Detection of Benign Intracavitary Lesions in Postmenopausal Women with Abnormal Uterine Bleeding: A Prospective Comparative Study on Outpatient Hysteroscopy and Blind Biopsy

Abstract Study Objective To evaluate the specificity of blind biopsy in detecting benign intracavitary lesions as causes of postmenopausal bleeding in comparison with directed biopsy via hysteroscopy. Design Prospective trial without randomization (Canadian Task Force classification II-1). Setting University hospital. Patients Three hundred nineteen postmenopausal women with abnormal uterine bleeding. Interventions All patients underwent both blind biopsy (Novak’s curette) and directed biopsy via hysteroscopy (after at least a week). All patients with benign intracavitary lesions underwent operative hysteroscopy to enable the removal of polyps and intracavitary myomas or endometrial resection if required. All patients with pathologic reports of complex hyperplasia and atypical hyperplasia (20 patients) underwent vaginal hysterectomy with bilateral adnexectomy. All patients with histology reports of endometrial carcinoma (15 patients) underwent abdominal hysterectomy, bilateral adnexectomy, and pelvic lymphadenectomy. Histopathologic findings from endometrial specimens obtained after operative hysteroscopy or uterine specimens obtained after hysterectomy were used as a reference test to establish the prevalence of disease. Measurements and Main Results The sensitivity, specificity, accuracy, and positive and negative predictive values of blind biopsy and hysteroscopy were assessed to distinguish benign intracavitary formations such as polyps, submucous myomas, and endometrial hyperplasia in postmenopausal patients with abnormal uterine bleeding. The level of agreement was evaluated by use of the coefficient of concordance κ. Blind biopsy showed a sensitivity of 11% and a specificity of 93%, with an accuracy of 59% in detecting endometrial polyps, a sensitivity and specificity of 13% and 100%, respectively, with an accuracy of 98% for submucous myomas, and values of 25%, 92%, and 80%, respectively, in diagnosing hyperplasia. On the other hand, hysteroscopy demonstrated a sensitivity of 100% and a specificity of 97%, with an accuracy of 91% in diagnosing endometrial polyps, a sensitivity and specificity of 100% and 98%, respectively, with an accuracy of 99% for submucous myomas. The coefficient of concordance κ (95% CI) was 0.12 for blind biopsy and 0.82 for hysteroscopy, corresponding, respectively, to slight concordance and almost perfect agreement with final pathologic diagnosis. Conclusions Blind biopsy (Novak’s curette) demonstrates very low sensitivity