The pubofascial anchor sling procedure for recurrent genuine urinary stress incontinence


Mr Chung Department of Obstetrics and Gynaecology, All Saints Hospital, Magpie Hall Road, Chatham ME4 5NG, UK.



To evaluate, in a preliminary study, the outcome of a modified pubovaginal sling operation with titanium bone anchors for recurrent genuine stress urinary incontinence (GSI) in women.

Patients and methods

This prospective study included 13 consecutive women who underwent the modified sling procedure between September 1994 and August 1996. The subjective and objective cure of urinary stress incontinence, and the occurrence of postoperative osteitis pubis, were assessed.


All 13 patients subjectively claimed complete urinary continence and 12 were objectively cured (12 patients agreed to undergo a repeat urodynamic study) during a median (range) follow-up of 26 (19–38) months. There were no cases of postoperative osteitis pubis, bladder injury or major complications. Mild suprapubic pain was a frequent and self-limiting complication.


This innovative modified sling procedure is effective for recurrent urinary stress incontinence, with no complication of osteitis pubis. We suggest that this procedure should be considered as a treatment for recurrent GSI and perhaps for primary GSI. A study incorporating a longer follow-up and more patients has been planned.


More than 100 surgical procedures have been described for the treatment of genuine urinary stress incontinence (GSI) [ 1]. Colposuspension remains the preferred operation for GSI, as reflected in a survey among members and fellows of the Royal College of Obstetricians and Gynaecologists [ 2]. Black and Downs [ 3], in a systematic review, suggested that colposuspension appeared to be more effective and longer lasting than other procedures. However, the evidence is less convincing for repeat procedures. Recently, there has been a trend suggesting that sling operations should be considered for the treatment of recurrent GSI, as they seem to have a good long-term outcome [ 4]. This report describes our experience with a sling procedure using a harvested strip of rectus fascia suspended by prolene or Ethibond sutures onto the suprapubic tubercle with small titanium bone anchors. The subjective and objective outcomes were evaluated and complications associated with this procedure assessed, particularly the possible occurrence of osteitis pubis.

Patients and methods

From September 1994 to August 1996, 13 consecutive women with recurrent GSI underwent a modified pubofascial sling procedure using titanium bone anchors (Mitek G2, Mitek Surgical Products, Norwood, MA, USA) by the same gynaecological surgeon in a district general hospital. All patients underwent a full urogynaecological assessment including a detailed history and examination, urine culture and twin-channel subtracted cystometry with simultaneous pressure/flow voiding studies. The criteria for selecting patients included those who had previously failed procedure(s) for urinary incontinence and subsequently had urodynamically confirmed stress urinary incontinence in the absence of detrusor instability.

Surgical technique

The procedure was similar to the modified Pereyra procedure described by Raz [ 5]. The bladder was catheterized with a 16 F Foley catheter and an inverted U-shaped vaginal incision made under the bladder neck. The Cave of Retzius was entered vaginally and a suburethral tunnel created by blunt finger dissection. A 5 cm transverse suprapubic incision was made, then a 5×1 cm detached strip of rectus sheath harvested and used as the sling. It was placed loosely 2 cm below the urethra at the bladder neck, to act as a hammock. It is important that the sling does not compress the urethra (Fig. 1). Each end of the sling was sutured with prolene 1/0 Z-type suture to prevent the bunching of the rectangular end of the fascial strip. The Pereyra needle was guided through to the vagina from above, threaded with the prolene suture and brought out above, through the abdominal incision. This was repeated on the other side and the fascial strip aligned loosely under the urethra. Cystoscopy was performed only if the urine became bloodstained.

Figure 1.

a, A drawing of the pubofascial sling procedure with the Mitek bone anchors (not to scale) shown in b.

A hole was drilled into each side of the suprapubic tubercle, 1–2 cm lateral to the midline, using an orthopaedic drill fitted with a guard to prevent drilling deeper than 6 mm. The Mitek G2 titanium anchor (5×1 mm and resembling a fishhook, Fig. 1b) was then threaded with the loose end of the suture and placed into the drilled hole using an inserter. A loose hammock (a ‘sling on a string’) was thus formed underneath the bladder neck. Meanwhile, the vaginal incision was closed with a polyglactin 910 (Vicryl Rapide) 2/0 suture. The prolene sutures were then tied down over the assistant’s index finger tip to create a tension-free top-end suspension. The abdomen was closed routinely. A 16 F Foley catheter was inserted suprapubically using a 16 F Add-a-Cath introducer (Femcare Ltd., Nottingham, UK). This catheter was clamped after 48 h and removed if the residual volume was < 100 mL on two occasions. All patients received cefuroxime 1.5 g intravenously during the operation and two further doses later, followed by a 5-day course of oral cephradine (500 mg four times daily).

All patients were assessed at 2 months and again about 2 years after the operation. A repeat urodynamic study was arranged, and pelvic radiographs (brim and anteroposterior views, Fig. 2) taken during the 2-year follow-up. Subjective cure was defined when the patient had no complaints of stress incontinence, and objective cure if the urodynamic study continued to show continence and a negative pad test. The Wilcoxon test was used as the nonparametric method to analyse the urodynamic variables, with median values given wherever applicable.

Figure 2.

An anteroposterior pelvic radiograph showing the Mitek titanium bone anchors in situ. There is evidence of osteosclerotic changes at the fibrous joint of the symphysis pubis which are unrelated to the bone anchors.


The median (range) age of the patients was 56 (45–72) years and the median parity 2 (1–5); in all, 19 incontinence operations and 18 pelvic operations had been performed previously ( Table 1). The median (range) estimated blood loss was 300 (280–500) mL and the median operative duration 70 (45–120) min. There were neither bladder injuries nor any other major operative complications.

Table 1.  Type and number of previous incontinence and pelvic operations Thumbnail image of

Eight patients complained of postoperative suprapubic pain; seven settled within a month after the operation. In three of these patients, the hammock was probably tied too tightly under the urethra, as noted by the operating surgeon. The pain settled spontaneously with oral analgesia. Another three patients had UTIs and two had superficial wound infections, which probably accounted for their immediate postoperative suprapubic pain. One patient had mild chronic intermittent right-sided suprapubic pain which persisted beyond 6 months; no cause had been found and her symptoms were not severe enough to necessitate any treatment. The median (mean, sd, range) time before removal of the suprapubic catheter was 3 (6.8, 9.3, 2–36) days; one patient required urethral dilatation and urethrotomy for postoperative voiding problems. The median (range) length of stay in hospital was 5 (4–12) days.

At the 2-month follow-up, all 13 patients were subjectively continent and free of urinary symptoms; no major postoperative morbidity was noted. At the next follow-up (19–38 months), all 13 patients remained subjectively continent. All but one patient (who refused) underwent a repeat urodynamic study, giving an objective cure in 12 of the 13. Two women who had stress incontinence during the urodynamic study and a positive pad test claimed that they leaked slightly during their daily activities. Hence, all patients reported full satisfaction and success after the operation. No de novo detrusor instability was noted in the series. One patient reported having slight difficulty emptying her bladder, whilst another developed mild urgency and urge incontinence. There was a noticeable but not significant reduction in the bladder capacity and maximum flow rate after the operation (P>0.05) and an increase in the residual volume (P>0.05). No woman had regrets about having the operation, except one who, despite the successful outcome, regretted having any surgery at all, as previous repeated vaginal repair procedures had narrowed her vagina to such an extent that sexual intercourse was impossible.

The median time before resuming normal activity or returning to work was 7 (1–18) weeks. There were no cases of sling erosion or osteitis pubis, as excluded by the absence of translucency on a brim and anteroposterior pelvic radiograph after 2 years.


Sling procedures have been proposed as the treatment of choice, usually for recurrent GSI as well as for intrinsic urinary sphincter incompetence. These procedures, first described by Giordano [ 6] in 1907, have been modified in several ways, including the use of artificial and porcine tissues. The pubofascial sling was first described by McGuire et al. in the 1970s [ 7]. Synthetic materials such as Gore-Tex [ 8] and Marlex [ 9] have also been used as slings and claimed to provide both constant tension and durability. However, sling erosion could be a potential complication as a consequence of foreign-body reaction. Apart from rectus sheath, other autologous tissues, e.g. fascia lata, were used and reportedly gave a good postoperative outcome [ 10]. Success rates of sling procedures vary, but Jarvis [ 11], in a review, suggested that long-term objective success might be better than with a repeat Burch colposuspension. In the present series, 12 of 13 patients were objectively cured, with all 13 reporting subjective cure. Barrington et al. [ 12] also reported a high success rate in a recent series with the modified rectus fascial sling for the treatment of both primary and recurrent GSI. The apparent success of sling procedures may be attributed to the mechanism suggested in the ‘hammock theory’ proposed by DeLancey [ 13]; the sling functions like an automobile antibraking system, i.e. it only activates when there is a sudden increase in abdominal pressure. The laxity of both the sling and the top-end fixation ligatures is an important element of this procedure, although there is no published evidence to quantify the required tension.

This sling procedure combines aspects of the pubofascial sling procedure and the Marshall-Marchetti-Kranz (MMK) procedure. Jarvis [ 11] mentioned that the latter procedure had a high long-term success rate. The MMK procedure has not been widely practised in the UK because of the risk of osteitis pubis, the incidence of which was around 3% according to a recent review paper [ 14]. However, the definition of osteitis pubis was unclear in this review and it was not distinguished from symphysitis pubis and periostitis pubis. Osteitis pubis implies inflammation with or without an infectious cause. Clinically, it presents as a dull to excruciating pain which is often self-limiting and might last for many weeks. Osteolytic lesions will be apparent radiologically to confirm the clinical diagnosis. A literature review revealed that the older suture materials, including chromic catgut [ 15], silk [ 16] and linen [ 17] gave a higher complication rate as they promote an intense inflammatory reaction. The advent of inert suture materials, e.g. polypropylene and polyglycolic acid, contributed to the decrease in osteitis in the 1990s. The postoperative suprapubic pain noted in the present series was partly attributable to the causes mentioned, but also to periostitis pubis as a consequence of shaving of the periostium of the inferior pubic ramus on entering the Cave of Retzius during the initial dissection.

The titanium Mitek G2 bone anchors are widely used in orthopaedic procedures of the shoulder [ 18] and foot [ 19], and their safety is well proven. They are also being used in head and neck [ 20] as well as orthodontic operations, with no evidence of osteitis (personal communication). A potential criticism of this procedure is the need to place these small titanium anchors into the suprapubic bone, as it constitutes an unnecessary surgical risk with possible serious bone infection as a long-term sequel. We believe top-end fixation with bone anchors paradoxically avoids the severe postoperative suprapubic pain incurred with other needle suspension procedures because of the erosion of the suture through the rectus muscle. However, mild suprapubic pain from infection or a tight ligature was common, and settled within a month after the procedure. In addition, there were no cases of osteitis pubis in the present small series. We have so far performed more than 110 modified Pereyra procedures using Mitek G2 bone anchors, with a total of 220 bone anchors being inserted, with no evidence of osteitis pubis since 1994 (unpublished). Leach [ 20] showed that bone fixation of suspension sutures into pubic bone decreased postoperative pain and improved long-term success in the modified Pereyra procedure, compared with other fixation points such as rectus fascia. No patients reported significant chronic pubic bone pain after surgery in this study.

In conclusion, this modified sling procedure is effective for recurrent GSI because it has a good success rate after 2 years, with low peri- and postoperative morbidity, particularly as the anticipated osteitis pubis was not a problem. However, mild suprapubic pain was common after surgery. Thus sling procedures using autologous tissue should be considered as a treatment for recurrent stress urinary incontinence and even for primary GSI. We intend to review this group of patients yearly for at least 5 years and to accumulate more patients.


We thank Mrs Margaret Hughes for her secretarial support and Sister Piper McNickle for her expertise in performing the urodynamic studies. This paper was supported by a grant from Ethicon Ltd, UK.