Intraobserver and interobserver agreement of grayscale typical ultrasonographic patterns for the diagnosis of ovarian cancer

The purpose of this study was to evaluate the intraobserver and interobserver agreement for identifying ovarian malignancy using typical grayscale ultrasonographic patterns. Digitally stored grayscale sonographic images from a random sample of 98 women with an adnexal mass submitted to surgery after a grayscale transvaginal sonography were evaluated by five different examiners with different degrees of experience in three European university departments of obstetrics and gynecology.

Masses in which the echo features were highly characteristic of a benign pathology were categorized as benign. Any cystic mass containing excrescences, thick septations, multiple irregular septations or solid component in which the echo architecture was not highly suggestive of benign histology was categorized as malignant. Intraobserver and interobserver agreement according to the level of experience were assessed by calculating the kappa index.

Of the 98 cases randomly selected, 28 (29%) were malignant masses and 70 (71%) were benign. Intraobserver agreement was good or very good for all examiners with different degrees of experience (kappa = 0.72 to 1). Interobserver agreement was good for all expert operators (kappa = 0.69 to 0.75). Interobserver agreement between experts and highly experienced operators was moderate or good (kappa = 0.51 to 0.63). Interobserver agreement between the moderately experienced operator and experts was fair to moderate (kappa = 0.29 to 0.46). Interobserver agreement between moderately and highly experienced operators was fair (kappa = 0.33).

Our results indicate that ultrasonographic malignant patterns are reproducible, even in moderately experienced examiners, although more experience is associated with better interobserver agreement.

CITATION  Ultrasound Med Biol. 2008 Nov;34(11):1711-6. Epub 2008 Jun 4.