Australian and New Zealand Journal of Obstetrics and Gynaecology 2005; 45: 187–190 187 Blackwell Publishing, Ltd. Original Article Paravaginal defects Paravaginal defects: a comparison of clinical examination and 2D/3D ultrasound imaging Hans Peter DIETZ, 1 Selina PANG, 2 Andrew KORDA 3 and Christopher BENNESS 3 1 University of Sydney, Western Clinical School, Penrith, 2 Prince of Wales Hospital, Hong Kong SAR, China, and 3 Royal Prince Alfred Hospital, Sydney, Australia Abstract Background: Paravaginal defects are often assumed to be the underlying anatomical abnormality in anterior compart- ment descent. Neither clinical examination nor ultrasound assessment are generally accepted diagnostic modalities. Aims: To compare clinical examination and translabial 3D ultrasound in the detection of such defects. Methods: Fifty-nine women without previous prolapse or incontinence surgery were seen prospectively. Clinical and ultrasound assessments were carried out in blinded fashion. 3D translabial ultrasound was undertaken after voiding and supine. Volumes were acquired at rest, on Valsalva and on levator contraction. Loss of paravaginal support (‘tenting’) in the axial plane was taken to signify paravaginal defects. Results: Paravaginal defects were reported clinically in 14 cases on the left (24%), 19 times on the right (32%). Two 3D ultrasound examinations did not yield satisfactory volumes, leaving 57 for analysis. Neither midsagittal nor coronal views yielded data that correlated with clinical assessments. In the axial plane there was absence of tenting at rest in 32/57 (57%) patients, but this did not correlate with clinical findings. Loss of tenting on Valsalva was observed less often (21/57, 37%) and was weakly associated with clinically observed lateral defects (P = 0.036). Conclusions: Pelvic floor ultrasound in midsagittal, axial or coronal planes does not correlate well with clinical assess- ment for paravaginal defects. This could be due to poor clinical assessment technique or limitations of the ultrasound method. On the other hand, paravaginal defects may be uncommon or clinically irrelevant. On present knowledge, the paravaginal defect has to be regarded as an unproven concept. Key words: 3D ultrasound, incontinence, paravaginal defects, prolapse, translabial ultrasound. Introduction Paravaginal defects, that is, gaps in the endopelvic fascia, have been postulated as the underlying anatomical abnormality in anterior vaginal wall descent. 1–3 Cystocoele and stress urinary incontinence is thought to be due to defects in the endopelvic fascia, supposedly as a result of traumatic childbirth. 2 Two main types of defect have been hypothesised- the lateral paravaginal defect, due to separation of the endopelvic fascia from the arcus tendineus fasciae pelvis, and the central defect due to development of a central gap in the endopelvic fascia supporting bladder neck, trigone and posterior bladder wall. The integrity of pelvic fascial structures is notoriously difficult to assess however, even at surgery, and there is no agreement at present regarding the validity of the concept. 4 Usually, assessment is undertaken by clinical examination, evaluating the depth of the vaginal fornices and the presence/ absence of vaginal rugae. To date, this technique has never been fully validated. In order to make detection of paravaginal defects more objective, some authors have used transabdominal ultrasound, with conflicting results. 5–8 Paravaginal defects are supposed to be present when an asymmetry of the caudal contour of the bladder is observed in an oblique coronal view. However, as both uterine position and size as well as rectal filling can influence results and as it is impossible to standardise the assessment plane due to an absence of land- marks in the coronal plane, it is not surprising that the method has been found unreliable. 7 Furthermore, while one would expect defects to become more pronounced and more easily identifiable on Valsalva, transabdominal ultrasound does not allow imaging on Valsalva due to dislodgement of the transducer on contraction of the abdominal muscles. Correspondence: Assoc. Prof. Hans Peter Dietz, University of Sydney, Western Clinical School, Nepean Campus, Nepean Hospital. P.O. Box 63, Penrith, New South Wales 2751, Australia. Email: hpdietz@bigpond.com Received 12 July 2004; accepted 4 December 2004.