Posterior colporrhaphy: its effects on bowel and sexual function


  • Margie A. Kahn,

    Research Fellow , Corresponding author
    1. Urogynaecology Unit, Department of Obstetrics and Gynaecology, St. George's Hospital, London
      Correspondence: Dr M. A. Kahn, Urogynaecology Unit, Department of Obstetrics and Gynaecology, St. George's Hospital, Blackshaw Road, London, UK SW17 0QT, UK.
    Search for more papers by this author
  • Stuart L. Stanton

    1. Urogynaecology Unit, Department of Obstetrics and Gynaecology, St. George's Hospital, London
    Search for more papers by this author

Correspondence: Dr M. A. Kahn, Urogynaecology Unit, Department of Obstetrics and Gynaecology, St. George's Hospital, Blackshaw Road, London, UK SW17 0QT, UK.


Objective To determine the anatomical cure rate of posterior colporrhaphy and its effect on bowel and sexual function one to six years later.

Design Retrospective observational study.

Setting Urogynaecology Unit, St George's Hospital, London.

Participants Two hundred and thirty-one women who underwent posterior colporrhaphy.

Main outcome measures Anatomical and symptomatic cure of rectocoele.

Methods The charts of 231 women who underwent 244 posterior colporrhaphies between 1 January 1989 and 4 January 1994 were reviewed. One hundred and seventy one (74%) were interviewed; 140 (61%) were examined. Mean follow up time was 42.5 months (range 11–74).

Results Two hundred and nine women had prior or concurrent vaginal and/or bladder neck surgery including 38 previous posterior colporrhaphies. Postoperatively prolapse symptoms due to rectocoele decreased (64%vs 31%). Constipation (22%vs 33%), incomplete bowel emptying (27%vs 38%), incontinence of faeces (4%vs 11%) and sexual dysfunction (18%vs 27%) increased. Those with incontinence of stool were more likely to have had two or more posterior colporrhaphies. Sixty-two percent felt that they improved over all after surgery. Additional postoperative symptoms included: vaginal and/or perineal splinting (33%), soiling and/or inability to wipe clean (16%), rectal digitation (23%), incontinence of flatus (19%), and rectal and/or vaginal pain (22%). Thirty-three women (24%) had large rectocoeles, seven of whom did not have impaired bowel emptying.

Conclusions Posterior colporrhaphy corrects the vaginal defect in 76% of women. It does not necessarily correct and may contribute to bowel and sexual dysfunction, particularly in those requiring multiple procedures. The presence of the anatomical defect does not imply dysfunction. The prevalence of bowel symptoms suggests the need for close questioning about bowel habits and the selective use of bowel investigations for some women before surgery.


The rectocoele is a herniation of the rectum through the rectovaginal septum causing a protrusion of the posterior vaginal wall. Symptoms include a lump, backache, incomplete bowel emptying, and the need to splint the vagina in order to empty the rectum. In 1867 Simon1 originated the term ‘posterior colporrhaphy’ to describe the operation which was then used to support the uterus in case of prolapse, but is now used by gynaecologists to cure the rectocoele. Dyspareunia is a major post-operative complication1–3. The effects of posterior colporrhaphy on bowel function are less well documented. Initially coloproctologists became interested in repair of the rectocoele because its neglect resulted in surgical failure of other anorectal surgery4,5. In 1967 Marks6 recognised that anorectal pathology may not be resolved by the vaginal approach alone and Pitchford7 recommended a combined transanal-vaginal approach. Numerous authors8–14 have since reported on the success rates and complications of transanal or combined procedures, but there is little information concerning anorectal symptoms in association with posterior colporrhaphy9. The purpose of this observational study is to estimate the anatomical cure rate of posterior colporrhaphy and its effect on bowel and sexual function.


From 1 January 1989 to 4 January 1994, the records of 231 women who underwent 244 posterior colporrhaphies by or under the supervision of the senior gynaecologist (S.L.S.) were scrutinised. One hundred and forty women (61%) agreed to a clinical examination and an interview in person, and a further 31 women (13%) underwent a telephone interview only. The remaining 60 women could not be located or did not agree to interview or examination. Sexual dysfunction was defined as coitus that was uncomfortable or impossible because of anatomical factors. All preoperative signs and symptoms were obtained from the women's records, including the degree of rectocoele, noted traditionally as first, second or third degree. Postoperative complications were also obtained from the case notes. At the long term follow up assessment, often because of vaginal narrowing, combined vaginal-rectal examination was performed. We have used ‘small’ to correspond to first degree and ‘large’ to correspond to second or third degree rectocoele. Anatomical cure was defined as the presence of no rectocoele or a small one at follow up examination. The mean follow up time was 42.5 months (range 11–74 months). The indications for repair were symptoms of a lump or incomplete bowel emptying or slackness at intercourse in the presence of a rectocoele.

Statistical comparison of preoperative symptoms and symptoms at the follow up assessment was by McNemar's test. Fisher's exact test was used to analyse differences in those who had two or more posterior repairs. The χ2 test with Yates' correction was used to detect any association between large rectocoeles and bowel symptoms. The statistical programme used was Arcus Pro-stat Version 3.28.

Surgical technique

On the morning of the operation two glycerine suppositories were given. One gramme cephadrine intravenously and one gramme metronidazole rectally were given with the premedication. The posterior colporrhaphy was performed through a vertical incision in the posterior vaginal wall. The levator ani muscles were exposed and sutured together. The most superficial of the deep sutures drew together the bulbospongiosus and transverse perinei muscles. Excess vaginal mucosa was trimmed prior to closing the vagina. Polyglycolic acid sutures were used throughout (number one for the levator ani and zero for the vaginal mucosa) except for perineal skin which was closed with No. 00 chromic catgut. A Foley catheter was inserted for 48 h. One to two sachets of ispaghula husk and 15–30 mL lactulose were given daily. The woman was discharged after the first normal bowel movement, usually on the fourth or fifth postoperative day.


The preoperative characteristics of the 231 women at the time of surgery were: mean age, 56 years (range 27–83); mean birthweight of heaviest infant, 3.7 kg (1.8–6); and median number of vaginal deliveries, 3 (0–8). Two hundred and eleven women (91%) had prior or concurrent vaginal or bladder neck surgery, including 38 (16%) previous posterior colporrhaphies. Twenty women had no other concurrent or previous procedures other than posterior colporrhaphy and only 13 of these 20 were able to be contacted.

Table 1 shows the comparison of preoperative symptoms and symptoms at the follow up assessment. Of the 171 women who were followed up, 109 complained of a preoperative lump; the remainder had had slackness at intercourse or incomplete bowel emptying. Of the 140 women who were examined at the follow up assessment, 51 (36%) with symptoms of a lump or pressure had no large prolapse other than rectocoele at the preoperative evaluation; the corresponding number postoperatively was 32 (23%). Of the 18 women who were incontinent of faeces, eight had two or more posterior colporrhaphies. Statistical analysis showed a strong association between faecal incontinence and a history of more than one posterior colporrhaphy (χ2= 47, P < 0.0001). Other symptoms did not show such an association. One hundred and six women (62%) felt that their bowel symptoms had improved after surgery; 21 (12%) experienced more difficult defaecation. Forty-four (26%) had no change in their bowel function.

Table 1.  Pre operative symptoms and symptoms at follow up (n= 171). Values are shown as n (%).
SymptomPre-operativeFollow upDe novoχ2P
  1. *n= 140; McNemar's test.

Lump, pressure109 (64)53 (31)16 (9)35.6< 0.0001
Lump, pressure (rectocele only, on examination)*52 (36)32 (23)15 (11)6.60.01
Incomplete bowel emptying46 (27)65 (38)38 (22)6.30.02
Constipation38 (22)56 (33)39 (23)4.80.03
Faecal incontinence7 (4)18 (11)14 (8)5.90.01
Sexual dysfunction30 (18)47 (27)27 (16)6.90.009

Except for the symptom of a lump or pressure, each symptom for which preoperative data existed showed increased prevalence at the follow up examination (Table 1). De novo means development of a symptom post-operatively that was not present pre-operatively. Insufficient preoperative data were available for comparison in Table 2. If the 60 women who did not participate had no symptoms whatsoever, the prevalence of follow up problems would still be significant. For example, 57/231 (25%) would have had to perform digitation. Any selection bias would underestimate rather than overestimate the problems.

Table 2.  Additional symptoms at follow up interview (n= 171).
Symptomsn (%)
Vaginal/perineal splinting57 (33)
Manual emptying40 (23)
Vaginal pain38 (22)
Incontinence flatus32 (19)
Difficulty wiping clean27 (16)
Faecal urgency22 (13)
Diarrhoea13 (8)

At the follow up examination, there were 33 women (24%) with large rectocoeles, which gives an anatomical cure rate of 76%. Of the 107 women who had an anatomical cure, 69 (64%) had impaired bowel emptying (e.g. constipation, incomplete emptying, rectal digitation or splinting). Of the 33 failures, 26 (79%) had impaired bowel emptying. This did not correlate with the presence of a large rectocoele (χ2= 1.75, P= 0.19).

Thirty women (13%) had one or more subsequent operations: correction of other utero-vaginal prolapse (n= 15); posterior colporrhaphy (n= 13); perineotomy or scar revision (n= 9); and rectal mucosal prolapse operation or haemorrhoidectomy (n= 5). Examination at the follow up assessment showed that 35 (25%) women had one or more other types of prolapse: cystocoeles (n= 17); uterine (n= 7); and enterocoele (n= 12).


This retrospective study shows that posterior colporrhaphy is associated with an increase in bowel and sexual dysfunction. This may reflect the limitations of a retrospective study, progression of intrinsic bowel dysfunction, or pelvic floor disorder or the effect of the surgery itself. Chronic constipation has many causes15 and may be associated with the development of a rectocoele16,17 and with uterovaginal prolapse18. The women in this study had bowel symptoms at the follow up examination which were independent of the presence of rectocoele. Repair of the rectocoele does not always solve and may contribute to the functional problem, either producing bowel symptoms or aggravating bowel disorders already present, particularly faecal incontinence, which was strongly associated with two or more posterior colporrhaphies. For women with preoperative bowel symptoms it would be wise initially to prescribe bulking agents, lactulose, and suppositories or to carry out bowel investigations to evaluate the significance of their bowel disorders.

Since 211 women (91%) underwent prior or concurrent vaginal or bladder neck surgery it could be argued that it was this surgery rather than the posterior colporrhaphy which was responsible for the bowel and sexual dysfunction; for example, colposuspension is known to predispose to the development of posterior wall prolapse19. However, posterior colporrhaphy is an operation that is designed to improve bowel function, and if other surgery contributed to the follow up bowel problems, clearly posterior colporrhaphy did not solve them. Other authors have argued that sexual problems are more common with posterior colporrhaphy than with other vaginal surgery1,20.

As all the posterior colporrhaphies were undertaken or directly supervised by the same gynaecologist, it could be argued that the results of the study would not be general, and that better results may be obtained by other gynaecologists. However, the senior author has a specific interest in vaginal surgery, and his technique is the same as described in many textbooks21–33. Current collaboration with a colorectal surgeon has emphasised the importance of bowel symptoms before and after pelvic and vaginal surgery.

The most recent longer term study9 examined 35 women with rectocoele repaired transanally and 29 women repaired by posterior colporrhaphy at four years and found that 54% complained of constipation, 34% had partial incontinence, 17% complained of persistent rectal pain and 22% had vaginal tightness or sexual dysfunction at follow up, similar to our experience. Eighty percent reported improvement after surgery. The only difference between posterior colporrhaphy and transanal repair was the greater frequency of postoperative pain with the posterior colporrhaphy.

Long term sexual dysfunction with posterior colporrhaphy has many causes: increasing age and postmenopausal atrophy together with vaginal surgery may cause sexual dysfunction1–3,20. Over half of the women in our study were postmenopausal at the time of surgery and less than one-third were on hormone replacement therapy at follow up. Most had had other vaginal or bladder neck surgery.

Levator muscle plication is thought to be a fundamental part of posterior colporrhaphy21–33. The lateral and cephalad extent of muscle bites are variously described in the literature, either as ‘very wide’ and ‘deep’ or as ‘very superficial’ and just capturing that portion of levator fascia which inserts into the perineal body which is the perineorrhaphy34. This part of the operation is thought to cause pain and dyspareunia28,35 because of pressure atrophy of the included muscle fibres35 and subsequent scarring.

An alternative technique to improve bowel function is by transanal plication of the defect in the inner circular muscle after dissection of the rectal mucosa8,10–14. With this operation subjective improvement of bowel function has been reported in 80% to 98%8,10–14. A recent article by Mellgren et al.36 quotes an 88% improvement in bowel function in 24 constipated patients one year following posterior colporrhaphy. Sullivan et al.8 did not report any dyspareunia following transanal repair, but Arnold et al.9 reported a 21% prevalence following both procedures.

Attempts have been made to refine the criteria for rectocoele repair using anorectal investigations. Anismus (paradoxical sphincter contraction during straining) diagnosed by puborectalis electromyography, balloon expulsion, or defaecography has not been shown to affect the outcome of rectocoele repair37. Directed repair of the defect using proctography has also been disappointing38. A large volume required to produce an urge to defaecate in faecally incontinent women without previous pelvic surgery is the only physiological measurement predictive of a good clinical outcome for transanal repair39.

The current classification of grading the size of rectocoeles is inadequate since a narrowed introitus can interfere with visualisation. Rectal examination is necessary to determine size and to identify coincidental enterocoeles. Defaecography can help identify unsuspected rectocoeles and enterocoeles40 although correlation of symptoms with amount of barium retention has yet to be shown41.

The strength of the evidence of an association between posterior colporrhaphy and the development of bowel and sexual dysfunction in this study is limited, since the quality of the information in the women who came for an examination may be better than that of women who were interviewed over the telephone, and this combined with the lack of information from the nonresponders and the weakness of retrospective data may make some of the findings less certain. However this study suggests that the hypothesis that posterior colporrhaphy can cause bowel and sexual dysfunction should be tested further in a prospective observational study including anorectal physiological measurements and imaging. We have started this study.


M.A.K. was supported by a grant from the South Thames Regional Health Authority Research and Development Directorate.