Objective: To define the etiologies of postmenopausal bleeding [PMB] categorized by clinical finding and endometrial histopathology and explore the diagnostic performance of endometrial biopsy compared to either fractional curettage [F&C] or hysteroscopic biopsy or hysterectomy.

Materials and Methods: The retrospective study of 100 consecutive postmenopausal women presenting with PMB at the out-patient department, Ramathibodi Hospital between January 1 and June 30, 2015. They were investigated for the causes of bleeding by vaginal examination and endometrial biopsy for histopathologic examination. Some patients received additional investigations or operations for final diagnosis, either F&C, directed hysteroscopic biopsy, or hysterectomy. The accuracy, sensitivity, and specificity of the endometrial biopsy was analyzed.

Results: From 100 consecutive PMB patients, the bleeding sources were vagina (3%), cervix (2%), and endometrium (95%). The etiologies of PMB were mostly from endometrial sources, including proliferative and other benign endometrial pathology (45%), atrophic endometrium (22%), endometrial polyp (12%), endometrial hyperplasia (7%), and endometrial carcinoma (9%). Besides, a small part of PMB resulted from atrophic vaginitis (3%) and cervical polyp (2%). The pathologic reports of 68 endometrial samplings revealed proliferative and other benign endometrium (48.5%), atrophic endometrium (14.7%), endometrial polyp (7.3%), endometrial hyperplasia (8.8%), endometrial cancer (7.3%), and inadequate endometrial tissue (13.2%). Some patients received further investigations or operations, either F&C, or directed hysteroscopic biopsy or hysterectomy. The discrepancy of the pathology, between endometrial sampling and further investigation or operations, was found in 37.5% (6 in 16 of patients). The sensitivity and specificity of endometrial biopsy for benign and malignant endometrial lesions were 62.5% and 100%, respectively. The positive predictive value [PPV], the negative predictive value [NPV], and the accuracy were 100%, 72.73%, and 81.25%, respectively.

Conclusion: The most common causes of PMB were benign in nature, however, endometrial carcinoma accounted around 9%. The diagnostic performance of endometrial biopsy for PMB was relatively low, demonstrated by low sensitivity and some risks of undetectable endometrial precancerous and cancerous lesion.

Keywords: Postmenopausal bleeding, Endometrial biopsy, Endometrial hyperplasia, Endometrial cancer