Reliability of adnexal mass mobility in distinguishing possible ectopic pregnancy from corpus luteum cysts

Abstract

Study objectives: Pelvic ultrasonography for ruling out ectopic pregnancy is one of the most common emergency ultrasonographic applications but may be frustrating when a potential ectopic mass cannot be differentiated from a corpus luteum cyst. An adnexal mass adjacent to an ovary that fails to move independently from the ovary under sonographic palpation is thought unlikely to be an ectopic. We evaluate the utility of pelvic mass separation from adjacent ovary as an indicator of mass identity (corpus luteum versus other) in ruling out ectopic pregnancy. Methods: This was a retrospective quality assurance review of all pelvic ultrasonographic studies on first-trimester complication patients treated during an 18-month period at an academic Level I emergency department with hospital-based emergency ultrasonographic credentialing. All pregnant patients presenting with signs or symptoms suspicious for ectopic pregnancy were scanned when credentialed faculty were available. Each ultrasonographic examination was videotaped in entirety per department policy. When no intrauterine pregnancy was seen and an adnexal mass that was suspicious for either corpus luteum or an ectopic pregnancy was encountered, sonologists attempted to separate the mass from the ovary with abdominal palpation and endovaginal transducer movement. All such scans were reviewed by the ultrasonographic director and assistant director. A database recorded results of attempts to separate masses, quantitative β–human chorionic gonadotropin findings on follow-up, and final outcome. Patients for whom data were incomplete or inaccurate or follow-up records were missing were excluded from analysis. Separation of mass and ovary was defined as movement of the mass away from the ovary, sliding past the ovary, or rotation past the ovary. Statistical analysis included descriptive statistics, agreement analysis, sensitivity, specificity, negative and positive predictive value, and 95% confidence intervals (CIs). Results: Agreement between the reviewers was excellent for mass separation (κ=0.9). For 2 patients, records were inaccurate, and the patients were not pregnant on blood testing. A total of 63 patient scans fit the criteria when an ectopic pregnancy could neither be ruled in or out and a mass was present directly adjacent to the ovary and was not clearly identified as an ovarian cyst. Of these, 19 (30%) were ectopic pregnancies on final diagnosis, 13 (22%) were live pregnancies, and 31 (48%) were miscarriages. Among the 41 patients for whom there was no separation of mass and ovary, the final diagnosis was ectopic pregnancy in 2 (5%) patients. Lack of mass and ovary separation yielded a sensitivity of 95% (95% CI 67% to 99%) and a specificity of 77% (95% CI 55% to 92%) for ruling out an ectopic pregnancy. The negative predictive value of no mass and ovary separation was 88.7% (95% CI 66.9% to 98.7%) for the presence of an ectopic pregnancy. Separation of the mass and ovary yielded a sensitivity of 77% (95% CI 55% to 92%) and a specificity of 95% (95% CI 83% to 99%) for an ectopic pregnancy. The positive predictive value of mass separation from the ovary was 89% (95% CI 67% to 99%) for an ectopic pregnancy. Conclusion: Although frequently taught as a definitive diagnostic maneuver, lack of separation of an adjacent mass and ovary does not preclude the mass from being an ectopic pregnancy. However, according to the negative predictive value, lack of separation of mass and ovary makes it unlikely that the mass is an ectopic pregnancy and may help in risk stratification.