Table 5.
Contingency tables used two alternative cut-off values to compare TDM-based indications by UPLC-MS/MS and EMIT or adjusted EMIT by Deming Linear regression.
| VRZ determined by EMIT with cut-off at 1 mg/L | Total | p-value | VRZ determined by EMIT and adjusted by DLR with threshold at 1 mg/L | Total | p value | |||||
|---|---|---|---|---|---|---|---|---|---|---|
| VRZ determined by UPLC-MS/MS with a cut-off at 1 mg/L | <1 | ≥1 | <1 | ≥1 | ||||||
| <1 | 12 | 7 | 19 | p Fisher<0.0001 | <1 | 15 | 4 | 19 | p Fisher<0.0001 | |
| ≥1 | 0 | 103 | 103 | ≥1 | 0 | 103 | 103 | |||
| Total | 12 | 110 | 122 | p McNemar = 0.016a | 15 | 107 | 122 | p McNemar = 0.125 | ||
| VRZ determined by EMIT with cut-off at 5.5 mg/L | VRZ determined by EMIT and adjusted by DLR with threshold at 5.5 mg/L | |||||||||
| VRZ determined by UPLC-MS/MS with a cut-off at 5.5 mg/L | ≤5.5 | >5.5 | ≤5.5 | >5.5 | ||||||
| ≤5.5 | 89 | 11 | 100 | p Fisher<0.0001 | ≤5.5 | 99 | 1 | 100 | p Fisher<0.0001 | |
| >5.5 | 0 | 22 | 22 | >5.5 | 1 | 21 | 22 | |||
| Total | 89 | 33 | 122 | p McNemar=0.001a | 100 | 22 | 122 | p McNemar = 1.0 | ||
DLR, dynamic linear regression; EMIT, enzyme multiplied immunoassay technique; UPLC-MS/MS, ultra-performance liquid chromatography-tandem mass spectrometry; VRZ, voriconazole.
a
p Fisher significance implies a significant correlation. In contrast, p McNemar's test indicates a significant discordance.