Endometrial thickness as measured by transvaginal sonography: interobserver variation.

Transvaginal sonography of the uterus has become an important tool for diagnosing endometrial pathology in women with postmenopausal bleeding. One parameter that has been claimed to be important for curly detection of endometrial pathology in this group of women is the endometrial thickness. The purpose of this study was to evaluate the reproducibility of measurements of endometrial thickness as performed by five inexperienced doctors and one doctor experienced in the transvaginal sonography technique. The endometrial thickness as measured by the experienced doctor was regarded as the 'true measurement', and the mean discrepancy from this true measurement was 1.5 mm for the inexperienced doctors. No endometrial pathology was found at the histopathological examination after dilatation and curettage (D & C) when the endometrial thickness was measured as <or= 4 mm by both the inexperienced and the experienced doctors. However, there were considerable differences between the results obtained by the five inexperienced doctors. If this method is to be used for identifying those women who will not have a D & C performed, based on the findings of a thin endometrium (<or= 4 mm), training is needed in order to minimize the error. We believe that this will make measurement more accurate and it may then be used to exclude endometrial abnormality in women with postmenopausal bleeding. Copyright 1994 International Society of Ultrasound in Obstetrics and Gynecology