Table 4. Test performance of the IOTA diagnostic strategies, subjective assessment and RMI when using a meta-analysis approach on centre-specific data.
| Approach | AUC (95% CI) | Sens, % (95% CI) | Spec, % (95% CI) | LR+ | LR– | DOR (95% CI) |
|---|---|---|---|---|---|---|
|
One-step strategies | ||||||
| LR1 | 0.930 (0.917–0.942) | 93.7 (91.4–95.4) | 77.6 (70.9–83.0) | 4.17 | 0.08 | 40.8 (30.0–55.4) |
| LR2 | 0.918 (0.905–0.930) | 90.2 (86.9–92.8) | 78.9 (73.2–83.7) | 4.28 | 0.12 | 31.2 (23.1–42.2) |
| SA | 0.914 (0.886–0.936) | 92.5 (89.4–94.8) | 87.7 (83.2–91.2) | 7.53 | 0.09 | 72.9 (49.8–107) |
| RMI | 0.875 (0.853, 0.894) | 67.1 (61.4–72.4) | 90.6 (87.3–93.1) | 7.15 | 0.36 | 17.5 (13.1–23.4) |
| SRMal |
NA |
95.3 (93.1–96.9) |
74.1 (67.7–79.7) |
3.68 |
0.06 |
49.1 (34.9–69.0) |
|
Two-step strategies | ||||||
| SR+SA | NA | 91.8 (89.1–93.9) | 89.0 (85.2–92.0) | 8.38 | 0.09 | 75.7 (55.6–103) |
| LR2+SA | NA | 92.3 (89.5–94.5) | 84.8 (80.4–88.3) | 6.06 | 0.09 | 58.7 (43.4–79.4) |
| SD+SRMal | NA | 95.7 (93.5–97.1) | 73.6 (66.7–79.5) | 3.62 | 0.06 | 50.5 (35.7–71.6) |
| SD+LR2 | NA | 91.1 (88.1–93.5) | 78.1 (72.4–82.9) | 4.17 | 0.11 | 32.8 (24.6–43.7) |
| SD+SA |
NA |
93.0 (90.0–95.1) |
86.5 (81.8–90.1) |
6.88 |
0.08 |
68.5 (47.7–98.3) |
|
Three-step strategies | ||||||
| SD+SR+SA | NA | 92.5 (89.6–94.6) | 87.6 (83.5–90.7) | 7.44 | 0.09 | 70.7 (51.7–96.5) |
| SD+LR2+SA | NA | 93.1 (90.5–95.0) | 83.7 (79.2–87.4) | 5.71 | 0.08 | 57.6 (42.3–78.6) |
Abbreviations: AUC=area under the receiver-operating characteristics curve; CI=confidence interval; DOR=diagnostic odds ratio; IOTA=international ovarian tumour analysis; LR+=positive likelihood ratio; LR–=negative likelihood ratio; LR1=logistic regression model 1; LR2=logistic regression model 2; LR2+SA=LR2 as a first stage test and SA for tumours in which LR2 yields a predicted risk of malignancy of ⩾5% but <25% NA=not applicable; RMI=risk of malignancy index; RMI-1=risk of malignancy index-1; SA=subjective assessment; SD=simple descriptor; SD+SRMal=SDs as a first stage test, SRs for tumours unclassifiable by the descriptors with all tumours in which SRs are inconclusive being classified as malignant; SD+LR2=SDs as a first stage test and LR2 for those tumours in which the descriptors are not applicable; SD+SA=SDs as a first stage test and SA for those tumours in which the SDs are not applicable; SD+SR+SA=SDs as a first stage test, SRs for tumours in which the descriptors are not applicable, and SA for masses in which the SRs are inconclusive; SD+LR2+SA=SDs as a first stage test, LR2 for tumours in which the descriptors are not applicable, and SA for masses in which LR2 yields a predicted risk of ⩾5% but <25% Sens=sensitivity; Spec=specificity; SR=simple rules; SRMal=SRs as a first stage test with all tumours in which SRs are inconclusive being classified as malignant; SR+SA=SRs as a first stage test and SA for tumours in which the SRs are inconclusive.