Diagnostic Accuracy of Transvaginal Sonography for Abnormal Uterine Bleeding [31G

Abstract

INTRODUCTION: The prevalence of intrauterine abnormalities in patients with abnormal uterine bleeding is 46%. Historical data gives the transvaginal sonography (TVS) a sensitivity and specificity of 56% and 73% for detecting intracavitary pathology respectively, however the image resolution of the ultrasound machines has improved significantly since those studies. The objective of this study is to calculate the sensitivity, specificity, positive and negative predictive values of TVS. METHODS: Retrospective chart review of patients who had a TVS and a subsequent hysteroscopy performed in the operating room at The Western Pennsylvania Hospital. 500 patients were needed for a 95% Confidence Interval of 0.06% RESULTS: Between January 2013 and September 2015, 2351 hysteroscopies were performed. Using a randomization software, 500 randomly selected patients that met inclusion criteria underwent chart review. The hysteroscopic findings were: normal cavity (243), endometrial polyp (157), atrophic endometrium (34), submucosal fibroid (25), other (41). We found a sensitivity of 63%, specificity 80.9%, positive and negative predictive values 67.9% and 77.4% respectively, a significant improvement comparing to historical data, however far from the most accurate studies, the sonohysterogram and diagnostic hysteroscopy. CONCLUSION: A false positive rate of 11% is still high if the next step in management will be a hysteroscopy performed in the operating room. In the last 2 years 258 hysteroscopies were performed in the operating room with a false positive diagnosis of intracavitary lesion. This results support the current recommendations of performing a sonohysterogram or office hysteroscopy after a TVS finding of intracavitary pathology.