >2 days) without viral load testing | 67% (63–71%) | 82% (79–85%) | 24% (20–28%) | 97% (96–99%) | 33% | 18% | 0% |
WHO-based criteria (CD4 failure* or Stage 4 disease) without viral load testing [see Figure 1] | 31% (27–35%) | 87% (84–90%) | 16% (13–19%) | 94% (92–96%) | 69% | 14% | 0% |
WHO-based criteria (CD4 failure* or Stage 4 disease) with viral load testing | 31% (27–35%) | 100% | 100% | 94% (92–96%) | 69% | 0% | 15% |