Ultrasound screening for congenital uterine anomalies


Correspondence: Dr D. Jurkovic, Early Pregnancy and Gynaecology Ultrasound Unit, Department of Obstetrics and Gynaecology, King's College Hospital, Denmark Hill, London SE5 8RX, UK.


The aim of this study was to investigate the prevalence of congenital uterine anomalies in 1046 women attending gynaecological ultrasound clinics for a variety of indications. Using three-dimensional ultrasound, anomalies were found in 55 women (5.4%), including 32 (3.1%) with an arcuate uterus and 23 (2.3%) with major anomalies. The prevalence of uterine anomalies was similar to the findings in women undergoing elective sterilisation, but lower than in studies of women with recurrent miscarriage.

The true prevalence of congenital uterine anomalies in the general population is not known; previous studies reporting on uterine anomalies examined women with a history of infertility or recurrent miscarriage. One of the reasons for this pre-selection of study populations was the invasive nature of the tests for the examination of the uterus including hysterosalpingography, hysteroscopy and laparoscopy. Three-dimensional ultrasound, that has recently been introduced into clinical practice, enables noninvasive and accurate diagnosis of congenital uterine anomalies1. We used this new technique to determine the prevalence of congenital anomalies in women attending gynaecological ultrasound unit for a variety of indications.


This was a prospective observational study involving pelvic ultrasound examination in 1046 women attending our gynaecological ultrasound unit. The women were either self-referred for ovarian cancer screening or they were referred for a scan by their consultant gynaecologists or general practitioners because of suspected pelvic abnormalities (Table 1). Exclusion criteria were pregnancy, previous hysterectomy or myomectomy, and referral for investigation of infertility or recurrent miscarriages.

Table 1.  Indications for ultrasound examination and prevalence of congenital uterine anomalies. Values are given as n or n (%). IUCD = intrauterine contraceptive device.
IndicationnArcuate uterusMajor anomalies
Screening for ovarian cancer2536 (2.3)4 (1.6)
Pelvic pain2807 (2.5)7 (2.5)
Abnormal uterine bleeding2508 (3.2)6 (2.4)
Clinical suspicion of fibroids1266 (4.8)4 (3.2)
Clinical suspicion of ovarian cyst1034 (3.9)2 (1.9)
Localisation of IUCD341 (2.9)0 (0)
TOTAL104632 (3.1)23 (2.3)

Each woman first underwent a conventional B-mode transvaginal ultrasound examination using a 7.5 MHz mechanical probe (Combison 530 3D Voluson, Kretztechnik, Zipf, Austria). The longitudinal axis of the uterus, from the isthmus to the fundus, was first defined and then a series of transverse sections were obtained. When there was any duplication or splitting of the endometrial echo, fusion anomalies were suspected and in these women a three-dimensional ultrasound scan was used to determine the diagnosis. In the three-dimensional ultrasound examination the uterus was visualised in a longitudinal plane and a three-dimensional volume was generated by the automatic rotation of the mechanical transducer through 360°. The volumes, in the shape of truncated cones with a depth of 4.3–8.6 cm and a vertical angle of 90°, were analysed on-line using the technique of computer-generated planar reformatted sections. Although the images were similar to those of conventional two-dimensional sonography, with this technique it is possible to obtain an unlimited number of sections through the uterus which cannot be seen on routine scans. The acquisition of three-dimensional volumes is about 10 s, but complete analysis is achieved in 3 to 10 minutes.

Congenital uterine anomalies were classified according to the criteria of the American Fertility Society2. Thus in an arcuate uterus the fundus is normal but the uterine cavity is concave, in a septate uterus the cavity is completely or partially divided by a septum, and in a bicornuate uterus there is a fundal cleft of at least 1 cm in depth. Cases with rudimentary uterine horns were classified as unicornuate uteri.

Demographic details and ultrasound findings were entered into a computer database at the time of examination and this database was used to obtain details on the age of the women, indications for the scan and the ultrasound findings on uterine morphology. The significance of the association between the prevalence of congenital uterine abnormalities and both the indication for the ultrasound scan and the age of the women was examined using the χ2 test. For age analysis the following groups were used: 19-29 years, 30-39 years, 40-49 years and > 49 years.


Successful examination of the uterus was achieved in 1022/1046 women (97.7%). The uterine cavity could not be clearly visualised due to the presence of large calcified anterior fibroids in 12 women, thin atrophic endometrium in nine women, an intrauterine contraceptive device in two women and previous endometrial resection in one woman. These 24 women were excluded from further analysis.

Congenital uterine anomalies were observed in 55 women (5.4%), including 32 (3.1%) with an arcuate uterus and 23 (2.3%) with major anomalies; there were 16 women with a subseptate uterus, four with a bicornuate uterus, two with an unicornuate uterus and one with uterine agenesis. There was no significant association between the prevalence of uterine anomalies and either the indication for referral (χ2= 1.76; P > 0.05; 5 df; Table 1), or the women's age (χ2= 2.4; P > 0.05; 3 df).


In this study of women attending a gynaecological ultrasound unit the prevalence of all congenital uterine anomalies was 5.4%, while the prevalence of major anomalies was 2.3%. Although our study population may not be representative of the general population, as most women attended for the scan because of a specific complaint, there was no significant association between the prevalence of anomalies and either the indication for the scan or the age of women.

These findings are compatible with those of three previous studies that attempted to define the prevalence of congenital uterine anomalies in low-risk populations by examining women undergoing elective sterilisation. Cooper et al.3 found congenital uterine anomalies in 20 (6.2%) of 323 women at hysteroscopic sterilisation, Ashton et al.4 reported major anomalies in 16(1.9%) of 840 hysterosalpingograms following transcervical tubal sterilisation and Simon et al.5 found major anomalies in 22 (3.2%) of 679 women that had a hysterosalpingogram five months after sterilisation by laparoscopy or laparotomy. Our findings are also compatible with the 1 % to 7% prevalence of uterine anomalies in studies utilising laparoscopy or hysterosalpingography to investigate women with infertility. In contrast, as expected, our prevalence of anomalies was lower than the 10% to 25% reported in studies of women with recurrent miscarriages.

The advantages of three-dimensional ultrasound in the study of uterine morphology are first, the technique is noninvasive, secondly, it provides quantitative information on parameters such as fundal indentation or septum length, which is not possible with the endoscopic methods or hysterosalpingography, and thirdly, the stored ultrasound data may be re-sliced in a way which provides standardised views of the uterus which is also impossible with the traditional techniques. In addition, three-dimensional ultrasound may improve our understanding of the pathophysiology of conditions such as recurrent miscarriage in the context of uterine anomalies and help to identify those women that may benefit from surgery; recent reports have shown that the prognosis in individuals with the same type of uterine anomaly is varied and have questioned the value of metroplasty in the treatment of multiple pregnancy loss6. The disadvantage of three-dimensional ultrasound is that it may not provide adequate examination in the presence of large uterine fibroids. Additionally, the technique is currently confined to a few specialised centres.