Table 2. Summary of Phase II validation results.
| A) Predicate PLG/CD4 vs. Clinic FACSPresto (accuracy) | B) Laboratory FACSPresto vs. Clinic FACSPresto (venous vs. capillary sampling) | |||
|---|---|---|---|---|
| Absolute CD4 count | CD4% lymphocytes | Absolute CD4 count | CD4% Lymphocytes | |
| Number of samples (n) | 118 | 118 | 118 | 118 |
| %Similarity Mean ±STDev (Range) | 107.6±10.56 | 96.7±5.8 | 105.1±8.8 | 95.3±5.8 |
| %Similarity %CV | 9.81 | 6.08 | 8.38 | 6.18 |
| Bland-Altman bias (Mean ±STDev) | 65.2±99.3 | -153±2.46 | 50.2±92.79 | -2.8±2.73 |
| Bland-Altman 95% LOA | -129.5 to260.6 | -6.36 to 3.29 | -131.7–232.1 | -9.22 to 2.48 |
Capillary filled FACSPresto™ compared to (A) predicate PLG/CD4 method and (B) venous blood manually filled cartridges in the laboratory (FACSPresto™) (S2 Table)