Table 3 |
||||||||||
| Immunohistochemical strategies, statistical results | ||||||||||
| Strategy | TP | TN | FP | FN | S (%) | Sp (%) | PPV (%) | +LR | −LR | |
| Markers considered in a parallel manner | ||||||||||
| 1 | HR + CK7 + P-15 | 10 | 42 | 0 | 19 | 35 | 100 | 100 | 999.999 | 0.66 |
| 2 | HR + CK7 + P-15 + MAG | 7 | 42 | 0 | 22 | 24 | 100 | 100 | 999.999 | 0.76 |
| Markers considered in a serial manner | ||||||||||
| 3 | HR----CK7 | 28 | 20 | 22 | 1 | 97 | 48 | 56 | 1.84 | 0.07 |
| 4 | HR----MAG | 21 | 39 | 3 | 9 | 70 | 93 | 88 | 9.80 | 0.32 |
| 5 | HR----MAG----CK7 | 28 | 22 | 20 | 1 | 97 | 52 | 58 | 2.03 | 0.07 |
HR: estrogen and progesterone receptors; CK7: cytokeratin 7; P-15: gross cystic disease fluid protein (GCDFP-15); MAG: mammaglobin; TP: true positives; TN: true negatives; FP: false positive; FN: false negative; S: sensitivity; Sp: specificity; PPV: positive predictive value; + LR: positive likelihood ratio; − LR: negative likelihood ratio.
Noriega et al. Diagnostic Pathology 2012 7:73 doi:10.1186/1746-1596-7-73
