Ultrasonography-Based Triage for Perimenopausal Patients with Abnormal Uterine Bleeding


Four hundred thirty-three women who were older than 39 years, were not menopausal, and who presented with abnormal uterine bleeding comprised the study population for this paper. Rather than using endometrial biopsy, Goldstein and colleagues developed an algorithm based on published literature and their own prior studies that used endovaginal ultrasonography to triage patients. Those who had a normal endometrial stripe on days 4-6 of a bleeding cycle were diagnosed as having "dysfunctional uterine bleeding" and were not evaluated further. Sixty-five percent of the women in the study fell into this group. The remaining 35% of women were subjected to saline infusion sonohysterography. Of these 153 patients, 61 were found to have normal endometrial architecture and were not evaluated further. The remaining 92 patients fell into four categories, including women with polyps, submucous myomata, thickened endometrium, and in two patients the procedure was technically inadequate. (See table.) Women with evidence of polyps had hysteroscopy and biopsy. In all cases, a polyp was confirmed pathologically.

Goldstein et al reviewed the published literature on endometrial biopsy and conclude that "undirected sampling will be fraught with error" They likewise point out that "hysteroscopy is very operator-dependent." They conclude that, based on their study, "ultrasonography is a safe, readily available modality for noninvasively imaging the endometrial cavity." Finally, they end their paper with the statement that, "undirected endometrial sampling is only appropriate if one first demonstrates that the endometrial process is indeed global and not focal." (Goldstein SR, et al. Am J Obstet Gynecol 1997;177:102-108.)


For a long time, we have been aware that neither traditional dilatation and curettage under general anesthesia nor office endometrial sampling is 100% accurate. Unfortunately, no one has ever published an appropriate study to show that anything else is better. Though there are large series of hysteroscopic procedures performed in women with abnormal uterine bleeding, and although many cancers have been diagnosed with this technique, no one has ever adequately compared the two techniques.

Recently, vaginal ultrasonography has emerged as a new tool for the evaluation of abnormal uterine bleeding. When the endometrial thickness is below a certain benchmark (and the published studies disagree as to what thickness represents an abnormality), several papers have suggested that nothing further needs to be done. Specifically, some authors (including this paper) suggest that endometrial biopsy is not necessary. Unfortunately, most of these papers (including this paper) do not validate their negative vaginal ultrasounds with a tissue biopsy nor do they present follow-up data. In this study, for example, Goldstein et al conclude that 79% of the women studied had no pathology based solely on the results of vaginal ultrasonography. I imagine that everyone who reads the paper will ask themselves, "Were there any cancers missed in this group of normal women?" Of course, we will never know since Goldstein et al did not obtain tissue.

This article caused me to sit back, put my feet up on my desk, and think about the future: I wonder what we will be doing to evaluate abnormal uterine bleeding in


Using Endovaginal Ultrasonography to Triage Patients

Findings Number % of total

Abnormal Bleeding (Premenopause) 433 100% Abnormal Endovaginal Ultrasound 153 35% Abnormal Saline Infusion Ultrasound 92 21% Polyps 58 13% Submucuous myoma 22 5% Thick endometrium 10 2% Technically inadequate 2 0.5%

premenopausal women in 10 years? I doubt office endometrial biopsy will be considered sufficient. On the other hand, I also doubt that routine endovaginal ultrasonography followed by saline infusion hysterography (as suggested by Goldstein et al) will be the method of choice. Blind endometrial biopsy simply misses too much, and endovaginal sonography is too expensive and dependent on technical ability. Office hysteroscopy using a microhysteroscope (By the way, what will the next generation of hysteroscopes be called? Metamicrohysteroscopes?) may be the first evaluation step. Perhaps we will have a better treatment that is so specific that only those women who fail to respond need further evaluation.