Table 1.
Accuracy of Endorectal Ultrasound
| Study | Year | N | T-Staging (Overall) (%) | T1 | T2 | T3 | T4 | N-Staging (%) |
|---|---|---|---|---|---|---|---|---|
| Badger et al16 | 2007 | 95 | 71.6 | – | – | – | – | 68.8 |
| Landmann et al20 | 2007 | 134 | – | – | – | – | – | 70 |
| Skandarajah et al14 | 2006 | 2718 | 81.8 | 76 | 75 | 88 | 87 | – |
| Ptok et al10,* | 2006 | 3501 | 65.8 | 76.4 (402) | 56.0 (1208) | 71.2 (1780) | 48.6 (111) | – |
| Kulig et al25 | 2006 | 29 | – | 89.2 | 96.2 | – | – | – |
| Giovannini et al22,† | 2006 | 35 | 71.4 | 71.4 | ||||
| Kim et al18 | 2006 | 85 | 69 | 56 | ||||
| Zammit et al24 | 2005 | 117 | 76.4 | – | – | – | – | 73.6 |
| Bali et al26 | 2004 | 29 | 79 | – (0) | 50 (4) | 84 (25) | – (0) | 59 |
| Garcia-Aguilar et al15,* | 2002 | 545 | 69 | 47 (105) | 68 (153) | 70 (131) | 50 (8) | 64 (238) |
| Kim et al23,‡ | 2002 | 33 | – | – | 84.8 | 75.8 | – | 66.7 |
| Hünerbein et al21,§ | 2000 | 30 | 83 | 71.4 (7) | 90.9 (11) | 50 (2) | 100 (1) | – |
| Gualdi et al27 | 2000 | 26 | 76.9 | 76 | ||||
| Akasu et al19 | 2000 | 154 | ||||||
| Blomqvist et al28,‡ | 2000 | 49 | 84.1 | 89.8 | 77.6 | 63.3 | 91.8 | 60 |
| Kruskal et al29 | 2000 | 26 | 76.9 | |||||
| Kim et al13,‡,§ | 1999 | 89 | 81.1 | 100 (4) | 50 (12) | 87 (66) | 71 (7) | 63.5 |
| Massari et al17,§ | 1998 | 75 | 90.7 | 86.7 (15) | 88.9 (18) | 91.4 (35) | 100 (7) | 76 |
*
Accuracies are really PPVs (positive predictive value) (TP/TP [true positive] + FP [false positive]) for each UTx.
†
Accuracies are based on four TN (true negative) groups (T1/T2N0, T3N0, T3N1, T4N1)-T and N accuracy not given independently.
‡
(For N-staging) Accuracy for each Tx = total # of correct calls by ultrasound (TP and TN) for each T per total number of patients.
§
Accuracies represent sensitivities (TP/TP + FN [false negative]) for each pTx. Overall sensitivity calculated by adding the total TP made per total number of patients.