Transvaginal sonography (TVS) has been shown to be the most effective means to screen for ovarian cancer. TVS is associated with a high sensitivity and specificity. However, the positive predictive value associated with TVS in the diagnosis of malignancy is low. A morphologic scoring index for use with TVS has been used at the University of Kentucky since 1991. The current study was performed to more fully evaluate the efficacy and interobserver variation in ultrasonographic morphology index scores attributed to ovarian tumors. Ultrasound records of 213 patients from five participating centers were reviewed by three independent observers. Morphology index scores were assigned to each tumor in a blinded fashion. The morphology index scores were then compared with the final histopathologic findings. One hundred sixty- nine patients had benign tumors and 44 patients had ovarian malignancies. The mean morphology index scores were significantly higher in malignant ovarian tumors (MI 7.3 ± 1.9) than in benign ovarian tumors (MI 3.3 ± 1.8). Statistical evaluation of the morphology index scores revealed a sensitivity of 89% and a positive predictive value of 46%. Interobserver variation was lowest in assessing ovarian volume and higher in the evaluation of wall structure and septal structure. A multilogistic regression model was used to evaluate the predictive power of each component of the morphology index. The use of a morphology index is an effective and cost-efficient method of increasing the positive predictive value of TVS screening for ovarian cancer. Use of this index in large numbers of patients will generate data which should help refine appropriate structural scoring categories and reduce interobserver variation.
ASJC Scopus subject areas
- Obstetrics and Gynecology