Table 3 The overall sensitivity and specificity (benign vs malignant) of the ADNEX model with and without the inclusion of serum CA125

From: Evaluating the risk of ovarian cancer before surgery using the ADNEX model: a multicentre external validation study

Cutoff

Patients with riskcutoff, N (%)

Sensitivity with 95% CI

Specificity with 95% CI

ADNEX with CA125

1%

559 (91.6%)

100.0% (97.4–100.0)

11.9% (9.1–15.5)

3%

479 (78.5%)

100.0% (97.4–100.0)

30.6% (26.3–35.3)

5%

383 (62.8%)

99.0% (94.9–99.8)

53.2% (48.2–58.1)

10%

315 (51.6%)

97.3% (93.5–98.9)

67.7% (63.0–72.0)

15%

281 (46.1%)

94.4% (90.0–97.0)

75.2% (70.7–79.2)

20%

253 (41.5%)

90.6% (85.2–94.1)

79.3% (75.1–83.0)

30%

226 (37.0%)

86.3% (80.4–90.6)

83.9% (80.1–87.2)

ADNEX without CA125

1%

557 (91.3%)

100.0% (97.4–100.0)

12.4% (9.5–16.0)

3%

490 (80.3%)

100.0% (97.4–100.0)

28.0% (23.9–32.6)

5%

374 (61.3%)

98.9% (95.7–99.7)

54.7% (49.9–59.3)

10%

317 (52.0%)

96.7% (92.9–98.5)

67.1% (62.5–71.3)

15%

289 (47.4%)

94.5% (90.1–97.0)

72.7% (68.2–76.7)

20%

261 (42.8%)

90.7% (85.5–94.1)

77.6% (73.4–81.3)

30%

225 (36.9%)

84.6% (78.6–89.2)

83.4% (80.0–86.6)

  1. Abbreviations: ADNEX=The Assessment of Different NEoplasias in the adneXa; CA125=cancer antigen 125; CI=confidence interval.
  2. When using a 1% or 3% cutoff, confidence limits are calculated through use of Wilson’s score confidence interval method with continuity correction (Newcombe, 1998). For the other cutoffs, confidence limits are calculated using logistic regression to combine results after multiple imputation.