To Name :
To Email :
From Name :
From Email :
Comments :

Clinical Reviews

Assessment of abnormal uterine bleeding: 3 office-based tools

Sophisticated, user-friendly tools now available for office diagnosis of dysfunctional bleeding are underutilized. Here, a thorough assessment of transvaginal ultrasound, saline-infusion sonography, and hysteroscopy.

May 2003 · Vol. 15, No. 5


  • Identification and measurement of the endometrial echo and descriptions of the echogenicity and heterogeneity of the endometrium are key to defining endometrial health.
  • The introduction of intracervical fluid (saline-infusion sonography) during transvaginal ultrasound is one of the most significant advances in ultrasonography of the past decade.
  • Hysteroscopic visualization has several advantages: immediate office evaluation, direct visualization of the endometrium and endocervix, and the ability to detect minute focal endometrial pathology and to perform directed endometrial biopsies.
  • Sometimes a combination of procedures may be the best way to determine the cause of abnormal uterine bleeding.

Because office-based physicians tend to feel comfortable relying upon endometrial biopsy or dilation and curettage (D&C) to evaluate abnormal uterine bleeding, newer tools—transvaginal ultrasound (TVUS), saline-infusion sonography (SIS), and hysteroscopy—see far too little utilization. Although these modalities are remarkably user-friendly when employed correctly, only 28% of gynecologists perform office hysteroscopy, and even fewer use SIS.1

This article reviews indications for use, sensitivity and specificity, advantages and disadvantages, special considerations including cost issues, and suggestions for incorporating these modalities into gynecologic practice.

For most patients, neither endometrial biopsy nor D&C is particularly helpful in the assessment of abnormal uterine bleeding (see “Limitations of D&C and endometrial biopsy”). Does a negative biopsy signify health? An absence of intracavitary pathology? Or does it mean that we failed to sample the culprit?

Fortunately, abnormal uterine bleeding usually is attributable to endometrial conditions other than cancer, such as atrophy, hyperplasia, polyps, or fibroids. These pathologies may be benign, but they are still a nuisance. When left undetected and untreated, they can cause the patient much worry, along with weeks or months of unnecessary medical therapy and surgical procedures.

Limitations of D&C and endometrial biopsy

Dilation and curettage (D&C) detects benign pathology in about 80% of patients with menstrual dysfunction.1 It is most likely to detect the problem when pathology affects the endometrium globally.

A recent study reported on 105 postmenopausal women with bleeding and an endometrial echo of more than 5 mm who were evaluated with both hysteroscopy and D&C.2 Although 80% of the women had pathology in the uterine cavity, and 98% of the pathologic lesions manifested a focal growth pattern at hysteroscopy after D&C, the whole or parts of the lesion remained in situ in 87% of the women. In addition, D&C failed to detect 58% of polyps, 50% of hyperplasias, 60% of complex atypical hyperplasias, and 11% of endometrial cancers. When disease was global, D&C detected 94% of abnormalities.

Focal disease therefore mandates operative hysteroscopic-directed biopsy and removal of suspicious pathology.

Endometrial biopsy can be performed in the office without anesthesia—a great advantage. The technique is most helpful in “dating” the endometrium and diagnosing endometrial cancer or hyperplasia. Unfortunately, blind endometrial biopsy studies are frequently returned with pathology reports of insufficient tissue, atrophic changes, mucus and debris, scanty tissue, no visible endometrial tissue, endocervical tissue, or proliferative or secretory endometrium.

Further, when a blindly performed biopsy reveals normal histology, it does not necessarily rule out other pathology. In addition, a biopsy via endometrial suction curette frequently misses focal lesions such as endometrial polyps and submucosal fibroids. Global noninvasive surveillance of the endometrium is more effective at detecting such focal lesions.

Investigators who performed endometrial biopsy prior to hysterectomy in patients with known endometrial cancer demonstrated that the sensitivity of diagnosing endometrial cancer with a biopsy via endometrial suction curette increases when the pathology affects more than 50% of the surface area of the endometrial cavity.3 However, biopsy failed to detect cancer in 11 of 65 patients in whom the malignancy affected less than 50% of the endometrium. These 11 false negatives included 5 cases of endometrial polyps, 3 malignancies that affected less than 5% of the endometrium, and 7 cancers that affected less than 25% of the endometrium.


1. Holst J, Koskela O, von Schoultz B. Endometrial findings following curettage in 2,018 women according to age and indications. Ann Chir Gynaecol. 1983;72:274-277.

2. Epstein E, Ramirez A, Skoog L, Valentin L. Dilation and curettage fails to detect most focal lesions in the uterine cavity in women with postmenopausal bleeding. Acta Obstet Gynecol Scand. 2001;80:1131-1136.

3. Guido RS, Kanbour-Shakir A, Rulin MC, Christopherson WA. Pipelle endometrial sampling: sensitivity in the detection of endometrial cancer. J Reprod Med. 1995;40:553-555.

Transvaginal ultrasound

TVUS is quick, convenient, inexpensive, and comfortable for the patient. It can evaluate the endometrium utilizing gray scale, color or power Doppler, contrast media (SIS), or 3-dimensional ultrasound technology. In addition, TVUS permits visualization of the adnexa and pelvic organs, including the bladder and cul de sac. Among the abnormalities detectable with TVUS are fibroids (including submucosal leiomyomas) and endometrial polyps. Not surprisingly, an experienced operator is crucial for a precise diagnosis.

Optimal TVUS evaluation includes endometrial measurements in the sagittal plane, with the bilayer thickness measured from the proximal to distal myometrialendometrial junctions. In the coronal view, measurement should be from the cervix to the fundus. When intraluminal fluid is present, each endometrial thickness should be measured separately (in single layers), and the combined endometrial thickness should be expressed as the sum of the 2 layers.

Characterize the endometrium. Sonographic hallmarks increasingly are used to describe uterine pathology. These include endometrial thickness, morphology, endocavitary lesions, borders, and myometrial invasion. In addition, a number of practitioners have attempted to describe the sonographic texture of the endometrium in post-menopausal women to increase sensitivity in the detection of disease, improve diagnosis, and guide treatment.

The endometrium can be characterized as homogeneous, diffusely inhomogeneous, or associated with focally or diffusely increased echogenicity. Textural inhomogeneity is present in cases of endometrial cancer.2 A homogeneous endometrium less than 6 mm thick is commonly associated with tissue insufficient for diagnosis. If focal or diffuse increased echogenicity occurs with a thin endometrium, SIS or hysteroscopy is more sensitive than TVUS in identifying the endometrial abnormality.

We rely increasingly on descriptions of the echogenicity and heterogeneity of the entire endometrial echo, as well as on echo measurements, to define endometrial health. Most experts use an endometrial echo of 5 mm as the cutoff for significant endometrial disease. A number of investigators have noted that the endometrial thickness in hyperplasia often ranges from 8 mm to 15 mm in post-menopausal women, but the number of reported cases is too low to draw firm conclusions based on endometrial echo alone.3,4

A prospective evaluation of 200 postmenopausal women with endometrial echo ranging from 3 mm to 10 mm noted that homogeneity, thin endometrium, and sonographically demonstrable central endometrium with symmetry were associated with absence of pathology. In contrast, heterogeneity and high echogenicity were indicative of pathology.5

Endometrial echo measurement and morphology increase sensitivity in predicting endometrial disease and can point to the need for ancillary testing with SIS or hysteroscopy. For example, in the postmenopausal patient, an endometrial echo of less than 5 mm, in combination with a negative endometrial biopsy, might be the only evaluation the patient needs if she responds to medical therapy for endometrial atrophy. If symptoms persist, office hysteroscopy or SIS could be performed to rule out endometrial polyps and endometrial hyperplasia (which is less likely).

Saline-infusion sonography offers an exquisite view of the endomyometrial complex that cannot be obtained with transvaginal ultrasound alone.

A thickened endometrial echo (more than 5 mm) at the initial TVUS in the postmenopausal patient should be evaluated promptly with SIS. In most of these patients, polyps or fibroids are the cause of bleeding.

Perform imaging at optimal time. In postmenopausal women, the endometrial thickness remains constant unless the patient is taking hormone replacement therapy (HRT) or tamoxifen.6 Thus, it is easier to detect or predict endometrial pathology in this population.

In premenopausal women, TVUS is most likely to detect fibroids during the early follicular stage because the endometrium is thin, usually measuring 2 mm to 4 mm. Endometrial polyps and submucosal fibroids are best viewed when a trilayered midfollicular endometrium is present, while uterine synechiae are visualized most clearly during midcycle. Adhesions appear as hyperechoic irregular structures that vary in size from 2 mm to 6 mm. They interrupt the continuity of the endometrial layer and appear much different from polyps, which tend to be round, symmetrical, and uniform.7

Diagnosing endometrial hyperplasia among premenopausal women based on an absolute endometrial measurement is more difficult than it is in postmenopausal women because of the wide range in thickness associated with the menstrual cycle (TABLE 1). Although endometrial hyperplasia or cancer is more likely when the endometrial echo is thicker than anticipated based on age or menstrual phase, it can only be proven definitively with histologic sampling.

Sensitivity and specificity. Because of the monotonous nature of the endometrium in healthy postmenopausal women, TVUS has higher sensitivity in this age group than in reproductive-age women. In the postmenopausal population, the sensitivity for detecting uterine pathology is 87%, while specificity is 82%.

Office hysteroscopy offers immediate evaluation and direct visualization of the endometrium and endocervix.

In premenopausal women, polyps were the pathology most likely to be missed, according to one literature review.8 TVUS identified only 275 of 344 polyps in this population—a sensitivity of 80%. When submucosal fibroids were located near the endometrium, the diagnostic sensitivity rose to 94%. It was difficult to discern location (i.e., submucosal or intramural or polyps) with TVUS alone, however.

Another investigation compared the sensitivity and specificity of TVUS and endometrial biopsy for the detection of endometrial disease in 448 postmenopausal women who took estrogen alone, cyclic or continuous estrogen-progesterone, or placebo for 3 years.9 Using a threshold value of 5 mm for endometrial thickness, TVUS had a positive predictive value (PPV) of 9% for detecting any abnormality, with 90% sensitivity, 48% specificity, and a negative predictive value (NPV) of 99%. Using this threshold, biopsies would be indicated in more than half of the women, of whom only 4% actually had serious disease.

Using a threshold thickness of 4 mm, Gull et al10 evaluated TVUS for the detection of endometrial cancer and atypical hyperplasia. For endometrial thicknesses exceeding 4 mm, TVUS had a sensitivity of 100%, specificity of 60%, a PPV of 25%, and an NPV of 100%. No woman with an endometrial thickness of 4 mm or less was found to have endometrial cancer.

Special considerations.

  • Intracavitary fluid. On occasion, TVUS reveals “naturally occurring SIS”—that is, the spontaneous appearance of intracavitary fluid, which is probably secondary to cervical stenosis. (The name of this phenomenon refers to the iatrogenic fluid that is intrinsic to SIS.) Endometrial fluid also is observed in women who use tamoxifen, diethylstilbestrol, or megestrol acetate and in those who have a hematometra. In addition, about one quarter of patients with an endometrial malignancy have fluid in the endometrial cavity.11
  • Endometrial thickness. As noted, TVUS is excellent for ruling out endometrial abnormalities in postmenopausal patients because of the consistent thickness of the endometrium. When the endometrium is especially thickened, however, its usefulness is limited.

Observations in more than 5,000 women consistently note that an endometrium of less than 5 mm, measured as a double-layer, is most often associated with a pathology reading of “tissue insignificant for diagnosis”—hence atrophy.12 When the endometrium is thicker than 5 mm, there is a greater chance of detecting polyps, endometrial hyperplasia, and endometrial cancer. Fewer than 0.12% of cancers were missed in this series.

When a skilled physician performs office hysteroscopy, the complication rate is less than 1%.

If the endometrium is thicker than 6 mm, then SIS or hysteroscopy is more helpful than TVUS in reaching a conclusive etiology for uterine bleeding. Further evaluation (hysteroscopy or SIS) is usually recommended when TVUS measurements of the endometrium are greater than 5 mm.13 Although a low cut-off, such as 4 mm or 5 mm, is associated with improved sensitivity, it sacrifices specificity.


Transvaginal ultrasound: Endometrial thickness32





Early proliferative








Postmenopausal on hormone replacement therapy


Saline-infusion sonography

In SIS, saline is infused into the endometrial cavity during TVUS to enhance the view of the endometrium. This constitutes one of the most significant advances in ultrasonography of the past decade. SIS can provide a wealth of information about the uterus and adnexa in patients with abnormal bleeding. It offers an exquisite view of the endomyometrial complex that cannot be obtained with TVUS alone. It differentiates between focal and global processes and improves overall sensitivity for detecting abnormalities of the endometrium.

Did you miss this content?
Is supplemental ultrasonography a valuable addition to breast cancer screening for women with dense breasts?