Management of the posterior compartment


Dr A. Monga, Princess Anne Hospital, Southampton University Hospitals Trust, Southampton, UK.


A rectocele is a herniation of the rectum through the recto vaginal septum causing a protrusion of the posterior vaginal wall. The prevalence of rectocele ranges from 20% to 80% of the general population.1 Hendrix et al.2 performed a cross-sectional analysis of women who enrolled in the women's health initiative of a hormone replacement therapy clinical trial. The total population was 27,342 women. At baseline pelvic examination the prevalence of rectocele was approximately 18%, whether or not the women had undergone a prior hysterectomy. African/American women in this particular study demonstrated the lowest predisposition for prolapse. Parity and obesity were strongly associated with the increased risk for rectocele. However, whether or not black women are protected from prolapse remains a contentious issue as a Nigerian study has shown rectocele may be present in up to 11.4% of women over the age of 40.3 The natural history of prolapse has been very poorly studied, Handa et al.4 performed annual pelvic examinations on women over a 2–8-year period. The annual incidence of rectocele was 5.7 cases per 100 women years. While the progression rate for stage 1 rectocele was 13.5 per 100 women years there was a regression rate of 22 per 100 women years. This shows spontaneous regression is common in stage 1 prolapse. It is therefore very important that stage 1 prolapse be managed conservatively in the first instance as many will resolve.

Rectocele was once thought to be a condition affecting only multiparous females and resulting from obstetric damage or increased tissue laxity with ageing and menopausal atrophy. However, rectocele has been observed in asymptomatic volunteers during defaecography.5 Obstetric events are usually the major predisposing factor. Traumatic delivery and damage to nerve, muscle and connective tissue will result in altered function and anatomical position of the pelvic tissues. The rectal fascia may separate from the perineal body causing a low transverse defect, although higher isolated defects in various parts of the recto vaginal tissue are also described.

Chronic constipation and straining may also cause damage to the recto-vaginal septum over time and result in a rectocele.

Finally, posterior vaginal prolapse is extremely common after colposuspension, which is likely to change the vaginal axis and cause a strain on the posterior vaginal wall detaching it from its fascial attachments.Table 1 shows aetiological factors in the development of a rectocele.

Table 1.  Aetiology of rectocele.
Obstetric injury
Poor episiotomy repair
Disordered defecation
insensate rectum
Poor connective tissue


Denonvilliers first described a layer of fascia he found in males, which he termed the ‘recto-vesical septum’. Nichols and Milley later recorded the existence of the septum in surgical dissections and autopsies of fresh female cadavers in a personal communication quoted by Richardson,6 who states that ‘the recto-vaginal septum and uterosacral ligaments provide suspensory support of the perineal body from the sacrum’. Posterior to the rectovaginal septum lies the rectovaginal space, which provides a plane for dissection. In between the rectovaginal septum and the rectum is the pararectal fascia; inside this fibrous muscular layer lie blood vessels, nerves and lymph nodes, which supply the rectum. The para-rectal fascia, originating from the pelvic sidewalls, divides into anterior and posterior sheaths, which encompass the rectum. These layers provide additional support to the anterior rectal wall. The fascia extends downwards from the posterior aspect of the cervix and the cardinal–uterosacral ligaments to attach to the upper margin of the perineal body and laterally to the fascia over the levator ani muscles. Further support is then provided by the levator ani basal tone maintaining a closed urogenital hiatus. DeLancey7 has further substantiated this. He dissected 42 fresh and 22 fixed cadavers and also performed histological and macroscopic sections. He confirmed the presence of this fascia. He also stated that muscular pelvic floor closure helped to relieve fascial stress and prevent damage occurring to the tissue. Shafik et al.8 have suggested that women with rectocele have a defect in their abdomino–vaginal pressure transmission and have lower resting vaginal pressures than healthy volunteer controls. This needs further substantiation.



Women with rectocele may be completely asymptomatic and therefore this may be an incidental finding at the time of examination. Symptoms may be nonspecific or specific to a rectocele. The woman may complain of a lump or a feeling of discomfort or dragging, she may have low back pain, which can be attributable to the prolapse. Common complaints include constipation, although this may be longstanding and affect the whole of the lower bowel. Other symptoms include a feeling of incomplete bowel evacuation with a need to splint either the vagina or the perineal area and sometimes even digitation of the rectum is required to empty the bowel. This often leads to increased bowel emptying frequency and may result in pocketing and soiling. It is not infrequent for a woman to report anal incontinence as a result of pocketing. Many women will not volunteer these symptoms and may need careful sensitive history-taking. A rare symptom is anal incontinence during intercourse; any prolapse might restrict coitus or lead to complete abstinence.


Examination is usually performed in the dorsal supine position or the left lateral position. After inspection of the external genitalia, the cervix and vaginal vault and vaginal walls are assessed, e.g. for atrophy. The prolapse is usually examined at maximum strain or Valsalva manoeuvre and staged according to the classification system of the International Continence Society. A recto-vaginal examination can help to demarcate a rectocele and will help identify defects. It also helps to differentiate rectocele from enterocele and will allow the perineal body to be assessed. Often there is no perineal body and therefore at surgery attention may need to be paid to this area. Other concomitant pathology must also be expected, e.g. stress incontinence is present in 24% of women with rectocele9 and Thompson et al.10 have shown that 33% of women have coexistent ano-rectal prolapse.

One of the problems with vaginal assessment is that the normal movement of the recto-vaginal axis is unknown and there are very few reliability or reproducibility studies available examining this concept. Furthermore the accuracy of clinical evaluation when compared with surgical evaluation is poor. Burrows et al.11 showed the sensitivity and positive predictive value of clinical evaluation for all defects was less than 40%.


A number of investigations have been performed to try and improve the diagnostic sensitivity of rectocele. It is debatable as to how helpful these investigations are.

Defecating proctocography

A technique that involves filling the rectum with barium paste with the consistency of stool and opacifying the vagina (Fig. 1). Fluoroscopy is performed with the patient at rest, on contracting the pelvic floor muscles, on straining down and then during defecation. Examination allows grading of the rectocele and the assessment of post-defecography trapping. Trapping is thought to be associated with incomplete bowel emptying. Although definitions differ, the definitions of Mellgren et al.12 have been widely utilised. If the maximum distension is less than 2 cm the rectocele is considered small, if it is more than 2 cm and less than 4 cm it is considered moderate and if it is greater than 4 cm the rectocele is considered large. However, there are problems with this and Shorvon et al.5 showed that 60% of normal volunteers have rectoceles greater than 2 cm distension.

Figure 1.

Defecography is probably a good diagnostic tool to help exclude anismus, which if present may predispose to poor surgical outcome. It also recognizes intussusception, which can cause constipation and requires different surgical treatment.

However, there are few reproducibility studies and the inter- and intra-observer errors are high.13,14 Freimanis et al.15 examined 21 asymptomatic volunteers and diagnosed rectocele during defecography in 67%. Only four managed to evacuate the barium paste and therefore results in symptomatic patients should be interpreted with caution. It has also been demonstrated that barium trapping does not correlate with the symptom of incomplete bowel emptying regardless of the volume retained.16,17 This was further substantiated comparing asymptomatic and constipated patients with no significant difference in trapping.18 Halligan and Bartram19 examined 11 women with rectocele, 11 with rectocele and trapping and 11 with no rectocele and no trapping and there were no differences in symptoms of incomplete bowel emptying.

Isotope defecography has been used as it allows the rate of evacuation to be assessed and the amount of trapped material to be quantified. The advantage of this technique still needs further evaluation.


Although ultrasound has been used to try and assess the size of prolapse it has not been very useful in this situation. However, endoanal ultrasound is an important investigation of management of women with anal incontinence and if an anal sphincter defect is present then this may require discussion with a colorectal surgical colleague to decide whether the rectocele should be repaired alone or in conjunction with an anal sphincter repair.

Anal manometry, electromyography, nerve conduction studies

Although these have been used to assess pelvic floor damage that may predispose to rectocele, they have little functional utility in the day-to-day management of patients with this condition.

Colonic transit studies

If a patient is thought to have chronic constipation and the constipation is widespread throughout her lower bowel, then transit studies demonstrating this may prevent unnecessary rectocele surgery. A repair may not help the patient with chronic constipation.



Recent attention to functional outcome after surgery has led to a more conservative approach first. Surgery does not always cure patient's symptoms and may lead to new symptoms such as dyspareunia. The major focus of conservative therapy is aimed towards providing symptom relief. This may involve the use of vaginal oestrogen therapy for atrophy. The other major symptom is altered bowel function and therefore attention is paid to improve patient's dietary fibre, the appropriate use of laxatives and suppositories, and improvement of defecatory habits. Women can be taught to splint their vagina and perineal body to try and improve defecation. Mimura et al.20 have shown that in 32 women with large rectoceles that a significant improvement can be obtained in bowel emptying by improving recto-anal coordination by using behavioural therapy in the form of biofeedback.

Heit et al.21 have performed studies evaluating clinical factors that may be predictive of treatment choice for patients with pelvic organ prolapse. The probability of choosing expectant rather than surgical management increased with higher preoperative pain scores and decreased as the severity of the prolapse increased. Older women were also less likely to choose surgery and patients that had prior prolapse surgery that had failed tended to choose conservative therapy. Signs and symptoms that are predictive of poor surgical outcome include preoperative pain and the use of laxatives.


Table 2 shows the aims for surgery as applied to rectocele repair. There are a number of surgical techniques that have been described to effect repair of a rectocele. The three most commonly utilised are posterior colporrhaphy, transanal repair and fascial defect repair. The fascial defect repair may rely on the closure of discreet defects or may actually involve plication of the fascia and usually this (probably) utilises some of the muscular tissues of the rectum. There is currently much debate as to the use of mesh and both synthetic and organic meshes have been tried. There are little data to support their use at present and studies have been small case series only. There is a paucity of good surgical outcome data in relation to rectocele repair. This is another area for further research.

Table 2.  Aims of surgery.
• Restore anatomy and function
• No deleterious effect on anatomy and function
• Good long term success rates
• Low complication rates
• Beneficial effect on an individuals quality of life
• Technique should be easy to learn and produce reproduceable results

The aims of surgery should be to restore anatomy and function without any deleterious effect and ideally procedures should have good long-term success rates and low complication rates. It is important to assess functional outcome as this has previously been ignored and also to demonstrate a beneficial effect on an individual's quality of life after an intervention. Ideally the techniques should be easy to learn and produce reproducible results when taught to others.

Posterior colporrhaphy

This procedure was very commonly used and involves a vaginal incision to perform levator ani plication. Usually absorbable sutures are used both for the levator ani plication and for closure of the vaginal mucosa after any redundant epithelium has been excised. There are very few published studies in the literature. Kahn and Stanton22 performed a retrospective observational study of 230 women undergoing levator ani plication repair for a rectocele over a 5-year period. There were 171 women interviewed and 140 examined with a mean follow-up time of 43 months (11–74 months).Table 3 shows the preoperative and postoperative symptoms and it is quite clear that symptoms of constipation, incomplete bowel emptying and faecal incontinence deteriorated and there was an increase in sexual dysfunction. Mellgren et al.23 in a small series of 25 women had already shown that bowel function improved but there was new dyspareunia in 20% of the study population. It is because of these poor functional results that many surgeons are no longer practising levator ani plication.

Table 3.  Symptoms before and after posterior colporraphy.
 Pre-op (%)Post-op (%)
  1. Kahn & Stanton, 1998.

Incomplete emptying2738
Faecal incontinence411
Sexual dysfunction1827

Transanal repair

The transanal approach was first used by Sarles et al.24 The technique involves an incision of the anterior rectal mucosa approximately 1 cm above the dentate line. The submucosal plane is sharply dissected 8–10 cm from the anal verge. The rectal muscle and recto-vaginal septum is then plicated usually using interrupted absorbable sutures. The mucosal flap that is created is then excised. Most of these procedures are performed by colorectal surgeons and their indications for surgery are often different to the rectocele encountered by gynaecologists. Disordered defecation is the most common symptom. No attention is paid to the vaginal lump. The studies are variable with varying lengths of follow up but success rates for improvement in bowel function are reported between 79% and 98%.25–29

If the patient uses preoperative vaginal digitation or rectal digitation they are thought to have a better outcome and Ayabaca et al.30 reported a 73% improvement in anal incontinence. Ho et al.31 have suggested that transanal repair may predispose to anal sphincter injury because of the stretching of the anal canal. In summary this technique appears to give good bowel function results on short-term follow up but there are little data regarding sexual function or the vaginal lump.

There are few comparative studies using the transanal vs the transvaginal approach but there is one limited randomised study32: 33 women underwent trans-anal repair and 24 transvaginal posterior colporraphy. Both procedures were equally effective for improvement in bowel emptying, not surprisingly dyspareunia was more frequently reported in the vaginal group. Many of the transanal group still complained of a vaginal lump.

Van Dam et al.33 reported on 89 women who underwent combined transanal and trans-vaginal repair. They reported 71% improvement in bowel emptying, 8% new faecal incontinence and 41% new dyspareunia. This combined procedure has not been further investigated.

Fascial defect repair

This technique involves incising the vagina in the midline and very carefully dissecting off all fascia and tissue so it lies on the rectal side. This tissue is then used to either plicate in the midline, or discrete defects can be identified at the time of surgery and closed. Rectal examination often facilitates both of these procedures. If any perineal body strengthening is required it can be performed with one index finger in the rectum.

A number of studies have been performed and have shown improvement in bowel function and sexual function and led to improvement in nonspecific symptoms of prolapse (Table 4 shows the outcome of four of these studies).34–37 The major benefit is the reduction in complications rather than a better long-term result. Obviously further follow-up studies are required.

Table 4.  Fascial defect repair
  1. Percentage figures pertain to improvement.


In our unit we have proceeded to perform fascial repair under local anaesthesia as a day case procedure without an anaesthetist or sedation. Dorflinger et al.38 followed a series of these women and equally good functional results were obtained under pure local anaesthesia compared with general anaesthesia. The reduction in hospital stay and quicker return to work and cost effectiveness of the procedure is being evaluated.

Newer techniques

There are a number of early studies looking at the use of Vicryl mesh, Prolene mesh and Porcine Dermis augmentation of the repair of rectocele. The short-term data are not very different to published series on standard fascial defect repair. Careful studies are required to evaluate the role of these meshes, as they may be an unnecessary expense and have inherent complication rates.

Colorectal surgeons have reported the use of a transanal stapling technique to repair rectocele. There are very early data and further investigation is required.


Rectoceles are common and pose a significant health problem. They are associated with nonspecific symptoms and also symptoms of sexual dysfunction and disordered defecation. Better obstetric management and episiotomy repair might prevent some rectoceles from developing, but significant numbers will still require surgery. The role of imaging and other investigations need careful evaluation, as their clinical role has not been defined. The outcome of conservative management is unknown and needs further study. There are little published medium or long-term data looking at surgical techniques and their outcome. There is a realisation that attention to functional outcome is important and this had led to development of the fascial defect repair. Further follow-up studies are necessary.