Introduction

Vesicovaginal fistula (VVF), although not life-threatening, is a distressing condition having a major psychosocial impact on both the patient and her partner. The reported incidence and aetiology differ between developed and developing nations. The incidence in the developed world is estimated between 0.3% and 2.0% with gynaecological surgery being the most common cause [1, 2]. In the developing world, the World Health Organization has estimated that about 80,000 to 1 million women have VVF with prolonged or obstructed labour being the leading cause [3]. Various approaches to surgical repair have been described in the literature. Both transabdominal (TA) and transvaginal (TV) approaches are common surgical procedures. While the anatomical closure rate following both the procedures seem comparable [4], there is a dearth of literature on the long-term urinary and sexual outcomes following successful repair. It is often presumed that TV approaches may lead to vaginal shortening and impair sexual function while TA approaches may cause more neuro-vesical dysfunction owing to more bladder mobilization. However, understanding of the quantum of such effects over the long term and objective evidence in terms of patient-reported outcomes are lacking. Most studies are limited to a short-term follow-up of about 6 or 12 months [5, 6] only when sexual activity is generally medically prohibited or not voluntarily initiated for a major portion of that duration. Also, many studies have used non-validated tools for such assessment. De novo urge urinary incontinence (UUI) or stress urinary incontinence (SUI) may also be related to the type of surgical approach used for repair. An abdominal approach may be associated with a higher incidence of bladder spasms in the early post-operative period [7]. However, whether there is any residual effect over the long term is largely unknown. Incontinence during sexual activity for any reason by itself may impair sexual activity [8]. Against this background, we tried to analyse the long-term continence of the patient and sexual functions of the couple with validated self-reported questionnaires after successful VVF repair and also compare the results between the TV and TA approach cohorts.

Methods

Study design

This cross-sectional observational analysis was done in a single tertiary care unit after institute ethical board approval. Medical records of patients who underwent VVF repair between April 2011 and December 2019 were reviewed. All women who underwent successful VVF repair by either a TV or TA approach and had a minimum follow-up of 6 months were included in the final analysis. The success of the procedure was defined by clinical history and examination. Women with isolated ureterovaginal fistula, vesicouterine fistula, urethrovaginal fistula, radiation-induced fistula, concomitant other pelvic surgical procedure, and repairs by minimally invasive techniques or colpocleisis were excluded. Women who had known neurological diseases (or developed them before the last follow-up) that could impact bladder or sexual function were also excluded. Pre-, intra- and post-operative details were retrieved from electronic data software. The interim follow-up details were recorded from the outpatient cards of each patient. Prospective cross-sectional data were collected on this retrospective cohort. Individual couples were contacted by telephone and called to the follow-up clinic for the assessment of functional outcomes. The questionnaire forms were filled out by the couples themselves. For those who were unwilling to travel or could not fill out the forms on their own, a co-investigator (NK) blinded to the type of surgical approach facilitated the filling out of forms in person or through a structured telephone interview as necessary.

Surgical procedures

The decision on the surgical approach depended on the location, size, number, prior repair of the fistula, and vaginal calibre, but was largely influenced by the operating surgeon’s preference. Briefly, in TV repairs the vaginal flaps were elevated, after circumscribing the fistula. When two or more fistulae were present, they were coalesced to a single fistula before repair. The final size of the fistula was then taken into consideration. The fistula was closed with absorbable sutures and the repair was completed by a three-layered watertight closure of the bladder and vagina. Excision or incorporation of the fistula margin in the sutures and tissue interposition was at the discretion of the surgeon, though interposition was used in most redo repairs. In TA repairs, a midline infra-umbilical incision was made and fistula repair completed by a trans-vesical O’Conor technique as described previously [9]. An interposition tissue was placed between the bladder and the vaginal closure. Suprapubic catheterization was used very sparingly in either approach at the behest of the operating surgeon.

Post-operative and follow-up care

Anti-cholinergics were given to all patients in the post-operative period and continued for 3 weeks when the Foley catheter was removed. All the patients were advised to refrain from sexual intercourse and heavy weight lifting for 3 months.

The patients were followed at 1, 3 and 6 months and then on an as-needed basis. The evaluation included clinical history and pelvic examination. Women with complaints of post-operative urinary leak were examined with a speculum, cough-stress test and/or office cystoscopy and vaginoscopy to assess for failed VVF repair, de novo urge or stress urinary incontinence.

Study outcomes

Patient-reported outcome measures (PROM) with respect to urinary and sexual function, as assessed by direct questioning, the International Consultation of Incontinence Questionnaire–Short Form (ICIQ-SF) and the Female Sexual Function Index (FSFI) [10, 11], were used as primary outcome measures. Both English and Hindi versions of the questionnaires were available. The ICIQ-SF consists of questions evaluating the frequency, amount and overall impact of urinary incontinence to calculate the ICIQ-SF score with range 0–21 (no urinary incontinence to severe urinary incontinence). The FSFI questionnaire contains 19 self-administered questions describing the six domains of the female sexual function namely desire, arousal, lubrication, orgasm, satisfaction and pain. In addition, we also assessed the husband's satisfaction subjectively on a verbal rating scale of 0 to 5 (0, no sexual activity; 1, very dissatisfied; 2, moderately dissatisfied, 3, equally dissatisfied or satisfied; 4, moderately satisfied: 5, very satisfied). The overall incidence of any urinary incontinence (urge or stress) and sexual dysfunction at the last follow-up was recorded. The respective outcomes in the TV and TA patient cohorts were compared. SUI and UUI were defined as per ICS definitions [12, 13] and recorded at least 3 months beyond the fistula repair whereas avoidance of intercourse or a FSFI score of ≤ 26.5 was defined as sexual dysfunction [14]. Secondary outcomes measured were impact of age at presentation, parity, redo repair, obstetric cause, fistula site, fistula size and type of repair on urinary or sexual outcomes.

Statistical analysis

The data on continuous variables were expressed as mean ± standard deviations whereas categorical variables were expressed as proportions. The Student t-test and Mann-Whitney U test were used for comparing continuous variables for normal and skewed distribution, respectively. Chi-square test and Fisher exact test were used for comparing categorical variables. A p value < 0.05 was considered statistically significant in the study. Univariate and multivariate logistic regression analyses were done to find various factors impacting urinary and sexual outcomes. All analyses were performed using the STATA 14.0 software (StataCorp, College Station, TX, USA).

Results

Patient demographics

A total of 81 patients had successful VVF repair during the study period and were available for follow-up. Thirteen failed repairs, 7 laser weldings, 2 laparoscopic/robotic repairs and 11 patients lost to follow-up were not included in the analysis. The baseline characteristics are shown in Table 1. The mean age was 37.5 ± 8.6 years (range, 19 to 58 years) and the mean fistula diameter was 12.9 mm (range, 3 to 30 mm). The most common cause of VVF was hysterectomy in 69.1% (56 patients). An obstetric cause was seen in 20 (24.7%) women, with 17 (20.9%) and 3 (3.7%) following lower segment caesarean section and prolonged labour, respectively. Of note, 33 (40.7%) women had previous failed repairs. In ten patients (12.3%), three or more fistulae were noted. The fistulae were located at the supra-trigonal and trigonal regions in 67 (82.7%) and 14 (17.3%), respectively. One patient required concomitant ureteric reimplantation for associated ureteric injury. The mean follow-up duration was 29.8 ± 19.1 months (range, 6 to 80 months).

Table 1 Patient demographics
Full size table

The study group included 28 (34.6%) TA and 53 (65.4%) TV repairs. Women with TA repair had a significantly larger fistula size (14.5 ± 4.8 vs. 11.9 ± 4.2 mm, p = 0.01), longer operative time (144.5 ± 19.8 vs. 103.6 ± 18.7 min, p = 0.001) and greater estimated blood loss (146.9 ± 77.9 vs. 92.5 ± 25.3 ml, p = 0.001) than TV repair (Table 2).

Table 2 Peri-operative outcomes
Full size table

Sexual function outcomes

A total of 72 (88.9%) patients reported being sexually active before developing VVF, and 20 (24.7%) patients had intercourse, even with VVF. Most patients who were sexually inactive before developing VVF became sexually active after VVF repair and were not excluded from this cross-sectional analysis as such. These were mostly young unmarried women with VVF of post-traumatic or post-hysterectomy aetiology. Nine (11.1%) women were sexually inactive for various reasons even after successful repair. This included two women having urine leakage during coitus, three with fear of urine leakage, one with fear of VVF recurrence, two young women with traumatic VVF who were virgins and one who did not reveal the reason. None of the women in our cohort had separated or divorced from her spouse.

The FSFI questionnaire was not available for a further nine patients, as they did not complete the questionnaire. Finally, 63 women (22 in the TA group; 41 in the TV group) completed the questionnaire at the time of the last follow-up and were available for analysis. Partners of 63 women also completed the husband satisfaction scores. Overall, 17/63(27.0%) women had a FSFI score ≤ 26.5. Excluding 2 young women with traumatic VVF who were virgins, and 9 of those who did not fill out the FSFI questionnaire, overall 24 (7 + 17) out of 70(34.3%) had sexual inactivity/dysfunction post VVF repair.

Both the groups were comparable in all the domains (desire, arousal, lubrication, orgasm, satisfaction and pain) of the FSFI questionnaire (Fig. 1a). Likewise, the overall mean FSFI score in the TA vs. TV group (28.7 ± 6.1 vs. 30.9 ± 5.2, p = 0.13) was comparable (Fig. 1b).

Fig. 1
figure 1

Sexual and urinary functional outcomes scores

Full size image

The mean husband satisfaction scores also did not differ between the two groups (4.2 ± 0.8 vs. 4.4 ± 0.7, p = 0.27; Fig. 1c). Partners of women in the TA group reported being ‘very’ and ‘moderately’ satisfied in 40.9% (9/22) and 40.9% (9/22), respectively, compared to 56.1% (23/41) and 31.7% (13/41) reported by partners of women with TV repair. Four (18.2%) partners of women in the TA group stated being ‘equally’ satisfied or dissatisfied compared to five (12.2%) partners of women in the TV group.

Urinary function outcomes

In the follow-up, 15 (18.5%) women had urinary dysfunction with 7 (8.6%) and 8 (9.9%) women developing de novo UUI and SUI, respectively. We found a significantly higher rate of UUI in TA repair compared to TV repair (17.9% vs. 3.8%, p = 0.04) (Table 2). The patients with UUI were treated with anticholinergics whereas women with SUI were advised to perform pelvic floor strengthening exercises. In the subsequent follow-up, all these patients were socially continent requiring no protection and none of them required any surgical intervention.

A completed ICIQ-SF questionnaire was available for 75 women (TA group, 26; TV group, 49) at the time of the last follow-up. Six patients did not consent to complete the questionnaire. The two groups did not differ in the mean ICIQ-SF score (0.7 ± 1.7 vs. 0.5 ± 1.4, p = 0.59; Fig. 1d).

Factors affecting sexual and urinary outcomes

In the univariate logistic regression analysis, we found age and parity impacted sexual dysfunction. However, in the multivariate analysis only parity [odds ratio 2.29, 95% confidence interval (CI): 1.20–4.37, p = 0.01] was found to be the most significant factor impacting sexual dysfunction in women with successful VVF repair (Table 3).

Table 3 Univariate and multivariate analyses of factors causing sexual dysfunction
Full size table

Likewise, type of repair, site and size of the fistula were found to impact urinary dysfunction in the univariate analysis. However, in the multivariate analysis only size (odds ratio 1.21, 95% CI: 1.04–1.41, p = 0.01) and site (odds ratio 4.98, 95% CI: 1.21–20.50, p = 0.02) of the fistula were found to significantly impact urinary dysfunction (Table 4).

Table 4 Univariate and multivariate analyses of factors causing urinary dysfunction
Full size table

Discussion

Women with VVF suffer major psycho-social issues with significant impact on normal well-being [15]. Though a reasonably high success rate of VVF repair is reported, studies have shown some women experience urinary and sexual problems even after successful repair, having a negative impact on quality of life [5]. In the present study, we had a success rate of 86.2% (81 out of 94) for VVF repair and the success rate of the TA and TV group was similar (82.4% vs. 88.3%, p value = 0.54) in this regard. Among 81 women with successful VVF repair, we found 34.3% and 18.5% women developed sexual and urinary dysfunction, respectively, over long-term follow-up. Multivariable analysis of our data revealed parity as the most significant factor impacting sexual dysfunction whereas fistula at trigone and fistula size significantly impacted urinary dysfunction.

Sexual health is an integral part of one’s normal well-being and it is not surprising that women with VVF desire to retain or regain their sexual function after surgery. While the reported literature mainly focuses on the success rate in terms of anatomical closure, most studies fail to concentrate on the sexual or urinary health of the women after VVF repair. The few studies that are available focus on short-term sexual function, limited to 6 to 12 months after VVF repair [5, 6]. Our study reports these functional outcomes over a much longer term after VVF repair, mean follow-up being 2½ years. One very recent study [16] in a much smaller set of 36 cases reported a significant improvement of sexual function as well as urinary distress with repair of the VVF at a median follow-up of 40.5 months. These results are consistent with the findings of our study, although this study is limited by only eight cases done through the abdominal route.

The issue of the TA versus TV approach to VVF with respect to sexual or urinary outcomes is often debated. Contrary to the results from the few available studies [5, 6, 16], many experts and textbooks continue to raise concerns about whether the transvaginal procedure is associated with more sexual dysfunction in the form of vaginal shortening, reduced lubrication or dyspareunia. We also found that both approaches did not differ in terms of FSFI scores and were in fact comparable across all the individual domains of sexual function. Although Lee et al. noted sexual dysfunction in two patients who underwent TA repair, they also did not find any statistical difference in the overall FSFI scores when patients were stratified by the approach to repair [17]. Overall it seems, although sexual dysfunction is relatively prevalent affecting about one-third of such women, it is unlikely related only to surgical repair itself and the type of repair per se does not seem to have much bearing on this effect. In our study, 24.7% women continued to be sexually active with good function even before fistula repair. Pope et al. reported as much as 64.3% of women in their cohort to be sexually active with VVF [6]. Such high variability in percentages again reflects that sexual function is highly dependent on individual couple choices and acceptability, though it may be beneficial to actively seek out those who tend to become fearful of it (as discussed earlier) and develop secondary sexual dysfunction.

Sexual outcomes are objectively better revealed by subjective assessment using validated self-reported questionnaires [5]. However, PROMs also evaluate the emotional or psychogenic parts of the dysfunction from the patient perspectives. Many may avoid sexual intercourse for the fear of VVF recurrence and have strained relationships due to prolonged sexual abstinence, dyspareunia etc. [6, 18]. In our study, 7 out of 81 patients totally avoided sexual activity even after VVF repair indicating the devastating impact of VVF on the psychology of some patients due to this disease. Urine leakage during coitus or fear of it was the prime fear for these women, highlighting the clinical significance of directly asking for these sequelae long after VVF has been surgically corrected. This is important, because the true subjective feeling about one’s own sexual activity is important for sexual health. Many of these problems can be medically or surgically corrected or perhaps even simple counselling will alleviate the apprehensions and may help improve sexual well-being for these patients.

Unique to our study, we have tried to assess various factors that may impact sexual or urinary outcomes after VVF repair. Notably, none of the anatomical fistula factors or approaches affected the sexual outcome. Only age at the time of surgery and parity were found to significantly impact it. Apart from experiencing emotional and psychological effects because of fistulae, many women lose their spouses and as many as 20% may experience changes in their marital status like divorce [6, 15]. However, no divorce or separation from the spouse was reported in our series. Partner expectations about the surgical repair are to be valued in the context of the sexual health of the couple. To the best of our knowledge, this is the first study to assess the husband (partner) satisfaction scores in the evaluation of sexual function of the couple after successful VVF repair. Enquiring specifically about the reasons for sexual dysfunction enabled the women to provide unprompted explanations. While the TA and TV groups did not differ in the husband satisfaction scores, interestingly, most partners reported levels of ‘moderately’ and ‘very’ satisfied. This highlights the very limited impact of VVF repair on male partner sexual satisfaction, if any.

Another unique feature of our study is that we also report on the incidence of de novo stress and urge urinary incontinence in our study over a long-term follow-up. Although bladder spasms are common with a Foley catheter in-situ in the immediate post-operative period, few women report urinary incontinence even after successful anatomical closure of the VVF in the form of urge or stress incontinence [7]. In fact, this negatively affects the quality of life of the women despite successful anatomical closure. In the present study, we found 18.5% women had urinary dysfunction following successful repair and a higher UUI rate was noticed in TA repair compared to TV repair (17.9% vs. 3.8%) (Table 2). However, this result is confounded by the presence of larger fistula size in the TA group. Nonetheless, it is conceivable that the essential steps of bivalving and long suture lines on the urinary bladder in a typical O’Conor technique may precipitate UUI [7]. However, all the women were relieved of the symptoms in the subsequent follow-up with conservative treatment. Likewise, either the inciting event or iatrogenic injury to the supports of the pelvic organs can result in stress incontinence. Fistulae at the bladder neck and proximal urethra are more prone to SUI after VVF repair [17]. The reported incidence of SUI varies between 3% to 30% [19]. We found that a trigonal site and large size of the fistula are independent risk factors for the development of urinary dysfunction over the long term.

There are some limitations in the present study. Although validated English and Hindi versions of the FSFI [20, 21] and ICIQ-SF (available at iciq.net, from Bristol Urological Institute) questionnaires were used, not all patients could fill them out on their own. A blinded co-investigator helped fill in the questionnaires for some patients, which places some limitation on the quality of data. Also, the study only included cross-sectional follow-up data over the medium to long term and did not indicate how sexual or urinary health may evolve over time after VVF repair. Moreover, we did not assess for lower urinary tract symptoms as such and looked only at urine incontinence and its impact. Yet, we believe that the urinary and sexual functions according to the PROM questionnaires provided a more subjective assessment of the patient’s perspectives over the long term. Also, we did not assess the functional scores before the fistula repair. Although this would have provided a baseline score to enable the comparison before and after the VVF repair, we believe that the data presented here with post-operative functional outcomes between the two approaches provide valuable information for counselling women undergoing VVF repair. Multiple surgeons were involved with some preferring the abdominal approach for most cases while others preferred the vaginal approach. In that sense actually the results may be more generalizable, mitigating the effect of surgical expertise to some extent.

To conclude, sexual and urinary dysfunction is found in a considerable number of patients after VVF repairs. However, our data suggest comparable long-term sexual and continence outcomes between TA and TV repairs.