Table 1.
Intraclass correlation between plasma and DBS (uncorrected, mean hematocrit corrected and individual hematocrit corrected) HIV-RNA results, Zimbabwe National Microbiology Reference Laboratory, August–December 2018 (n = 517).
| Intraclass Correlation (ICC) of Plasma with | ICC * | (95% CI) | p-Value |
|---|---|---|---|
| All specimen (n = 517) | |||
| Uncorrected DBS | 0.640 | (0.586, 0.688) | <0.001 |
| Mean hematocrit corrected DBS | 0.859 | (0.835, 0.809) | <0.001 |
| Individual hematocrit corrected DBS | 0.809 | (0.777, 0.837) | <0.001 |
| Subgroup-plasma result ≥ 1000 (n = 65) | |||
| Uncorrected DBS | 0.635 | (0.462, 0.761) | <0.001 |
| Mean hematocrit corrected DBS | 0.857 | (0.769, 0.912) | <0.001 |
| Individual hematocrit corrected DBS | 0.805 | (0.694, 0.878) | <0.001 |
| Subgroup-plasma result < 1000 (n = 452) ^ | |||
| Uncorrected DBS | 0.000 ^ | (−0.091, 0.091) | 0.500 |
| Mean hematocrit corrected DBS | 0.000 ^ | (−0.091, 0.091) | 0.500 |
| Individual hematocrit corrected DBS | 0.000 ^ | (−0.091, 0.091) | 0.500 |
DBS = Dried blood Spot; * ICC using absolute agreement (two-way mixed, single measure); ^ negative average covariance among items. This violates reliability model assumption.