Introduction

Bladder outlet obstruction (BOO) is an uncommon entity in women, the incidence being only 2.7–8% of women referred for voiding symptoms [1]. Female urethral stricture (FUS) is a more distinctive subgroup of female BOO and accounts for 4–13% cases of female BOO presenting with lower urinary tract symptoms (LUTS) [2]. The lower incidence compared to male strictures is attributed to the shorter length, straighter course and greater anatomical mobility of the female urethra and the protection provided by the bony symphysis over its entire length [3]. Although true FUS is a rare subset of BOO, the incidence was 4% in the Nitti et al. series [4]. Urethral dilatation, often as an office ambulatory procedure, has been practiced for a long for a variety of nonspecific female voiding complaints [5]. However, various urethroplasty techniques have gradually gain ground in recent years for treating women with true FUS [6]. We aimed to report all the surgical interventions and their effectiveness in treating FUS in the literature to date. The primary outcome was to report the different interventional techniques reported in the literature for treating FUS. The secondary outcomes were the success rates and complications of the various techniques described.

Methods

Sources

All retrospective and few prospective case series are included in this review, which followed the PRISMA statement (Preferred Reporting Items for Systematic Reviews and Meta-analysis) [7]. We carried out an extensive electronic search through the PubMed and EMBASE databases. Search terms used were 'female or women' combined with the terms 'urethral stricture or urethral stenosis' or 'urethral dilatation' or 'urethrotomy' or 'urethroplasty' on January 2021. Filters used were 'humans, female, age more than 19 years and the English language'.

Study Selection

Retrospective or prospective case series of women with FUS who underwent surgical intervention were included. For inclusion, we define the female urethral stricture as ‘A fixed anatomical narrowing of the female urethra diagnosed by radiological evidence, urethral calibration or direct endoscopic visualization causing symptomatic voiding symptoms and dysuria'. Papers not conforming to our inclusion criteria of ‘female urethral stricture’ such as urethral fistulas, diverticulum repair, congenital urethral abnormalities and post-pelvic fracture urethral loss were excluded. Also, full-text articles not providing the outcome data separately and citations from conference proceedings were excluded. Cross-references from the selected articles were scanned and hand-searched to identify any missed article. A pro forma was made in duplicate to record the extracted data from each of the selected full-text articles and checked by two of the authors (JC and NV).

Participants

Women with idiopathic urethral strictures.

Intervention

Surgical intervention grouped into two main categories (urethral dilatation and augmentation urethroplasty with local flap, local graft or oral graft).

Comparison

Success rates among the different intervention techniques.

Outcome measures

Stricture-free rate, improvement of flow rate (Qmax) and complications.

Data extraction and quality assessment

Two of the authors (JC and NV) independently performed the abstracting, followed by the screening of full-text articles for quality assessment and results. Data include participant characteristics, intervention, outcome assessment tools and the results. Any dispute regarding the quality of the studies was resolved by the senior author, AC. Success was defined as ‘No requirement of further intervention or surgery after the primary intervention’. We assessed the risk of bias in the included studies by using the Oxford (Centre for Evidence-based medicine, 2009) levels of evidence [8] as well as the JBI (Joanna Briggs Institute) assessment protocol for case series [9]. Most of the selected studies were retrospective case series (Oxford level 4 evidence) with selection bias.

Evidence synthesis

Electronic search identified 106 abstracts which, after removing the case reports and conference citations, reduced to 64 abstracts. Subsequent hand-searching led to four more full-text articles. A total of 68 full-text articles were identified from years unspecified to up till now. Thirty-three papers were rejected because of duplicate data and inadequate outcome reporting. All in all, 34 articles were ultimately included for the review. From 34 studies, 488 individual patients were reported with outcome measures after undergoing intervention for urethral stricture disease (Fig. 1).

Fig. 1
figure 1

PRISMA flow chart

Full size image

Results

Methodological quality

There were consecutive recruitments of cases in five series only [22, 27, 28, 40, 41]. All of the studies were retrospective case series, except one that was prospective [30]. Although most of the series had a well-described follow-up of > 1 year, some had only < 9 months follow-up [30, 32]. The confirmation of diagnosis was made clinically and other methods such as VCUG, cystourethroscopy, urethral calibration and UDS or a combination of these modalities (Table 1). The outcomes were assessed both subjectively and objectively in all the studies by using valid instruments. Most of the studies had an adequate number of cases, except a few [14, 24, 28, 32, 35] with fewer than five cases each (included because of their pioneering works).

Table 1 Case series (34 studies) on FUS with effectiveness of different interventions and their outcomes
Full size table

Outcomes in definitions, diagnostic tests and aetiology of FUS

Female urethral stricture disease does not have any standardized definition or strict diagnostic criteria. Inclusion criteria widely vary in several of the available studies. The definition followed by Smith et al. [10] was ‘any fixed anatomical narrowing between the bladder neck and distal urethra of < 14 F preventing catheterisation and confirmed by urethrography or cystourethroscopy’.

In the majority of the studies, the authors used a combination of various diagnostic tests along with symptomatic assessment. The common among them was post-void residual (PVR), voiding cystourethrography (VCUG), uroflowmetry, urethroscopy and office urethral calibration. The video-urodynamic study (VUDS), transvaginal ultrasound (TV-US) and MRI were used more selectively. Urethral calibre < 14 F (measured in the office set-up) was regarded as diagnostic in some studies [10, 17, 19, 29], whereas in others, a lumen < 10 F [26], 12 F [39, 43], 16 F [31] or 17 F [11] was set as the calibre threshold. Maximal flow rate (Qmax) was regarded as a common criterion to diagnose and quantify the severity of stricture. A Qmax < 12 ml/s was set as the cutoff level in several studies [17, 19, 29, 30, 31, 32], whereas a value of 10 ml/s and 20 ml/s was the lower limit in two studies [26, 30], respectively. Voiding cystourethrography (VCUG) was used in most of the studies to confirm the diagnosis and stricture location radiologically as a ‘fixed anatomical narrowing between the bladder neck and distal urethra with ballooning of the proximal urethra above the stricture during voiding’. Postvoid residual urine (PVR) threshold value varies in several studies from 60 ml to 90 ml, or even 100 ml. Several studies used video-urodynamics (VUDS) or urodynamics (UDS) selectively, especially when there was high PVR, history of retention or high Qmax. Several authors used a Pdet Qmax value > 20 cm. H2O as the lower limit to define BOO, whereas others set it as 30 cm or even 50 cmH2O. In addition to Pdet Qmax, nomograms such as the AG nomogram, Schafaer linPURR diagram [15], Solomon-Greenwell pressure-flow nomogram [21] or Blaivas-Groutz nomogram [11, 36] were also used to diagnose female BOO. Urethroscopy was used in 25 studies to diagnose and evaluate the length or site of the stricture. MRI or transvaginal ultrasound (TV-USG) was used only occasionally (only in 5 series) for assessing the stricture (fibrosis and extent) or any local pathology. The diagnostic tests used to make the diagnosis are shown in Table 2.

Table 2 Diagnostic tests
Full size table

The exact aetiology of FUS is believed to be multifactorial and mostly idiopathic. The presumed aetiological factors were mentioned in 229 (46.92%) women out of a total of 488 cases. Of these, 214 (43.85%) women were presumed to have an iatrogenic stricture (prolonged catheterisation, traumatic vaginal delivery, gynaecological surgery), 10 (2.04%) post-traumatic strictures and 5 (1.02%) inflammatory strictures. The idiopathic variety constituted 50.20% (245 patients), whereas 14 (2.86%) patients were of unknown category.

Three hundred fifty-five patients had a history of prior intervention for urethral stricture disease. Of these, 347 (97.74%) women had a history of repeated urethral dilatation with or without urethrotomy, and 4 (1.12%) women had a history of prior urethroplasty. Among those patients with previous dilatation, two patients also had a cystostomy tube in place before the urethroplasty procedure. Two women (0.56%) had a prior history of urethral sling surgery.

Outcomes of surgical intervention for FUS

Urethral dilatation (with or without urethrotomy) and augmentation urethroplasty are the two main techniques described in FUS treatment. Augmentation is usually done by utilizing tissues from either the local area (vaginal or labial tissue as a flap or free graft) or distant site (oral mucosal free graft). The augmentation is done by putting the flap or graft either dorsally (12 o'clock position) or ventrally (6 o'clock position) in the urethra. The studies used different parameters, both subjective and objective, to assess the outcome. Subjective parameters were the symptoms, and the degree was measured by various quantification methods (PGI/S scale, AUA score, Likert scale, sexual function questionnaires). Objective parameters measured were uroflowmetry, PVR, urethral calibre size, cystourethroscopic findings, UDS and VCUG. Table 3 shows the 34 studies with different intervention techniques and their outcomes.

Table 3 Comparison and the outcomes and follow-up of the different techniques used in FUS management
Full size table

Outcomes of dilatation

The primary form of treatment in four studies was urethral dilatation [10,11,12,13] (Table 4). The urethra was dilated up to 30 F in 7 patients, 41 F in 93 patients and 39–43 F in 30 patients (a total of 130 patients), whereas in 7 women, it was unclear to what extent the dilatation was carried out. Fifty-three women had a history of prior urethral dilatation, whereas one had prior urethrotomy.

Table 4 Female urethral dilatation series
Full size table

In the series by Smith et al., all patients were advised to carry out daily CISC (intermittent self-catheterisation) [10]. Three of these seven patients did require further dilatation over and above CISC (in this context, the success rate would be much lower, only 43%). Romman et al. had a success rate of 51% (in the failure group, 39% had a history of previous dilatation compared to only 17% in the success group). Tearing of the meatus requiring hemostatic suture placement to the adjacent vaginal wall was associated with a better outcome as happened in 81% of the success group compared to 71% in the failure group [12]. The mean time to failure was 12 months. Duration of symptoms before presentation was found to be a significant factor [13]. There were no de novo incontinence reports or other significant complications in the dilatation series, except in the Popat et al. series [13], which reported a 17% incidence of urge incontinence after dilatation.

Overall, urethral dilatation’s success rate was 41.2% at a mean follow-up of 34.75 months. The success rate varies according to the number of previous dilatation sessions, as evident from the fact that the success rate comes down from 55.4% in women without prior dilatation to 14.9% with multiple previous dilatations (Table 4).

Outcomes of augmentation urethroplasty

Thirty-two studies reported outcomes after using various methods of augmentation urethroplasty totalling 351 patients. Vaginal or labial flaps were used in 11 studies (97 women) [11, 14,15,16,17,18,19,20,21,22,23,24,25,26,27,28,29]. Twenty-one studies reported the use of free grafts. Of these, 8 studies used local vaginal or labial grafts (85 women) [24,25,26,27,28,29,30,31], and 17 studies used oral graft (lingual or buccal mucosal), totalling 169 women [11, 19, 21, 24, 32,33,34,35,36,37,38,39,40,41,42,43]. The surgical approach was explained for each of the categories with graft placement location in the urethra (either ventrally at 6 o'clock or dorsally at 12 o'clock position).

Outcomes of vaginal or labial flaps

Vaginal or labial flaps have been used in 11 studies (97 women), as shown in Table 5. Ten studies (95 patients) described the use of vaginal flaps, whereas only one study (2 patients) described the labial flap method. Seventy of the 97 women had a history of UD or urethrotomy before urethroplasty. Tanello et al. reported using labia minora pedicled skin flap placed ventrally in the urethra in two patients with 100% success [14]. The rest of the studies used various types of vaginal inlay flap in U-, Y- or C-shaped fashion. Reversed U –shaped vaginal flap was used in 7 studies (63 women) [11, 16, 18, 19, 20, 21, 23]. The basic principle was the same: raising a reversed U-shaped vaginal flap anteriorly and placing this to the opened urethrotomy incision (inlay flap) at the 6 o'clock position. The lateral-based vaginal flap was used in two studies [14, 22]. Romero-Maroto et al. used the lateral-based flap based on Orandi’s male urethral stricture surgery [22]. The Y-shaped ventral vaginal flap was used in one study with a dorsal inlay (12 o'clock) approach [15]. In the same study, an upper urethral mucosal flap was used in one case. In this particular case, a mucosal flap from the proximally dilated urethra was utilized to augment the urethra ventrally (urethral mucosal flap urethroplasty). The C-shaped ventral vaginal flap was used in one study with a ventral inlay approach [17]. Seven patients underwent concomitant pubovaginal sling operation, four Martious flaps, one bladder neck collagen injection and diverticulectomy. De novo urge incontinence was reported in four patients (one received botulinum toxin injection), whereas stress incontinence was found in only two cases (one received trans-obturator tape). Recurrent UTI and inward urinary steam were reported in four cases.

Table 5 Female urethroplasty (vaginal and labial flap) series
Full size table

Overall, the mean success rate in these 11 studies (97 women) was found to be 92.54%, with a mean follow-up of 37.28 months. The definition of a successful outcome was different in several of the studies, and post-operative ISC (although not conforming to our definition) was done as a routine for some time during the follow-up period. Twenty of the 97 women were instructed to perform daily CISC post-operatively.

Outcomes of vaginal grafts

Table 6 shows the eight studies (85 patients) that reported the use of local tissue grafts (vaginal or labial) [24,25,26,27,28,29,30,31]. Dorsal onlay VG urethroplasty was done in five studies (66 women) [24,25,26, 30, 31], whereas labia minora graft (LMG) augmentation was done in three studies (19 women) [27,28,29]. In all VG studies, the graft was placed dorsally (12 o'clock position), whereas, in the LMG studies, it was placed ventrally (6 o'clock position). Repeated urethral dilatation was done in all 85 women (with additional urethrotomy in 10 and meatoplasty in 1 patient). Two patients had cystostomy pre-operatively. There was no incidence of de novo incontinence or the use of routine post-operative CISC. The reported successful outcome was 87.30% at a mean follow-up of 20.35 months.

Table 6 Female urethroplasty (vaginal and labial graft) series
Full size table

Outcomes of oral grafts

Table 7 shows the 17 studies (169 patients) where tissues from the oral cavity were used in augmenting the stricture [11, 19,20,21, 24, 32,33,34,35,36,37,38,39,40,41,42,43]. Most of the patients (165) had a prior history of UD or urethrotomy. BMG harvested from the inner cheek was used in 16 studies, although the graft application method was different. While 12 studies (82 women) used the dorsal onlay approach (12 o'clock), four studies (33 patients) used the ventral onlay (6 o'clock position) technique. In 22 patients with the ventral onlay technique, a concomitant Martius flap was used. In one study, an innovative ventral inlay approach was used [39]. According to the authors, ‘onlay’ means placing and quilting the BMG in the urethrotomy gap after fully mobilizing the urethra with a full-thickness U-shaped or circumferential perimeatal incision. ‘Inlay’, on the other hand, means placing the BMG in the gap produced by an intra-meatal longitudinal incision without fully mobilizing the full thickness circumferentially [39]. Circumferential BMG (tube graft) was used in one study (two patients) [19]. Dorsal lingual onlay graft was used in one study only [34] with 15 women. There was no incidence of de novo incontinence or significant complications. Routine post-operative CISC was not used.

Table 7 Summary of contemporary female urethroplasty (oral mucosal graft) series
Full size table

Most of the series reported a higher success rate (> 90%) [19, 24, 32, 34, 35, 39 40, 43]. However, interestingly, the most recent and largest retrospective series by Hampson (2019) reported only a 33% success rate with BMG [41]. The mean success rate was found to be 89.94% at a mean follow-up of 22.54 months.

Outcome in terms of complications

Incontinence, although uncommon, is the most common complication mentioned in all the urethroplasty series. Fifteen women (9.43%) experienced incontinence, of which 11 (73%) had stress and 4 (27%) urge incontinence. Most of the stress and urge incontinences were mild to moderate in nature and either stopped spontaneously or after pelvic floor exercise. Donor site morbidity was noticed in two cases of BMG urethroplasty. Other complications (transfusion, meatal stenosis or wound infection) were rare (incidence: one each). Recurrence or restenosis is considered a failure and is mentioned in Tables 3–7.

Discussion

In this comprehensive literature review, data for a total of 488 patients from 34 studies were available for analysis. All the studies on FUS have been published since 2000, indicating a growing interest in this uncommon entity. The principal findings are that the urethroplasty techniques have better outcomes (success rate: 87% to 92% at a mean follow-up of 26.34 months) than urethral dilatation (success rate: 41.25% at a mean follow-up of 34.75 months) that often results in rapid recurrence requiring further intervention in > 50% cases (Table 3).

Stricture aetiology was presumably identified in 47% women in our review (50% idiopathic, 44% iatrogenic, 2% trauma, 1% inflammatory). Although the idiopathic variety constitutes nearly 50% of cases [6, 10, 12], some recent series reported iatrogenic injury (instrumentation, surgery) as the most common cause followed by traumatic vaginal delivery [21, 41, 44]. The stricture site in women is usually the mid (mostly) and distal urethra [15, 30]. The symptoms of FUS are mixed irritative and obstructive [6]; the severity can be quantified by using the QOL score tools such as the PGI-S scale (Patient Global Impression of Symptom severity) [22, 23] or AUA score [15, 19, 27, 30, 36]. All the patients in our review had a long-term history of poor urinary flow and incomplete voiding.

A correct pre-operative diagnosis is a foundation for a correct operative approach and a successful outcome. However, the diagnostic criteria for FUS are still undefined. While Brannan et al. [45] suggested the pathological urethral calibre as ≤ 20 F, some others set the cutoff value as 12 F [23], 14 F [10, 19] or even 20 F [20, 25, 46]. Although urethral calibre seems to correlate with neither the severity of symptoms nor the surgical outcome, it may be valuable in assessing recurrence in a post-operative setting [47]. A low-flow, high-pressure pattern during voiding on UDS is suggestive of BOO. However, there are differences among authors regarding the cutoff values of Pdet Qmax (cmH2O) and Qmax (ml/s), for example: ‘> 42, < 9’ [4], ‘> 25, < 12’ [48], ‘> 20, < 15’ [49] or ‘> 35, < 15’ [50]. Most of the authors prefer to use simple uroflowmetry, urethroscopy, PVR estimation and VCUG as diagnostic tests. UDS is more useful to rule out bladder dysfunction in selected cases to help predict post-urethroplasty failure. MRI or TV-US may be considered for assessing the degree of periurethral fibrosis, diverticulum or any other local pathology [6].

In the absence of any clear consensus on the definition of FUS, we have used the criteria as followed by Osman et al. [6]: “A symptomatic, anatomical narrowing of the urethra based on a failure of catheterisation, urethral calibration, visual inspection, or endoscopy or radiography”. In this way, it is easier to decide upon the exclusion criteria such as meatal stenosis, primary bladder neck or a functional obstruction, neurogenic bladder, pelvic irradiation, urologic or gynecologic malignancy or trauma”. Accordingly, ‘success’ was defined as the ability to void after the procedure without further urethral dilatation (UD), whereas ‘failure’ was defined as recurrent LUTS requiring repeat UD, chronic CISC, additional surgery or urinary diversion [51]. In our opinion, an initial period of CISC after urethroplasty should not be regarded as a failure unless CISC is required on a regular or chronic basis to keep the urethral patency.

Urethral dilatation (UD) is usually the most frequently chosen intervention, often indiscriminately, because of its simple and straightforward nature [5]. However, repeated UD has the potential for further fibrosis and earlier recurrence, as evidenced by a higher failure rate in patients with prior dilatation [6, 12]. The stricture-free rate in different series varies from 14% to 59% [10,11,12,13]. Osman et al.’s review quoted a success rate of 47% (27.2% in patients with prior dilatation vs. 58% without previous dilatation) [6]. The corresponding figures in our review are 41.25% (55.4% vs. 14.9%). Overall, attempting UD as a first-line intervention is still relevant in FUS.

Rosenbaum et al. [48] suggested using optical urethrotomy (at the 3 and 9 o'clock positions) to minimize stricture recurrence. However, its long-term efficacy is unverified [47]. Ackerman et al. [52] recommended a judicious use of UD and urethrotomy; with rapid recurrence after UD, an earlier urethroplasty is advisable.

Although the decision to perform urethroplasty is essentially driven by the patient who wants to get rid of CISC and periodic dilatation, proper counselling is essential before surgery. Urethroplasty can be done using local vascularised flaps or free grafts (local or distant). Important factors determining the surgical approach depend on stricture length, site, length of the healthy proximal urethra, bladder neck integrity and surgeon’s experience [52, 53].

Vaginal or labial flaps are the most commonly used substitution female urethroplasty techniques with a mean success rate of 91% [6]. Advantages of local flaps include good mobility with excellent vascularity and can be easily harvested with minimal morbidity.

Vaginal grafts (VG) are also popular because they are hair-free, elastic, easily accessible and naturally wet with infection resistance. Also, VG is cosmetically sound and can be done under spinal anaesthesia. The mean success rate of VG was 80% [6]. Complications of VGs are vaginal narrowing and dyspareunia. Also, in patients with vulvovaginal atrophy, local tissue quality suffers. Labia minora graft (LMG) is as promising as that of BMG in quality and graft take. It is thin and can be easily harvested with minimal donor site morbidity or cosmetic effect. LMG is usually used as a ventral inlay at the 6 o'clock position with a mean success rate of 86% [27, 28]. In our review, the overall success rates of local flaps and grafts are 92% and 87%, respectively.

Buccal mucosal graft (BMG) is the most widely used augmentation material in recent series [21, 39,40,41,42], including our review. Although El-Kasaby et al. are credited with the first use of BMG in male urethroplasty [54], Barbagli [55] popularized the technique. The advantages of BMG are being wet and hair-free and having a good vascular pattern and graft take, minimal contraction and easy harvestability with low donor site morbidity. Most of the series reported a > 90% success rate [19, 32, 36, 39, 40, 43]. However, interestingly, the most recent and largest retrospective series by Hampson et al. in 2019 [41] reported a mere 33% success rate. Most of the authors applied a buccal graft as a dorsal onlay technique [11, 21, 36] instead of a ventral onlay [40, 43]. Lingual mucosal graft (LMG) was first used by Simonato et al. [47] for male urethroplasty. The graft is of thin, smooth oral mucosa and easily harvestable with minimum donor site morbidity [29, 34]. Oral grafts are the most commonly used tissue for urethral augmentation in our review, with a mean success rate of 90%.

The main concerns related to dorsal urethrolysis are the potential injury to the neurovascular bundle (NVB) and the striated sphincter. The potential advantages of the dorsal technique are decreased risk of incontinence, fistula or diverticular formation and meatal hypospadias. In addition, there is a more physiological voiding with a reduced risk of ventral scarring, thus facilitating any future anti-incontinence surgery. However, the potential disadvantage is NVB injury [11, 24, 32]. The ventral approach, on the other hand, is more prone to give rise to stress incontinence unless a synchronous pubovaginal sling is placed [11].

Irrespective of the urethrolysis technique used, the potential surgical complications are haematoma, wound gaping or infection. However, new-onset incontinence (mild to moderate) is the most common complication, which resolves spontaneously or with pelvic floor exercise and rarely needs corrective surgery.

In their review, Osman et al. [6] showed that UD had the lowest mean success rate of 47%, followed by a local flap rate of 91%, local graft rate of 80% and BMG rate of 94%. Similarly, Faiena et al. also reported the highest success rate with vaginal flap and BMG [56]. In our review, the corresponding figures are 41%, 92%, 87% and 90%, respectively. Therefore, all urethroplasty techniques had a higher mean success rate (80–94%) compared to UD (< 50%) (Tables 6 and 7).

All the studies included in this review are heterogeneous, with a small number of patients. All are short case series (level 4 evidence) of descriptive nature and varied criteria for inclusion, success and failure. There was a very low incidence of de novo stress incontinence in several of the studies [11, 10, 15, 26, 34, 41, 42]. It seems that some complications were underreported, or most of the strictures were located distally, which allowed treatment without sphincter injury. Follow-up periods were shorter in some of the studies. Also, any meaningful comparison of the results to stricture location and length is not possible because these were not routinely reported in most of the studies. Since a randomized controlled trial is not feasible for such an uncommon entity comparing UD and different urethroplasty techniques, the flaws mentioned above are expected to remain in such a review made from level 4 studies. Finally, it may be suggested that an international registry collect data prospectively for more robust and long-term evidence.

Conclusion

UD has a mean success rate of only 41% at a mean follow-up of 35 months. Also, the success rate significantly decreases with repeated dilatations compared to the first dilatation. The mean time to failure is 12–15 months. Thus, it has the lowest success rate with minimum complications. In contrast, augmentation urethroplasty has the highest success rate of 90% at a mean follow-up of 27 months. The success rate of local flap augmentation is similar to that of free grafts (92.54% vs. 89.16%). Among free grafts, the BMG success rate is just marginally superior to that of local grafts (91% vs. 87%) at a follow-up of 22 and 20 months, respectively. Despite insufficient data and high-level evidence, it seems that local vaginal graft or BMG is a valuable technique concerning success rate and technical simplicity. It is our opinion that initially UD should be tried with a period of observation with CISC. Patients requiring more than two dilatations or early recurrence should be counselled about urethroplasty in a centre with the necessary expertise.