Skip to main content
. Author manuscript; available in PMC: 2010 Jan 1.
Published in final edited form as: Gynecol Oncol. 2008 Nov 20;112(1):95–103. doi: 10.1016/j.ygyno.2008.10.005

Table 3.

Odds ratios from univariate logistic regressions models employing generalized estimating equations for the features of average tumor size, parametrial invasion and uterine involvement by modality based on the retrospective multi-observer readings, along with average ROC AUC and range over the readers for detection of histologic lymph node involvement. Thresholds used were for average tumor size >3cm versus <=3cm; for parametrial invasion bilateral or unilateral invasion versus no invasion; for uterine involvement 5:Definitely present versus <5

Response Modality Tumor size Parametrial involvement Uterine involvement
Odds ratio (95% CI) Average ROC AUC (range) Odds ratio (95% CI) Average ROC AUC (range) Odds ratio (95% CI) Average ROC AUC (range)
Histologic lymph node involvement specific to common iliac nodes MRI 2.8 (1.2, 6.5) 0.66 (0.63 - 0.69) 2.7 (1.2, 6.1) 0.60 (0.54 - 0.70) 5.7 (2.0, 16.6) 0.66 (0.58 - .79)
CT 1.1 (0.3, 3.3) 0.54 (0.42 - 0.64) 1.4 (0.5, 4.0) 0.51 (0.49 - 0.55) None recorded1 0.57 (0.46 - .66)
General histologic lymph node involvement (Any positive node) MRI 2.5 (1.3, 4.7) 0.63 (0.61 - 0.65) 2.5 (1.3, 4.7) 0.58 (0.54 - 0.62) 3.1 (1.2, 8.3) 0.57 (0.54 - .60)
CT 1.1 (0.6, 2.3) 0.55 (0.50 - 0.61) 1.5 (0.7, 3.2) 0.52 (0.50 - 0.54) 1.1 (0.2, 6.1) 0.58 (0.54 - .63)
1

Odds ratio could not be estimated as none of the cases which had para-aortic or common iliac metastases were coded as uterine involvement definitely present by any of the CT readers.