Bernard T. Haylen

Professor UNSW


Featured Publications



Bernard T. Haylen

St Vincent’s Clinic, Sydney. Australia.

As published in the International Urogynaecology Journal (on line)
The original publication is available at International Urogynecology Journal – 2006, 17;6:555-558


The retroverted uterus has been largely ignored in urogynecological research to date. The prevalence of the retroverted uterus is 79% more common in the urogynecological patient population (34%) than in the general gynaecological population (19%). Its diagnosis requires the use of (i) transvaginal ultrasound with (ii) an empty bladder.

Recent data demonstrates that the prevalence of grade 2-4 uterine prolapse for a retroverted uterus is 4.5 times that for an anteverted uterus. Alternatively, 69% grade 2-4 uterine prolapse involves the retroverted uterus.


The retroverted uterus has been largely ignored in urogynecological research to date. Until now, there has been only one citation in the urogynecological literature [1]. There are only 2 other citations referencing the retroverted uterus in association with bladder dysfunction [2,3].

The remaining 71 of a total of only 74 references in the peer-reviewed and published literature concerning the retroverted uterus address general obstetric and gynecological issues. Obstetric issues, related to incarceration of the gravid retroverted uterus, occur in 12 papers e.g [4]. Associated gynecological issues involve (a) the additional technical difficulty in chorionic villus sampling (5 papers: e.g. [5]); (b) the greater likelihood of failure with endometrial resection and thermal balloon ablation (5 papers: e.g.[6] ); (c) pelvic pain and dyspareunia treated with laparoscopic ventrosuspension (15 papers: e.g. [7]); (d) increased difficulties with embryo transfer ( 4 papers: e.g. [8]); (e) issues regarding the insertion and performance of IUD’s and diaphragms (3 papers: e.g [9]); (f) sonography and surgical issues (18 papers: e.g. [10]); (g) infertility issues (6 papers: e.g. [11]); and (h) abortion issues (3 papers: e.g. [12]).

Retroversion implies that the axis of the body of the uterus is directed towards the hollow of the sacrum away from its normal (anteverted) position overlying the bladder [13]. The longitudinal axis of the anteverted uterus is approximately at right angle to the vagina [14] whereas the axis of the retroverted uterus tends to be similar to that of the vagina [Figure]. The cervix of the retroverted uterus is anteriorly placed in the vagina close to the bladder or urethra as opposed to the position of the cervix of the anteverted uterus lying in the posterior fornix and directed infero-posteriorly. If angulation of the corpus on the cervix is extreme, the retroverted uterus might be additionally or alternatively termed “retroflexed” [15].



Examples of anteversion and retroversion. From Llewellyn-Jones D. Fundamentals of Obstetrics & Gynaecology. Vol 2 Gynaecology. 1970. Fig 21/1. P202. Faber & Faber. London. Reproduced by permission of Prof Suzanne Abraham.


Where clinical examination allows, an anteriorly placed cervix and a clearly retroverted fundus can be palpated. Bladder volume should be empty. Too often, the version of the uterus is clinically unclear.Ultrasound provides the most accurate diagnosis though conditions for optimal imaging had not been clarified. In a recent study of 204 women attending a subspecialist gynaecological ultrasound unit [16], it was determined that an accurate diagnosis of a retroverted uterus required (a) the use of transvaginal ultrasound because of the superior resolution and the proximity of the probe to the area of interest and (b) a completely empty bladder. These results confirm the identical observations of Freimanis and Jones [17]. It was noted that with 6 women [16], the uterus was anteverted with a full bladder and retroverted with an empty bladder (a retroverted uterus never became anteverted).

There were no clear prevalence figures for the retroverted uterus in gynecological patients. The above study [16] using the outlined diagnostic criteria determined that 39 of the 204 women had a retroverted uterus, a prevalence rate of 19% in a general gynaecological patient population. This compares with a mean prevalence figure of 20% (range 10-38%) from 15 of the aforementioned 74 citations [2,4,5,8,9,13-15,18-24] which contained prevalence figures. All these latter studies, however, were not systematic.

Similarly, there were no prevalence figures for women with a retroverted uterus in a urogynecological patient population. In a further study [25] of 592 women attending for the first urogynecological assessment including urodynamics, it was determined that 133 (34%) of women with uteri in situ had a retroverted uterus and 262 (66%) women had an anteverted uterus. Thus the prevalence of the retroverted uterus in a urogynecological patient population is 79% more frequent than in the general gynecological patient population. The main reason for this, as will be further clarified below, is the much greater incidence of prolapse amongst women with a retroverted uterus.


The retroverted uterus is normally a developmental occurrence [15] though acquired retroversion can occur with the effects of endometriosis, pelvic inflammatory disease and pelvic tumours (13,15). The retroverted uterus appears to have a familial tendency.

There are strongly held traditional theories, however, that a much greater percentage of retroverted uteri may be acquired, hypothesizing that an anteverted uterus can often become retroverted as it prolapses [23]. There is no evidence for this theory in the literature. In the author’s clinical practice over the past 16 years, with each of over 10,000 new urogynecology patients, all of whom were subject to transvaginal ultrasound (bladder empty), there has been no objective evidence to support this theory. A majority of women presenting with a prolapsing retroverted uterus will have been given some indication of this orientation of their uterus in the past, with no woman, so far, having produced clear prior ultrasonic evidence of an anteverted uterus.

It is acknowledged that a longitudinal study is required to prove that uterine retroversion is mainly developmental with little acquired component.


Developmental retroversion is symptomless [15]. However, due to the greater association with prolapse, women with the retroverted uterus are more likely to experience pelvic dragging sensation or pain, sacral backache (due to taut or turgescent uterosacral ligaments [13]) or eventually a more obvious lump. It is well recorded that the anteriorly placed cervix is far more likely to be impinged upon with intercourse [7], especially if it is prolapsing, causing dyspareunia on deep penetration.

Research connecting the retroverted uterus with the different urinary symptoms is not available. The following have been author’s clinical observations. Women presenting with cyclical symptoms of lower urinary tract dysfunction are more likely to have a retroverted uterus. Cyclical symptoms of voiding difficulty are typical of the retroverted uterus representing most likely a degree of extrinsic impingement of the anteriorly placed cervix on the urethra, most commonly premenstrual. This effect generally serves to exacerbate some degree of underlying voiding difficulty. Some cyclical urgency can accompany these symptoms.

As a further observation, the symptom of stress incontinence tends to be different in women with a retroverted uterus. There is a slightly protective effect of the retroverted uterus akin to that of that of mild prolapse making occult stress urine leakage more likely than is the case for an anteverted uterus. Direct coughing (applying downward pressure on the uterus) might enhance this protective effect leading to reduced leakage than the equivalent situation with an anteverted uterus. Leakage is exposed more with jolting movements such as running, jumping and jolting-type sports, where the protective effect of the retroverted uterus is lost. That protective effect can be also lost with a fuller bladder akin to the situation with prolapse [26].

Coital leakage may often be a reflection of the interaction of a retroverted uterus with underlying diagnosis of urodynamic stress incontinence. The protective effect of the retroverted uterus on the latter diagnosis is lost as the cervix of the retroverted uterus is impinged upon with intercourse.


Due to the anteriorly placed cervix, the anterior vaginal wall of women with a retroverted uterus can be observed to be functionally if not actually shorter than those with an anteverted uterus. This has some surgical implications which will be discussed below. In contrast the posterior vaginal wall tends to be functionally longer as there is no cervix filling the posterior fornix as is the case with the anteverted uterus. However, no significant increase has been demonstrated in the incidence of posterior vaginal wall prolapse in women with retroverted uteri compared to women with anteverted uteri [25].

In seeking the sign of clinical stress leakage, due to the slightly protective effect of the retroverted uterus, this examination might be best performed with retroverted uterus reduced, e.g in the left lateral position and using a Sims speculum. Incidently, access to the anteriorly-positioned cervix for the purpose of obtaining a cervical smear is often easier with the patient also in the left lateral position.


The greatest relevance of the retroverted uterus is in relation to its close association with prolapse. Of the 395 women with uterus in situ [25], the prevalence of grade 2-4 uterine prolapse in women with a retroverted uterus was found to be 4.5 times that in women with an anteverted uterus. Sixty-nine percent of Grade 2-4 uterine prolapse was found to involve the retroverted uterus. There were also greater prevalences a Grade 2-3 cystocoele (1.9x) and at least grade 1 enterocoele (4.7x) in association with a retroverted uterus than is the case with an anteverted uterus.

Women with a retroverted uterus [25] presented earlier (median age 52) with urogynecological problems than their equivalents with an anteverted uterus (median age 56) or an absent uterus (median age 64). There was no significant difference in women with anteverted and retroverted uteri in relation to parity.


It was noted by Symmonds that a uterus that is in a retroverted position is especially subject to prolapse; with the corpus aligned with the axis of the vagina, anything increasing intra-abdominal pressure exerts a pistonlike action on the uterus, driving it down into the vagina [13]. This would seem a very logical mechanism, one likely to be behind the majority of uterine prolapse and much vaginal prolapse associated with that uterine descent. In contradistinction, intra-abdominal pressure on an anteverted uterus tends to force it infero-posteriorly towards the rectum, where prolapse is far less likely.


There were no significant differences demonstrated [25] between women with (i) an anteverted uterus (ii) a retroverted uterus and (iii) an absent uterus in relation to the major urodynamic diagnoses: urodynamic stress incontinence, overactive bladder, voiding difficulty (abnormally slow and/or incomplete micturition) or recurrent urinary tract infections (2 or more in the previous 12 months).


The tendency towards a shorter anterior vaginal wall for women with a retroverted uterus carries surgical implications. There is less vaginal wall available for elevatory continence procedures such as colposuspensions. It is more likely the uterine cervix (if present) might be drawn into a difficult position behind the “ridge” created by the colposuspension, jeopardising the longer term success of the procedure. Fortunately, the availability of effective non-elevatory sling-type continence procedures such as the tension-free vaginal tape avoids this issue.


The retroverted uterus has been ignored for too long. It’s greater prevalence in the urogynecological than in the general gynecological population reflects its much greater association with uterine and other associated vaginal prolapse.

For 69% uterine prolapse and much associated vaginal prolapse to be associated with uterine retroversion, that entity becomes “core” to an appreciation of prolapse. Abdominal “pulsion” forces in an “axial” direction provide the most logical pathogenesis of at least the uterine prolapse component.


The literature search of Vanessa Logan, the clinical assistance of fellow nurses Louise Verity and Serena Schulz, the ultrasound review of subspecialist in gynaecological ultrasound, Glen McNally and the comments of colleagues including Dr Surya Krishnan are gratefully acknowledged.


  1. Haylen BT and Cerqui A (1999) Postpartum uterine retroversion causing bladder outflow obstruction: Cure by laparoscopic ventrosuspension. Int Urogynecol J 10:353-355.
  2. Weinberger MD and Julian TM (1995) Voiding dysfunction and incontinence caused by uterine retroversion. A case report. J. Reprod Med 40:387-390.
  3. Lenck LC, Albuisson E, Jacquetin B (1990) Correction of uterine retro-deviation with a celioscope. Operative technique. Complications. Effects on urinary function. Revue Francaise de Gynecologie et d’Obstetrique 85; 11: 603-610.
  4. Smalbraak I, Bleker OP, Schutte MF, Treffers PE (1990) Incarceration of the retroverted gravid uterus: a report of four cases. Eur Journ Obstet Gynecol Reprod Biol 39:151-155.
  5. Silver RK, MacGregor SN, Muhlbach LH, Kambich MP, Ragin A (1994)
  6. A comparison of pregnancy loss between transcervical and transabdominal chorionic villus sampling. Obstet & Gynecol 8;5:657-660.
  7. Amso NN, Fernandez H, Vilos G et al., (2003) Uterine endometrial thermal balloon therapy for the treatment of menorrhagia: Long-term multicentre follow-up study. Human Reproduction. 18; 5:1082-1087.
  8. Halperin R, Padoa A, Schneider D, Bukovsky I & Pansky M. (2003) Long-term follow up (5-20 years) after uterine ventrosuspension for chronic pelvic pain and deep dyspareunia. Gynaecol Obstet Invest 55: 216-219.
  9. Henne MB, Milki AA (2004) Uterine position at real embryo transfer compared with mock embryo transfer. Human reproduction. 19; 3: 570-572.
  10. Chi IC, Farr G, Dominik R, Robinson N. (1990) Do retroverted uteri adversely affect insertions and performance of IUD’s. Contraception 41: 495-506.
  11. Bernstein P (1995) A new uterine manipulator for operative laparoscopic hysterectomy. J Amer Assoc Gynecol laparosc 2;3:331-333.
  12. Pepperell RJ, McBain JC (1985) Unexplained infertility: a review. Brit J Obstet Gynaecol 92; 6:569-580.
  13. Fielding WL, Lee SY, Friedman EA (1978) Continued pregnancy after failed first trimester abortion. Obstet Gynecol 52; 1: 56-58.
  14. Symmonds RE. (1982) Relaxation of pelvic supports. In: Current obstetrics & gynaecological diagnosis and treatment. Editor: RC Benson. Lange Medical Publications. California. Chapter 12, p285-291.
  15. Shaw RW, Soutter WP, Stanton SL Eds (2003) Gynaecology Third Edition Churchill Livingstone, Philadelphia. p25.
  16. Llewellyn-Jones DE (1994) Uterovaginal displacements, damage and prolapse. In:Fundamentals of obstetrics and gynaecology. Editor: DE Llewellyn Jones Sixth Edition. Mosby. Sydney. Chapter 40, p285-290.
  17. Haylen BT, McNally G, Birrell W, Logan V, Verity L and Schulz S (2005) Is the retroverted uterus more common in urogynaecology than general gynaecology. Int Urogynecol J 16; (Suppl 2): S42.
  18. Freimanis MG, Jones AF (1992) Transvaginal ultrasonography Radiol Clin Nth Amer 30; 5:955-976.
  19. Bongers MY, Mol BWJ, Brolmann HAM. (2002) Prognostic factors for the success of thermal balloon ablation in the treatment of menorrhagia. Obstet Gynecol, 99; 6: 1060-1066.
  20. Nazar NA, Herve F, Vilos G et al., (2003) Uterine endometrial thermal balloon therapy for treatment of menorrhagia: long-term multicentre follow-up study. Human reproduction 18;5:1082-1087.
  21. Silver RK, Macgregor SN, Hobart ED (1992) Factors associated with multiple-pass procedure during chorionic villus sampling: a video analysis. Prenatal Diagnosis 12:183-188.
  22. Jackson D, Elliott JP, Pearson M (1988) Asymptomatic uterine retroversion at 36 weeks gestation. Obstetrics & Gynaecology 71;3:466-468.
  23. Kivijarvi, A. and Gronroos, M (1983) Positional factors of the uterus play a contributing part in IUD failure. Acta Obstet Gynecol Scand 62: 67-70.
  24. Thompson JD(1996) Retrodisplacement of the uterus. In Rock JA and Thompson J (eds) TeLinde’s Operative Gynaecology. Lippincott-Raven, Philadelphia, PA, USA pp67-70.
  25. Egbase PE, Al-Sharhan M, Grudzinskas JG (2002) Influence of position and length of uterus on implantation and clinical pregnancy rates in IVF and embryo transfer treatment cycles. Human Reproduction 15; 9: 1943-1946.
  26. Haylen BT, Logan V, Verity L, Schulz S, Zhou J (2005) What is the relevance of the retroverted uterus in urogynaecology. Int Urogynecol J. 16 (Suppl 2) S43.
  27. Yang A, Mostwin J, Genadry R, Sanders R. (1993) Patterns of prolapse demonstrated with dynamic fastscan MRI; reassessment of conventional concepts of pelvic floor weaknesses. Neurourol Urodyn 12;4: 310-311. 




Bernard T. Haylen, Dianne Avery, Tin Lok Chiu, Warwick Birrell

International Urogynecol J, 25(12):1665-1772; Neurourol Urodyn 33(6):900-901. Dual Publication


Introduction: Posterior (PR) vaginal compartment repairs have traditionally involved a subjective approach. We aim to quantify (PR-Q) such repairs using key anatomical indicators (KAI).

Materials and Methods:

At 50 consecutive PRs: (i) Perineal gap (PG); (ii) Posterior vaginal vault descent (PVVD); (iii) Mid-vaginal laxity (MVL vault undisplaced/displaced); (iv) Recto-vaginal fascial Laxity (RVFL) were measured. The total posterior vaginal length (TPVL) and from POP-Q, TVL, GH, Ap, Bp C, D were also measured. Surgical details deemed appropriate to each repair were recorded.


A mean preoperative PG of 2.5cm was reduced to 0.0cm postoperatively by excision (100% cases) with an average 21.6% increase in total vaginal length over that if the repair was commenced at the hymen. There was an average 25.0% reduction in the genital hiatus (GH).

Mean PVVD was (i) 5.3cm overall; (ii) 6.4cm for 31/50 (62%) undergoing sacrospinous colpopexy; (iii) 3.5cm for 19/50 (38%) with no ligamentous vault fixation. An approximate “cutoff” for PVVD of 5cm might assist differentiating cases where vault fixation may be desirable.

Up to 52% (1.4/2.7cm) of preoperative MVL displaced was due to vaginal vault descent. The MVL undisplaced (mean 1.3cm) may better guide vaginal mucosal trimming.

RVFL averaged just 0.8cmm with 22/50 (44%), RVFL generally 0.5cm or less, not requiring any RVF plicatory sutures.

Conclusions: It is possible to use KAI to assist the planning and execution of posterior vaginal compartment surgery. The PG, PVVD, MVL and RVFL can indicate surgical measures in the perineum, vaginal vault, vaginal skin and recto-vaginal space respectively. 


It has been suggested that posterior pelvic floor problems have historically been neglected in urogynecology1. The inaccuracy of clinical examination emphasizes the need to develop new clinical and/or imaging methods to evaluate posterior vaginal wall defects2, 3. At present the clinical findings must be considered a poor representation of surgical findings2.  As one of the two most common surgeries in gynecology and urogynecology4, 5, the posterior colporrhaphy has received relatively little analysis of how to optimize surgical technique. The traditional posterior colpoperineorrhaphy has been questioned as ineffective at relieving symptoms and for the potential to cause new symptoms6. Success rates for posterior repairs have been described as modest7, 8, 9.  The rates for recurrent rectocoele / posterior vaginal vault prolapse after posterior colporrhaphy are not inconsequential8, 9, 10. Such criticisms, we acknowledge, might seem a biased analysis of the literature, failing to recognize reports of successful posterior repair series, though it does indicate sufficient concern to warrant further analysis of the posterior vaginal compartment repair.

Surgery in the posterior vaginal compartment generally is the last part of multi-compartment vaginal pelvic organ prolapse (POP) surgery. Very often, it is preceded by an anterior colporrhaphy and possibly a vaginal hysterectomy. Perhaps because of the sequence of surgeries, there is the danger for time-factor or other reasons that less attention might be given to optimizing or even performing a “posterior repair (PR)”. There may at times be the attitude or observation “the back wall doesn’t look too bad” without a more careful inspection, let alone measurement, of the different potential defects. Numerous native tissue repairs have been suggested for the posterior compartment11. These are often general surgical repair techniques with no reference to any particular case-at-hand. Loss of any or all of Level I (vaginal vault), Level II (mid-vaginal) and Level III (perineal) support can lead to posterior vaginal compartment prolapse12. The posterior repair needs to assess and address defects in each of these areas.

We wish to nominate four key anatomical indicators (KAI) as a starting point for both planning individual surgeries and deciding which surgical component techniques might be appropriate or even necessary to achieve a sound anatomical and functional result. There has been, at times, a debate between two-layer (pre-rectal fascia and vaginal mucosa) and three-layer (the addition of sutures plicating the levator fascia) posterior vaginal repairs. A recent review11 concluded that midline fascial plication without levatorplasty was the procedure of choice for posterior compartment prolapse with no evidence to support the use of prostheses and grafts in this compartment. It might be the situation that such an approach overtreats some areas of the posterior compartment and ignores other areas.

Selecting the KAI individually is the challenge which must precede collating them into a schema. The POP-Q13, 14 has four main reference points for the posterior vaginal wall, Ap, Bp and Point C (representing the cervix, or vaginal vault if cervix is absent) and point D (posterior fornix if cervix is present) whilst the total vaginal length (TVL) and the genital hiatus (GH) might also be relevant. These are all referenced to the hymenal ring. There is, however, additional posterior compartment distal to hymenal ring at the top of the perineum which is not taken into account by the above POP-Q13, 14 measurements. It is this area that is most often the subject of defects following childbirth and/or episiotomy and/or previous posterior vaginal surgeries. It is also the area where there is no clear description as to where to begin the posterior compartment repair.

The following four anatomical indicators are nominated for consideration for terming them “key”. Some of these can only really be assessed accurately and all are best measured at the time of surgery for posterior vaginal prolapse. More often than not, there will be an anterior colporrhaphy and/or vaginal hysterectomy immediately preceeding this assessment. We call these four indicators: (i) the perineal “gap” (PG); (ii) posterior vaginal vault descent (PVVD); (iii) mid-vaginal laxity (MVL-vault undisplaced); (iv) recto-vaginal fascial laxity (RVFL). Each will be described in the methodology and medically illustrated. A surgical marker pen and surgical ruler is required for each (pre- and post) operative examination.

Patients and Methods

A consecutive series of 50 female patients were involved in the study. Each was assessed intraoperatively (pre- and immediately post) for consideration of a repair of any section (Levels I-III12) of the posterior vaginal compartment from posterior vaginal vault to perineum. Each had undergone at least an anterior colporrhaphy immediately prior to this intraoperative assessment with some having undergone an additional vaginal hysterectomy. Each was examined for the presence and size of the proposed key anatomical indicators (KAI). Each measurement was checked and verified by an observer/surgical assistant. Bladder was re-emptied prior to this examination to prevent any limitation of the extent of prolapse aspects by the presence of a full bladder15.

1: Perineal Gap (PG):

With the patient still anesthetized, in appropriate position, surgically prepared and draped, the line of the labia minora was visually followed posteriorly till the perineum is reached on either side. At this point, Moynihan (Littlewood) forceps were applied on each side. Gentle bilateral traction highlighted any deficient anterior perineum. Closer inspection allowed marking (by surgical marker) of the junction of the much thinned-out area medially and the start of thicker perineal tissue laterally, closer to the forceps. This section, between the surgical marks, we would term the perineal gap (PG). It is measured (surgical ruler) in centimetres (cm) to one decimal point. Figure 1 shows a diagram of how the PG might be assessed and measured.

KAI 1: Perineal Gap (PG) – Thinned out media area of the anterior perineum       







2: Posterior Vaginal Vault Descent (PVVD).

The total posterior vaginal length (TPVL) was measured from the centre of the perineal gap to the vaginal vault. This was recorded in cm (one decimal point). A point 1cm below and posterior to the vaginal vault line (prior hysterectomy) or junction of cervix and posterior vaginal vault (uterus-in-situ) in the midline was then gripped by a (third) Moynihan forcep oriented horizontally. Inferior traction was then placed on the forcep. The distance between the perineal gap and the point of attachment of the Moynihan under maximum displacement, the perineal gap Moynihan (PGMOYN) distance, was measured in cm (one decimal point).

The posterior vaginal vault descent (PVVD) was calculated by the formula:

PVVD = TPVL – PGMOYN (cm to one decimal point).

KAI 2: Posterior Vaginal Vault descent (PVVD):  Measurement bottom figure subtracted from measurement  top figure.



3: Mid-Vaginal Laxity (MVL- vault undisplaced):

The (high posterior vaginal) Moynihan was placed on slight tension towards the vaginal vault and steadied by a surgical assistant. The surgeon/ surgical assessor placed anterior traction (perpendicular to the posterior vaginal wall), using Gillies forceps, on the best estimate of the midline of the midpoint of the vagina supero-inferiorly (most accurately judged by half the TPVL down from the vaginal vault).  The MVL (vaginal vault undisplaced) was the length (cm to one decimal point) of the anteriorly displaced mid-vagina over the lateral vagina.

KAI 3: Mid-vaginal laxity: Laxity of the mid-vaginal skin.







A further note was made of the MVL (vaginal vault displaced) i.e. the mid vaginal laxity (cm to one decimal point). If instead of slight tension on the Moynihan towards the vaginal vault, there was inferior traction on the Moynihan and the MVL remeasured.

4: Recto-vaginal fascial laxity (RVFL):

Whilst there is some debate over the nature (existence) of this fascia overlying the rectum and beneath the vaginal mucosa11, this potential defect was nonetheless assessed once posterior vaginal surgery was commenced. The perineal gap (if present) had been excised and a midline incision had been made in the posterior vaginal wall from excised perineal gap up to the Moynihan (or vulsellum). Artery forceps support both sides of the posterior vaginal wall incision. The forcep at the apex of the posterior vaginal wall incision was again held under slight tension towards the vaginal vault. As with the MVL, the surgeon/ surgical assessor placed anterior traction using Gillies forceps (perpendicular to the recto-vaginal fascia) on the best estimate of the midline of the midpoint of the recto-vaginal fascia supero-inferiorly. The recto-vaginal laxity was the length (cm to one decimal point) of the any displaced recto-vaginal fascia (midline) over the lateral fascia.

KAI 4: Recto-vaginal fascial laxity: Laxity of the mid-vaginal fascia.






Notes were made of the actual different surgical components: (i) excision of perineal gap; (ii) excision of mid-vaginal (mucosal) laxity (A: 0cm; B: 1-5mm; C: 6-10mm; D: over 10mm on either side); (iii) number of recto-vaginal fascial (RVF) sutures employed; (iv) posterior vaginal vault support to the right sacrospinous ligament (sacrospinous colpopexy) using a Capio Slim device (Boston Scientific, Boston, U.S.A) and minimalized dissection right posterior fornix submucosally.

Additional notes were made of the findings immediately post-PR in terms of the (i) PG; (ii) PVVD; (iii) MVL (undisplaced/displaced); (iv) TPVL. Pre- and immediately post-PR measurements were made of POP-Q13, 14 measurements; (i) GH (genital hiatus); (ii) TVL (total vaginal length); (iii) Point C; (iv) Point D; (v) Ap; (vi) Bp.

It can be assumed repairs to reconstitute the posterior vagina mucosa were performed on all patients down to the surgical marks for the PG using a 1 Vicyl suture. For the perineum (including 1 or 2 deeper sutures), 2/0 vicryl rapide was used. Rectovaginal fascial plication (0 vicryl) was performed in the percentage of cases indicated. Dissection in the recto-vaginal space was minimalized or nil; in no case was levator or levator fascial plication used.

Institutional ethical approval was received for the study. All patients were subject to informed consent as part of this audit of surgical practice. Methods, definitions and units conform to standards jointly recommended by the International Continence Society and the International Urogynecological Society, except where specifically noted14.


Table 1 shows the demographic of the group of 50 patients including prior and intercurrent surgeries (apart from the intercurrent PR).

Table 2 shows the mean pre- and post-operative (cm to one decimal point) KAI: (i) PG, (ii) PVVD; (iii) MVL – vault undisplaced and displaced); (iv) RVFL as well as (v) the TPVL and the POP-Q measurements13, 14 of (vi) GH, (vii) TVL, (viii) point C, (ix) point D; (x)  Ap (xi) Bp.

Table 3 shows the surgeries performed. Average number of rectovaginal sutures inserted overall (50 women) was 2.2 or 3.8 for each of the 28 (56%) women receiving recto-vaginal fascial plication. Twenty-two (44%) women had no such suturing.  The width of vaginal skin excised on each side as part of the repair was approximately: (i) 0mm (8%); (ii) 1-5mm (72%); (iii) 5-10mm (20%); (vii) over 10mm (0%).


This study suggests a possible role for the PG, PVVD, MVL (undisplaced), RVFL as “key (anatomical/surgical) indicators (KAI)” to quantify specific defects in at different levels of posterior vaginal compartment surgery (PR), namely the perineum, posterior vaginal vault, vaginal mucosa and rectovaginal space respectively. From these indicators, specific surgical actions might be considered: (i) PG – consider excision; (ii) PVVD – consider SSC vault fixation if say the PVVD is over 5cm; (iii) MVL (undisplaced) – consider limiting vaginal mucosal trimming to under half this measurement bilaterally; (iv) RVFL – consider no dissection and no suturing if under 5mm. The TPVL (TVL to a lesser degree) was of some value in indicating the very little overall (1.4-2.2%) impact of the surgery on the posterior vaginal length. The TPVL is also needed to calculate the PVVD. Similarly the GH measurements confirmed anteriorized support to the perineum though no specific surgical points or actions. We saw no role as a surgical indicator for the POP-Q13, 14 measurements: Ap, Bp or Points C and D.  We acknowledge others may find these measurements useful intraoperatively..

The results challenge the traditional concept of a “rectocoele” as a Level II defect12 of rectovaginal fascia (septum)11, 16, 17, 18and accompanying vaginal mucosal laxity. Mean MVL (indisplaced) was only 1.3cm and mean RVFL was only 0.8cm. The results suggest that the “bigger” issues in the posterior vaginal compartment are at the perineum (Level III12 - mean PG of 2.5cm) and vaginal vault (Level I12 - mean PVVD of 5.3cm). The implication of this is that in the traditional posterior repair, the mid-vagina may have perhaps received to date more intervention than required with the vaginal vault and perineum perhaps receiving less surgical attention than needed. MRI evidence7 also supports the view that there is overall weakening and generalized deformation of pelvic floor tissues rather specific fascial defects, pointing to this as a reason for the relatively modest success rates of posterior vaginal repairs8, 9.

In commenting specifically on the proposed key surgical indicators we firstly discuss KAI 1: Perineal Gap (PG). The results indicate that in women undergoing surgery for posterior vaginal wall prolapse, a thinned out area of the anterior perineum, PG can be detected, measured and considered for excision in all patients, with the thicker lateral tissue to be brought into the midline for restoration of a thicker anterior perineum. This was a mean 2.5cm wide preoperatively reducing to 0.0cm postoperatively.  The results show vaginal length up to 21.6% longer than if the repair was started at the hymen with a 25.0% reduction in the GH. The GH has been noted to have a strong correlation with prolapse severity19. This restoration of the thickness of this area, the decreased GH and the accompanying anteriorization of the axis of the posterior vagina may be beneficial in terms of support for accompanying anterior colporrhaphies, with the latter seeming much higher in the vagina and less visible or invisible at the introitus post-PR. The new introitus is soft with no sign of a flap or other impediment to an anticipated return to comfortable sexual function. Further studies will be needed to confirm this. We have noted the restoration of a “ring” of labia minora around the entire vaginal introitus as a result of excision of the perineal gap.

 KAI 2: Posterior vaginal vault descent (PVVD): With the mean PVVD of 5.3cm overall and 6.4cm for 31/50 (62%) undergoing vault fixation using a sacrospinous colpopexy (SSC), the results indicate a significant posterior vaginal vault laxity (Level I defect12) in the majority of the women undergoing the PR. It is understandable that in the cases of those who did not have a concomitant hysterectomy that most would be cases of recurrent POP, 87% requiring SSC due to vault defects. In may initially be more surprising that 12/27 (44%) of those having a concomitant vaginal hysterectomy would require a SSC. We have previously suggested20 that the vaginal vault should be considered as an “area” rather than an “apex”, the former with anterior, posterior and lateral aspects. The majority of support from those uterine supportive ligaments, the uterosacral (USL) and cardinal, has been shown to be directed towards the anterior vaginal vault and wall with very little influence on the posterior vaginal vault and wall20, 21, 22. Even though the first pedicle of the vaginal hysterectomy, the cardinal-uterosacral confluence (CUSC)21would have been routinely attached and tied to the lateral vaginal vault bilaterally to provide post-hysterectomy support, additional posterior vaginal vault support was required, here by SSC, in 44% cases. A McCall culdoplasty procedure has also been used in this circumstance23 seemingly with satisfactory longer term results24.

There was a mean PVVD of 3.5cm for 19/50 (38%) with no ligamentous vault fixation. In these patients, one arm of the closing suture for the posterior vaginal mucosa was however left long so that it could be laid submucosally extending from apex to perineum. The mucosa was closed over that arm, thus giving the apex some looser non-ligamentous support. We acknowledge surgeons will have alternate non-SSC vault fixation techniques.

What level of PVVD indicates than posterior vaginal vault fixation, e.g. by an SSC is desirable. We would suggest an approximate “cutoff” of 5cm. Above 5cm PVVD, an SSC is more anatomically and surgically desirable; below that figure, the posterior vaginal vault support is less in question and it is much harder for the vaginal vault to reach the sacrospinous ligament. The 62% rate for SSC in our study suggests more vault fixation may be required generally. In a study by Kenton et al25, 71% cases involved anterior colporrhaphy and 45% posterior colporrhaphy, yet only 4% had vault suspensions (3% SSC and 1% sacrocolpopexy).

KAPD 3: Mid- Vaginal Laxity (MVL): In the anterior compartment, much of the mid-vaginal (Level II12) laxity (cystocele) has been shown to be due to vaginal vault descent26. We have shown, to our knowledge for the first time quantitatively by intraoperative measurement, that the majority (52%) of mid-vaginal mucosal laxity posteriorly is due to vaginal vault descent. This was demonstrated by the difference in measurements of MVL if the vault was reduced (MVL- non-displaced) versus the equivalent measure if the vault was under traction (MVL-displaced). There is the danger, if this is not recognized, with effective vaginal vault support, as necessary, of over-excising the posterior vaginal wall which might cause dyspareunia18 due to restricted dimensions. Lowder et al3 had shown a 30% correction of posterior vaginal prolapse defects by simulated clinical restoration of the vaginal vault support using a speculum.

With the mean MVL (non-displaced) of 1.3cm and the majority (72%) vaginal mucosal trimming in the 1-5mm group, the results suggest that trimming was “conservative” and limited to under half the MVL (non-displaced) on each side. With the postoperative MVL (non-displaced and displaced) both around 0.3cm, it is possible that the “under half MVL (non-displaced)” guideline for mucosal trimming can be improved upon.

KAI 4: Recto-vaginal fascial laxity (RVFL): The mean “fascial” defect with the posterior vaginal wall opened was only 0.8cm. Traditionally, as already noted, posterior vaginal defects including rectocoeles have been thought to involve primarily rectovaginal facial defects11, 16, 17, 18. Such defects appeared far less than expected with 22/50 (44%) not deemed to have sufficient RVFL to warrant a repair. This results in mid-vaginal component of the PR being one-layer (mucosa-only). In the other 28, the average number of sutures was 3.8.  It creates a much less invasive posterior repair if the entire recto-vaginal space is not opened by dissection only then to be followed by an attempt at restoration. Whilst this is an observational study of immediate preoperative and postoperative anatomy and not one of longer term surgical and functional outcomes, it does raise the possibility that standardized two or three layer repairs11 may be an “overtreatment” for the existing Level II12 defects.

This section of the study would have been significantly compromised unless a minimally invasive sacrospinous colpopexy (SSC) technique was available which, in the great majority of cases did not unduly disturb the recto-vaginal space. The use of the Capio Slim© suture applicator (Boston Scientific, Boston. U.S.A) allowed the insertion of an O Ethibond via a relatively small tunnel in the right fornix submucosally down to the right sacrospinous ligament. We plan to report on and illustrate this technique separately.

We believe one of the strengths of this study is to focus attention of surgical techniques for posterior vaginal prolapse appropriate to the measurements/defects present, rather than employing a “one-size-fits-all” standardized repair. The latter, by not being individualized to the patient’s anatomy, might leave some defects unresolved. It also has the potential to create new defects6. The weakness of the study is that it does not, at this stage, include longer term anatomical and functional outcomes. We will, in subsequent studies, be addressing these weaknesses. Other validation studies are also planned.  In this study we have used the PG as the introital reference point, rather than the hymenal ring, for the different measurements and as the starting point for the PR. It remains to be noted over time, whether this will prove beneficial functionally as well as anatomically, though we have no reason, at this stage, for suggesting otherwise.

The emphasis in publications on the cure of recto-vaginal defects as prime to posterior compartment surgeries was not supported by our study. The prime defect areas seemed to be perineal (100%) and vaginal vault (63%) with recto-vaginal fascial defects found in 56%. These did not appear to require a full recto-vaginal space dissection and the required suturing was generally not extensive. Limited defect-directed rectocoele repairs have been shown to have better functional27-30, though not necessarily anatomical11, results. In our view, the functional result might well be due to the more limited dissection. Follow-up surgical studies will be required to determine whether our more minimalist approach to the recto-vaginal space based on the limited defects encountered and thus less apparent risk of adverse functional outcomes (based on the defect-directed rectocele repair data27-30) will be as anatomically successful. It will have the advantage that the more numerous Level 112 and level III12 defects will have been addressed as required.

The importance of posterior vaginal compartment surgery may have been underestimated over time1. Three Levels of defects are applicable to the majority of PRs compared with perhaps only one Level (II) defect anteriorly, provided the vault component of any anterior defect is resolved by the posterior vault repair (e.g. SSC).. Not only do the component surgeries identified in this study address the defects of the posterior vagina and posterior vaginal vault, they may have an impact, if performed effectively, in providing ongoing support for the anterior vaginal wall and assisting in preventing recurrences. We encourage pelvic floor surgeons to consider the KAI we have outlined or at least to recognize the need to compartmentalize the different defects and PR surgical components. We have been encouraged by the reproducibility of the immediate postoperative anatomical results, in particular the reductions in PG, MVL and PVVD and GH as well as the maintenance of TPVL (TVL).


1: Soligo M (2007) Pasterior pelvic floor dysfunction: there is an immediate need to standardize terminology Int Urogynecol J 18:369-371.

2: Burrows LJ, Sewell C, Leffler KS Cundiff GW (2003) The accuracy of clinical evaluation of posterior vaginal wall defects. Int Urogynecol J 14:160-163.

3: Lowder JL, Park AJ, Ellisin R, Ghetti C, Moalli P, Zyczynski H, Weber AM (2008) The role of apical vaginal support in the appearance of anterior and posterior vaginal prolapse. Obstet Gynecol 111(1):152-157.

4: Smith FJ, Holman CDJ, Moorin, RE, Tsokos N (2010) Lifetime risk of undergoing surgery for pelvic organ prolapse. Obstet Gynecol 116(5):1096-1100.

5: Boyles SH, Weber AM, Meyn L. (2004) Procedures for pelvic organ prolapse in the United States, 1979-1997. Amer J Obstet Gynecol 191:108-115.

6: Kahn MA, Stanton SL (1998). Techniques of rectocoele repair and their effects on bowel function. Int Urogynecol J 9:37-47.

7: Lewicky-Gaupp C, Yousuf A, Larson KA, Fenner DE, DeLancey JOL (2010) Structural position of the posterior vagina and pelvic floor in women with and without posterior vaginal prolapse. Am J Obstet Gynecol 202(5) 497.

8: Fialkow MF, Newton KM, WeissNS (2008) Incidence of recurrent pelvic organ prolapse 10 years following primary surgical management: A retrospective cohort study. Int Urogynecol J 19(11)1483-1487.

9: Clark AL, Gregory WT, Smith VJ, Edwards R. (2003) Epidemiologic evaluation of re-operation for surgically treated pelvic organ prolapse and urinary incontinence. Am J Obstet Gynecol 189 (5): 1261-1267.

10: Whiteside JL, Weber AM, Meyn LA, Walters MD (2004) Risk factors for prolapse recurrence after vaginal repair Am J Obstet Gynecol 191:1533-1538.

11: Karram MM, Maher C (2013) Surgery for posterior vaginal wall prolapse. Int Urogynecol J 24:1835-1841.

12: DeLancey JOL (1992) Anatomic aspects of vaginal eversion after hysterectomy. Amer J Obstet Gynecol 166:1717-1728.

13: Bump RC, Mattiasson A, Bo K, Brubaker LP, et al. (1996) The standardization of female pelvic organ prolapse and pelvic floor dysfunction. Amer J Obstet Gynecol, 175(1):10 -11.

14: Haylen BT, Freeman RM, de Ridder D, Swift SE, Berghmans B, Lee J, Monga A, Petri E, Rizk D, Sand P, Schaer G (2010). An International Urogynecological Association (IUGA) – International Continence Society (ICS) Joint Report into the Terminology for Female Pelvic Floor Dysfunction. Neurourology & Urodynamics, 29:4-20. International Urogynecology J, 21:5-26.

15: Yang, A., Mostwin, J., Genadry, R., Sanders, R. (1993) Patterns of prolapse demonstrated with dynamic fastscan MRI; reassessment of conventional concepts of pelvic floor weaknesses. Neurourol Urodyn, 12(4): 310-311.

16: Richardson AC (1993) The rectovaginal septum revisited: its relationship to rectocele and its importance in rectocele repair. Clin Obstet Gynecol 36:976-983.

17: Richardson AC (1995) The anatomic defects in rectocele and enterocele. J Pelv Surg 1:214-221

18: Francis WFA, Jeffcoate TNA (1961) Dyspareunia following vaginal operations. J Obstet Gynaecol Br Commonw 68:1-10.

19: Ghetti C, Grerory WT, Edwards SR, Otto LN, Clark AL (2005) Severity of pelvic organ prolapse associated with measurements of pelvic floor function. Int Urogynecol J 16:432-436.

20: Haylen BT, VU D, Birrell W, Vashevnik S, Tse K (2012) A preliminary anatomical basis for dual (uterosacral and sacrospinous ligaments) vaginal vault support at colporrhaphy. Int Urogynecol J 23: 879-882.

21: Samaan A, Vu D, Haylen BT, Tse K (2014) Cardinal ligament surgical anatomy: cardinal points at hysterectomy. Int Urogynecol J 25(2):189-195.

22: Vu D, Haylen BT, Tse K, Farnsworth A (2010) Surgical anatomy of the uterosacral ligament. Int Urogynecol J. 21: 1123-1128.

23: McCall ML (1957) Posterior culdoplasty: surgical correction of enterocoele during vaginal hysterectomy; a preliminary report. Obstet Gynecol 10:595-602.

24: Montella JM, Morrill MY (2005) Effectiveness of the McCall culdeplasty in maintaining support after vaginal hysterectomy. Int Urogynecol J 16:226-229.

25: Kenton K, Sadowski D, Shott S, Brubaker L (1999) A comparison of women with primary and recurrent pelvic organ prolapse. Am J Obstet Gynecol 180:1472-1475.

26: Summers A, Winkel LA, Hussain HK, DeLancey JOL (2006) The relationship between anterior and apical compartment support. Am J Obstet Gynecol 194:1438-1443.

27: Cundiff GW, Weidner AC, Visco AG, Addison WA, Bump RC (1998) An anatomic and functional assessment of the discrete defect rectocoele repair. Am J Obstet Gynecol 179:1451-1457.

28: Glavind K, Madsen H (2000) A prospective study of the discrete fascial defect rectocoele repair. Acta Obstet Gynecol Scan 79:145-147

29: Porter We, Steele A, Walsh P, Kohli N, Karram MM (1999) The anatomic and functional outcome of defect-specific rectocoele repairs. Am J Obstet Gynecol 181:1353-1358.

30: Kenton K, Shott S, Brubaker L (1999) Outcome after rectovaginal fascial reattachment for rectocoele repair. Am J Obstet Gynecol 181:1360-1364.


Table 1: Patient demographic data (n = 50)

AGE 59.6
WEIGHT (KG) 68.6
HEIGHT (CM) 161.9
MENOPAUSE   40 (80%)
HRT   4
  17 (34%)
7/17 (41%)
10/17 (59%)
PREVIOUS VAULT SX (0-1) 3 (10%)


Table 2: Pre- and Post-PR measurements (average in cm) of key anatomical indicators (KAI) and equivalent POP-Q measurements (n =50).

  KAI   POP-Q  
  Pre-op Post-op Pre-op Post-op
Perineal Gap (PG) 2.5 0.0 (100%↓)    
Genital hiatus (GH)     3.6 2.7 (25%↓)
Total Post Vag Length (TPVL) 9.2 9.0 (2.2%↓)    
Total Vaginal Length (TVL)     7.3 7.4 (1.4%↑)
Post Vag Vault Desc (PVVD) 5.3 0.3 (94%↓)    
   - PVVD (SSC used)               6.4 0.0 (100%↓)    
   - PVVD (SSC not used)         3.5 0.8  (77%↓)    
Point C     -2.5 -6.3
Point D (only 7 patients with cervix present)     -4.1 -7.2
Point Ap     +0.1 -2.9
Point Bp     +0.2 -2.8
Mid-Vag Laxity (MVL:nondispl) 1.3 0.3 (77%↓)    
Mid-Vag Laxity (MVL: displ) 2.7 0.3    
Recto-vaginal fascial laxity (RVFL) 0.8      


SSC= sacrospinous colpopexy; nondispl= vaginal vault not displaced (i.e. held in placed); displ (vaginal vault on traction). 


Table 3: Surgeries performed in posterior vaginal compartment repair (n=50)

Excision of perineal gap 50 (100%) 0  (0%)
Sacrospinous colpopexy (SSC) 31 (62%) 19 (38%)
SSC if c/com Ant Rep only 20 ( 87%) 3 (13%)
SSC if c/com VH/Ant Rep 12 (44%)  15 (56%)
Rectovaginal fascial suturing 28 (56%) (ave 3.8 sutures) 22 (44%)
Excision of vaginal skin 46 ( 92%) (72% 1-5mm) 4 ( 8%)

c/com = concomitant

© Copyright Bernard T. Haylen 2017 -

Printed on: Sun 17 December 2017 at 1:17:01 PM AEDT