Prognostic Value of the Plasma Level of sFcγRIII in Patients With Neutropenia of Unknown Origin (n = 66) Regarding the Risk of Infection
| . | Cut-off . | |
|---|---|---|
| sFcγRIII < 100 AU . | ANC < 1,000/μL* . | |
| Sensitivity† | 1.00 (15/15) | 0.60 (9/15) |
| Specificity | 0.41 (21/51) | 0.67 (34/51) |
| Positive predictive value | 0.33 (15/45) | 0.35 (9/25) |
| Negative predictive value | 1.00 (21/21) | 0.85 (34/40) |
| . | Cut-off . | |
|---|---|---|
| sFcγRIII < 100 AU . | ANC < 1,000/μL* . | |
| Sensitivity† | 1.00 (15/15) | 0.60 (9/15) |
| Specificity | 0.41 (21/51) | 0.67 (34/51) |
| Positive predictive value | 0.33 (15/45) | 0.35 (9/25) |
| Negative predictive value | 1.00 (21/21) | 0.85 (34/40) |
*The optimal cut-off point for ANC was determined with ROC curve.
A positive test was defined as an sFcRIII level below 100 AU. Sensitivity, specificity, positive predictive value, and negative predictive value were calculated with 2 × 2 tables (infectionv sFcγIII < 100: n = 15, infection vsFcγRIII > 100: n = 0, no infection vsFcγRIII < 100: n = 30, no infection vsFcγRIII > 100: n = 21; infection v ANC < 1,000: n = 9, infection v ANC > 1,000: n = 6, no infectionv ANC < 1,000: n = 17, no infection vANC > 1,000: n = 34).