The Likelihood Ratio of Sonographic Findings for the Diagnosis of Hemorrhagic Ovarian Cysts

Abstract

The purpose of this study was to quantify the likelihood ratio (LR) of specific sonographic features and experienced sonologist assessment in diagnosing a hemorrhagic ovarian cyst and to better understand the diagnostic power of sonography in making this diagnosis. Two sonologists, blinded to the patient's clinical history, independently reviewed the sonograms of 252 adnexal masses. For each mass, each sonologist recorded features using a standardized checklist, which included observations regarding the presence of fibrin strands, a retracting clot, septations, and wall irregularity. Each reviewer independently chose 1 specific conclusion from a list of possibilities that included the diagnosis of a hemorrhagic ovarian cyst. Sonographic observations and sonologist predictions were compared with pathologic findings (n = 214) or follow-up sonography (n = 38). Of the 252 masses, there were 30 hemorrhagic cysts. One sonologist correctly identified 25 hemorrhagic cysts, with 1 false-positive, 2 false-negative, and 3 indeterminate calls (LR 185). The other sonologist correctly identified 24 hemorrhagic cysts, with 1 false-positive, 0 false-negative, and 6 indeterminate calls (LR 178). The 2 cases incorrectly diagnosed as hemorrhagic cysts were endometriomas at pathologic diagnosis. Specific sonographic observations and combinations of observations performed as follows: retracting clot (LR >67); fibrin strands (LR 40); fibrin strands and no septations (LR 100); and fibrin strands, no septations, and smooth wall (LR 200). Fibrin strands and a retracting clot are paramount observations in allowing high confidence in the diagnosis of hemorrhagic ovarian cysts. Approximately 90% of hemorrhagic ovarian cysts will exhibit at least 1 of these 2 features.