- Chi Chiung Grace ChenChi Chiung Grace ChenDepartment of Gynecology and Obstetrics, Johns Hopkins University School of Medicine
- and René GénadryRené GénadryDepartment of Obstetrics and Gynecology, University of Iowa College of Medicine
Obstetric fistula (OF) is a condition that remains prevalent in non-industrialized nations, mainly in sub-Saharan Africa and Southeast Asia where proper and timely obstetrical care is inaccessible, unavailable, or inadequate. The reasons for the delay vary from country to country where poverty remains a common thread, and understanding the many factors leading to the development of OF is critical in preventing this scourge that has been all but eliminated in industrialized nations. Preventive measures can be effective when developed in conjunction with local resources and expertise and should include patient education and empowerment in addition to educating and equipping healthcare providers. In the absence of such measures, patients develop an « obstructed labor injury complex » involving the genital, urinary, and gastrointestinal tracts. Many troublesome health consequences arise from this complex, including skin lesions from the caustic effects of urine, endocrine abnormalities such as amenorrhea and infertility, neuropsychological consequences such as depression and suicide, and musculoskeletal impairments such as foot drop and contractures.
Globally, evidence-based interventions are needed to address the debilitating and persistent medical, psychological, and social effects of this condition on its sufferers. While surgery offers the amelioration of symptoms, many patients may not have access to such care due to lack of funds, knowledge of surgical options, or availability of surgical facility. Even after successful repair of the fistula, patients may still suffer from persistent incontinence, stigma, and socio-economic hardship requiring special programs for support, rehabilitation, and reintegration. Additionally, the patients who are deemed inoperable require special counseling and care. Consensus is needed on standardizing care and outcome measures to improve the quality of care and to evaluate programs directed toward prevention that will render this condition obsolete.
A fistula is an abnormal communication between two hollow organs. Due to the anatomic proximity of organs in the pelvis, these abnormal communications may develop between the reproductive tract (vagina, cervix, uterus) and the lower urinary tract (urethra, bladder, pelvic ureters), and/or the lower gastrointestinal tract (anus, rectum) (Figure 1). Over time, the fistula/communication tract becomes epithelialized and it fails to close. This persistent communication allows bodily fluids (e.g., urine, feces, blood, discharge) from one pelvic organ to unnaturally flow into the other, causing irritation, infection, or uncontrolled leakage. The causes can be traumatic, infectious, inflammatory, vascular, or neoplastic. The fistulas can also be iatrogenic, following surgeries, including cesarean section or hysterectomy, or following treatment such as radiation therapy.
Note. Specifically, fistulas may develop between the lower reproductive tract (vagina, cervix, uterus) and the lower urinary tract (urethra, bladder, pelvic ureters (ureters not shown here)), and/or the lower gastrointestinal tract (anus, rectum).
This review focuses on obstetrical fistula (OF), a fistula that results from trauma associated with the process of labor and delivery, usually following a prolonged and obstructed labor. While this preventable condition has been almost eliminated in higher-income countries, it remains prevalent where emergency obstetric care services are unavailable, inaccessible, inadequate, and/or untimely. Delays in recognition of prolonged or obstructed labor and lack of access to skilled care are the main factors facilitating its development. Unexpectedly, as access to healthcare improves globally, the incidence of OF following an assisted vaginal delivery or a cesarean section may initially rise. In these cases, it is unclear if a patient developed the fistula from the obstructed labor leading to the assisted delivery or if the fistula occurred due to an unrecognized complication or inadequately repaired injury at the time of assisted vaginal delivery or cesarean section, as many cesarean sections in lower resource settings are being performed by inadequately trained healthcare providers. Most of the published literature concerning OF consists of retrospective reviews and expert opinions with few randomized controlled trials or longitudinal studies.
Fistulas of the genital tract may involve the urinary tract and/or the intestinal tract. The former gives rise to a genitourinary fistula (GUF) and the latter to an enterovaginal fistula (EVF). They may also occur in combination (GUF/EVF).
The most common presentation of a GUF is a vesicovaginal fistula (VVF), which is a fistula that involves the bladder and the vagina. A urethrovaginal fistula (UVF) involves the urethra and may occur in combination with a VVF or as an extension of a VVF into the urethra. GUF is a more inclusive term and may include fistulas involving the upper genital and urinary tracts including the cervix, uterus, and/or ureters.
The most common type of EVF is a rectovaginal fistula (RVF) that may also occur in combination with a GUF, usually in cases of severe obstructed labor. When isolated, EVF often involves the lower rectum and the anal canal (anovaginal fistula (AVF)) and may be due to improperly repaired or non-healing vaginal/perineal tears or episiotomies performed at the time of vaginal delivery.
OF results most often from obstructed labor and interventions or maneuvers to relieve it. Obstructed labor can occur secondary to cephalopelvic disproportion or malpresentation, or following a protracted and dystocic labor course. During normal labor the anterior vaginal wall, upon which the bladder base and the urethra rest, is compressed between the fetal head and the posterior surface of the symphysis pubis. As the compression lasts for a short duration and is relieved between contractions, no permanent damage occurs. However, in obstructed labor that is not promptly relieved, the intervening tissues are compressed and, if unrelieved, undergo necrosis by ischemia. Most often the anterior vaginal wall is affected, although sometimes the anterior lip of the cervix and the trigone of the bladder are also involved depending on the level of the obstruction. The resulting necrotic area sloughs off, usually between day 3–10 following delivery, removing a critical barrier between the involved organs and leaving behind an abnormal communication wherever the pressure was exerted (bladder, urethra, rectum), leading to a fistula formation and incontinence (Creanga & Genadry, 2007). Dense scarring of the surrounding tissues including neurovascular structures also results (Wall, 2006). The combination of vaginal scarring, cervical destruction, and, frequently, amenorrhea may result in secondary infertility (Arrowsmith, Hamlin, & Wall, 1996). Most women with obstetric fistulas suffer injuries involving multiple organ systems: obstetric labor injury complex has been coined to reflect all possible morbidities associated with the “field injuries” resulting in OF that may involve the urologic, gynecologic, gastrointestinal, neurologic, and musculoskeletal systems (Arrowsmith et al., 1996). Of 377 cases of fistulas reported from Ibadan, Nigeria, 369 (97.9%) were obstetric and 343 (91.0%) were a direct consequence of obstructed labor (Lawson, 1989).
Smaller pelvic size, as occurs in women who have experienced malnutrition (Konje, 2000) and younger women with immature anatomy, can make operative delivery or cesarean section necessary; in resource-poor areas this is often not available and the delivery does not proceed in a timely manner. It has been speculated that women who have not reached skeletal maturity due to young age at delivery and/or poor nutrition may be at increased risk for developing OF. These women have immature pelvises that may result in cephalopelvic disproportion, leading to a fistula at the bladder neck or higher in the vagina (Murray, 2002).
Also, with outlet obstruction, as can be seen in patients having undergone extensive genital cutting, the vaginal introitus which has been deliberately narrowed to a pinhole may also be a cause of delay of delivery. When coupled with a lack of transportation, scarcity of health facilities, or sociocultural traditions impeding access to proper healthcare, OF may occur.
Besides complications of pregnancy termination and obstructed labor, procedures required during assisted vaginal delivery, such as forceps delivery or symphysiotomy, destructive delivery practices such as craniotomy, and traditional practices including female genital cutting (FGC), gishiri and gurya cutting have also contributed to the occurrence of a genital fistula (Ouedraogo, 2018).
It is estimated that the main cause of OF in the developing world—obstructed labor—affects at least 6.5 million women per year. (Lewis & de Bernis, 2006). These women live in areas where the risk of obstetric complications is highest and access to competent care is lowest (Abouzahr, 2003). Obstructed labor is estimated to account for 8% of all maternal deaths per year, while 5–10% of those who survive are at risk of the development of OF. As the actual prevalence and incidence of OF is only based on such estimations, the true number of women living with OF is unknown. It is known, however, that OF does exceed the developing world’s health systems’ capacity to provide adequate preventive and corrective care.
The magnitude of the problem might be underestimated due to the fact that patients with OF are often living in seclusion and their issues might not be reported by their families or communities. Such social stigma might make accurate estimates difficult and often unreliable; indeed, respondents with OF are often affected by others’ attitudes and perceptions of their afflictions.
Age and Parity
Many clinic and hospital-based studies have examined the roles of age, parity, and labor circumstances in the occurrence of obstetric fistulas, but the lack of standardized methods of data collection, evaluation, and reporting limits their value. For example, the length of labor is dependent on the patient’s recollection of the time of fistula occurrence or of the time of admission for repair. As the woman may have labored for days at home alone or with an unskilled birth attendant before she was delivered elsewhere, location of delivery is commonly unreliable and a self-reported duration of labor relies on the self-assessment of the start of active labor, which is usually not precisely known. The duration of labor varies widely as it relies on the patient’s imprecise recollection, but it has been reported as about 2.5 to 4 days (Ibrahim, Sadiq, & Daniel, 2000; Wall, Karshima, Kirschner, & Arrowsmith, 2004). Over 60% of deliveries leading to OF were reportedly carried out by no provider or by unskilled providers, although such information is rarely provided (Kelly & Kwast, 1993a). When it is collected, it is usually from referral centers that more or less consistently list provider supervision of delivery.
Although generally considered an affliction of the young primiparous patient, OF is not specific to reproductive age or parity. In various hospital-based case series, the mean age of women with obstetric fistula ranges from 19 to 65 years, and in approximately half of women the fistula developed during their first delivery. The percentage of primiparas with fistulas varied from 32% in a study of 2,484 cases by Hilton and Ward (1998) to 81% in the study of 31 cases by Ibrahim et al. (2000). In a large 1996 review of the literature on the epidemiology of obstetric fistulas, 11.7% of patients treated at Addis Ababa Fistula Hospital had six or more children (Arrowsmith et al., 1996). Indeed, higher age and higher parity have been reported for women with OF: a higher age (2.5% younger than age 16) and a higher rate of fistulas in women who experienced repeated deliveries than in primiparas were reported in Pakistan (Ahmad, Nishtar, Hafeez, & Khan, 2005).
Access to Healthcare Factors
The occurrence of an OF often reflects inadequacies at different phases of the birthing process and the healthcare system supporting it. When difficulties arise in a home delivery, the patient’s and family’s lack of understanding of the birthing process, along with limited education, financial support, and transportation, lead to an increased risk of stillbirth and OF development when coupled with the lack of readily available medical resources and skilled care. This situation is particularly compounded by the distance a woman must travel in order to reach a health facility and the often inadequate resources and skilled care encountered upon reaching the facility. Delay is a common feature in the development of OF, with many frameworks discussing the three types of delay first described by Thaddeus and Maine (1994): delay in deciding to seek treatment, delay in actually arriving at health facility, and then delay in receiving adequate care. Though there exists a high proportion of cesarean section delivery in patients reported with OF, this does not prevent the development of OF for many reasons, including lack of timely performance or lack of availability of skilled medical personnel, as well as the inherent risk of adding further injury following prolonged obstruction. Financial and social barriers also often interfere with access.
In a 2018 systematic review of the literature on barriers and facilitators to prevention, the most noted barrier addressed is in the first phase of delay: the decision to seek care, particularly due to a lack of awareness of the dangers of unsupervised labor (Lufumpa, Doos, & Lindenmeyer, 2018). The most noted facilitators addressed are the decision to seek care and the quality of care received at a facility. Unfortunately, most of labor abnormalities occur during labor and remain outside the realm of predictability (Abraham, 2014). Attempts have been made to identify those patients most likely to sustain labor obstruction by developing a “fistula index” which would include short stature, contracted pelvis, a history of delivering large babies, and/or suspicion of a large-for-gestational-age fetus in the current pregnancy. However, the use of a “fistula index” along with an intrapartum partograph8 have not been studied or effectively implemented (Browning, Lewis, & Whiteside, 2014). It seems that community mobilization holds promises in helping to reduce birth-related death and OF (Seim et al., 2014).
The risk factors that lead to obstetric fistula development are thus many, mostly revolving around poverty and lack of access to timely and proper obstetrical care, particularly when labor is obstructed (Figure 2). The fistula patient is most often described as poor, illiterate, young, living in a rural area, and attempting to deliver at home without a properly trained birth attendant. Thus, obstructed labor, obstructed access to timely care, and obstructed reception of proper care are the main components of OF (Figure 3).
Other Risk Factors
The large majority of patients with OF live in rural areas that are quite remote and poor and where access to care and education is challenging. Geographical region, place of residence, educational status, age at first birth, age at first marriage, employment status, place of delivery, and follow-up of antenatal care during pregnancy were significant determinant factors of obstetric fistula in Ethiopia (Andargie, 2017).
While women who delivered at a health facility and followed antenatal care for more than a day were less likely to experience OF than those who delivered at home and had no antenatal care visits, there was an inverse relationship between age at first birth and the prevalence of OF.
Furthermore, those who had had schooling (whether primary, secondary, or higher level) were less likely to suffer obstetric fistula than illiterate women. The likelihood of women who resided in rural areas to suffer obstetric fistulas was 5.167 times more than those in urban areas.
The logistic model also showed employment status to be a predictor of OF risk: women who were unemployed were 1.33 times more likely to have experienced obstetric fistula than employed women.
The ability to read decreased the odds of OF symptoms by 13%, whilehigher odds of OF symptoms were observed for women of short stature (<150 cm) (odds ratio (OR) = 1.31; 95% CrI: 1.02–1.68) in a study that pooled 27 surveys, including responses from 332,889 women on available demographic and health surveys (DHS) and multiple indicators cluster surveys in sub-Saharan Africa (Maheu-Giroux et al., 2016). Several case series have also documented short stature in women with fistulas, with average heights of less than 150 cm, and smaller foot sizes (Tebeu et al., 2012).
The exact relevance of the various forms of FGC in the development of OF remains unclear (Browning, Allsworth, & Wall, 2010). In general, they have not been shown to be associated with an increased risk of fistula development; nonetheless, the scarring of the most severe forms of FGC will increase the risk of outlet obstruction, with the need for some intervention that might increase such risk. Also yankan gishiri, using a sharp razor blade for longitudinal cutting of the vagina, has been implicated in some series. In Tahzib’s report on the epidemiological determinants of major etiological factors, one-third of the total had undergone gishiri, and in 15% this was thought to be the main etiological factor (Tahzib, 1983).
Incidence and Prevalence
Although the exact incidence and prevalence of OF is unknown, it is known that OF closely parallels maternal mortality, with a higher prevalence encountered in regions with high maternal mortality (Adler, Ronsmans, Calvert, & Filippi, 2013). As both conditions are directly linked to the accessibility of emergency obstetric care, such access to proper care is essential in the prevention of OF (Raassen, Ngongo, & Mahendeka, 2014).
It had been estimated that OF affects between 2 and 3.5 million women in the developing world, including Southeast Asia and sub-Saharan Africa where it is primarily caused by obstructed labor. From the literature, the range of incidence in Southeast Asia and Africa varied from 0 to 4.09 OF cases per 1,000 deliveries. The prevalence estimates ranged from 0 to 81 per 1,000 women, while the frequency of stillbirth associated with OF ranged from 32% to 100%, with 92% reported in the largest studies (Cowgill, Bishop, Norgaard, Rubens, & Gravett, 2015).
The prevalence estimates were largely based on self-reported symptoms of urinary incontinence from DHS and other population-based surveys or expert opinions extrapolating from facility-based settings. These estimates range widely from 0.16% to 4.7% in sub-Saharan Africa and 0.08% to 2.7% in South Asia. However, the reliability of self-reporting is questionable in the absence of confirmatory physical examination as urinary incontinence is frequently reported following delivery.
Utilizing meta-analytic technique, Adler et al. (2013) estimated a pooled prevalence rate of 0.29 per 1,000 women of reproductive age globally; 1.60 per 1,000 women aged 15–49 in sub-Saharan Africa; and 1.20 per 1,000 women aged 15–49 in South Asia. On the other hand, by analyzing information collected in DHS of 19 countries in sub-Saharan Africa and multiple indicators cluster surveys, Maheu-Giroux et al. (2015) estimate that 3 in every 1,000 women have had symptoms of obstetric fistula in their lifetime (lifetime prevalence), while 1 in every 1,000 women currently has the symptoms (point prevalence). Some of these studies used non-validated instruments without confirmatory pelvic examinations. The latter is indeed impractical and costly to implement on a population level as it requires the mobilization and training of a large cohort of healthcare workers.
A novel approach that has been suggested to more accurately estimate the prevalence of OF involves a community outreach followed by a physical exam by nurse-midwives at lower-level facilities closer to where at-risk women live (Tunçalp, Isah, Landry, & Stanton, 2014). Thus, using self-report of constant urine leakage with clinical assessment for confirmation, Mocumbi et al. (2017) report an incidence of 1.1 per 1,000 recently pregnant women in a population-based study of health facility in high density area in southern Mozambique.
While OF seems to be decreasing in sub-Saharan Africa and Southeast Asia, OF from iatrogenic causes (e.g., cesarean sections, hysterectomies, etc.) is more frequently encountered (Raassen et al., 2014). Indeed the term “retrogressive evolution” has been coined to describe the increasingly prevalent iatrogenic fistulas resulting from cesarean sections (Hilton, 2016).
The range of physical and psychological problems associated with obstetric fistula adversely affects the quality of women’s lives in numerous ways (Ahmed, Genadry, Stanton, & Lalonde, 2007). The constant leakage of urine and offensive smells make it impossible for them to fulfill their social and familial roles as active community members, wives, and mothers. As a consequence, they are frequently either self-isolated or abandoned by their husbands and progressively ostracized by their communities. Social stigma adds to their significant suffering. Describing the devastating role of the condition on women, Harrison writes, “In the case of the girls with an obstetric fistula, the baby is usually stillborn and … together with the fact that her odor is offensive … [soon] her incontinence becomes confused with venereal disease, and the affected family feels a deep sense of shame. The consequences are devastating: the girl is initially kept hidden; subsequently, she finds it difficult to maintain decent standards of personal hygiene because water for washing is generally scarce; divorce becomes inevitable and destitution follows, the girl being forced to beg for her livelihood” (Harrison, 1983, p. 385).
The quality of life and social positions of these women are often dictated by the reaction of people around them and their own perception of the cause of the fistula and the resulting social stigma. Their main goal is to keep clean and neat, preserve their marriage, maintain their social standing, and earn an income. If these goals cannot be easily achieved, more often than not they resort to self-imposed ostracism and withdrawal from regular activities and social connections. More symptoms of depression, post-traumatic stress disorder, somatic complaints, and maladaptive coping result. Compared to patients attending an outpatient gynecology clinic, they also report significantly lower social support: integrating mental health into the treatment by trained healthcare workers may improve their overall care (Wilson, Sikkema, Watt, & Masenga, 2015). Psychological interventions based on theories of cognitive behavioral therapy and coping models have been used effectively for women recovering from surgical correction of OF. Resuming mobility, increasing social interaction, improving self-esteem, reducing internalized stigma, resuming work, meeting their own needs and the needs of dependents, meeting other expected and desired roles, and negotiating larger life issues were themes central to women’s experiences following surgery (Watt et al., 2015).
Many of these women, especially after failed surgeries, need social reintegration and support for their subsistence and bare survival. They also need to adapt their sexual and reproductive behavior to prevent the recurrence of OF. Post-repair counseling about fistula and risk factors for recurrence, community-based follow-up care, and linkages to income-generating opportunities are critical. Rehabilitation and social reintegration efforts should ensure that women regain healthy, productive lives, as most remain interested in participation in marriage, community life, and childbearing (Donnelly, Oliveras, Tilahun, Belachew, & Asnake, 2015). A reintegration scoring instrument was validated in Uganda, as a first step toward improving the measurement of post-surgical reintegration (El Ayadi et al., 2017). Coping strategies, including group therapy (Ojengbede et al., 2014) and religious coping (Watt et al., 2014), have been quite useful in reducing depression, low self-esteem, and suicidal ideation, while improving social support and reducing stigma.
OF is a preventable and treatable disorder. Its clinical presentation depends on the structures involved and may include a combination of injuries referred to as the “obstructed labor injury complex,” including the genital, urinary, and gastrointestinal tracts, as well as skin lesions from the caustic effects of urine (Figure 4), endocrine abnormalities such as amenorrhea, neuropsychological consequences such as depression, and musculoskeletal impairments such as foot drop and contractures. We shall review the most common fistulas affecting the genital, urinary, and gastrointestinal tracts, namely GUF and RVF.
VVF is the most common type of GUF of obstetrical origin in lower resource settings and describes any fistula which involves the bladder/urethra and the vagina (Figure 5). Although most cases of GUF in lower resource settings may be a direct consequence of obstructed labor, as access to healthcare improves, more GUF cases are iatrogenic (e.g., associated with cesarean section often performed by inadequately trained health personnel) and may involve the cervix, upper vagina, uterus, or ureters. In a large retrospective analysis of almost 6,000 women undergoing GUF repairs, 13.2% of GUF were thought to be the direct result of provider error (Raassen et al., 2014). In higher resource settings, the most common etiology is pelvic surgery.
Patients typically present with continuous leakage of urine or watery vaginal discharge days or weeks following obstetric trauma or pelvic surgery. The degree of leakage may depend on the location and size of the fistula as well as the specific viscera involved. For example, women with vesicouterine fistula (VUF) (Figure 6) may not have continuous leakage and may only present with intermittent leakage or cyclic hematuria resulting from urine mixed with blood during menses (Youssef’s syndrome; Youssef, 1957). Urinary consequences may also include infection, stones, and, in rare cases, compromise of renal function. As the obstructed labor may result in wide ischemic damage to all pelvic structure and not just the bladder, often these patients may suffer from “obstructed labor complex” and may have other pelvic symptoms such as irregular vaginal bleeding, pelvic pain, and dyspareunia. Chronic exposure to urine, even of a short duration, may result in irritation of the vulvar/perineal skin.
Diagnosis and Evaluation
The diagnosis and evaluation of lower urinary tract fistula is usually straightforward and begins with a thorough pelvic examination. A speculum exam can be used to look for a pooling of fluid. As there may be other etiology for increased vaginal fluid, including vaginitis and fallopian tube cancer, if the origin of the fluid is not clear, it can be collected, and creatinine level can be determined. A split speculum or a Sims’ speculum may be used to identify the fistula tract. Even if a tract is identified, any area of tissue puckering should be further explored with a thin probe (e.g., cervical os finger, lacrimal duct dilator, etc.) as more than one fistula tract may exist. If a fistula tract is not visualized in a patient with a suspicious history, a dye test can be performed with or without a tampon/gauze in the vagina (Figures 7a and 7b). This test involves instillation of colored water or saline (e.g., using blue food dye, methylene blue, etc.) into the bladder with either direct visualization of blue fluid pooling in the vagina or visualization of blue-stained tampon or gauze upon removal from the vagina (placed in the vagina prior to bladder instillation). Although the false positive or negative rates of this test are unclear, a transurethral Foley catheter can be placed to minimize leakage of stained fluid from the external urethral meatus, which might otherwise result in a false positive. If the gauze is wet but not stained, this may indicate a ureteral fistula. Although not definitive, if there is ureteral involvement, administration of oral phenazopyridine or intravenous methylene blue or indigo carmine may lead to staining of the vaginal gauze orange or blue, respectively (Figure 8).
Note. A Foley catheter is placed, and the bladder retrograde filled with dyed fluid. Gauze or tampon is placed in the vagina.
Note. If the tampon is stained blue after the bladder is retrograde filled with blue-dyed fluid, this is suggestive of a bladder fistula. If the tampon is stained orange after the patient ingests a medication such as phenazopyridine, this is suggestive of a ureteral fistula. If the tampon is stained both orange and blue, the patient may have both a bladder and a ureteral fistula.
Attention should be paid to the number and size of fistula (size of residual intact bladder); specific viscera involved (e.g., uterus and bladder, bladder and vagina, ureter and vagina, ureter and uterus, etc.); location relative to anatomic landmarks such as the urethra, ureter, and/or cervix; and condition of surrounding tissue. As will be discussed in more detail in the section on corrective management, an important detail to consider is the accessibility of the fistula for vaginal repair.
Although cystoscopy is not always required to identify a GUF, it may be helpful for identifying the proximity of the fistula to ureters and the presence of other abnormalities or foreign bodies within the bladder, such as stones or sutures. If bladder distension is not possible to facilitate cystoscopy, a Foley catheter can be placed within the fistula tract. If the GUF is suspected to involve the uterus, hysteroscopy may be helpful to confirm the diagnosis. If a ureteral fistula is suspected, imaging, including intravenous pyelogram/retrograde pyelogram, voiding cystogram, or computed tomography (CT) urography (based on resource availability), can be used to evaluate the kidneys and ureters.
Several classifications exist for GUF, but none are standardized or universally used, thereby limiting comparative outcome analysis across studies. Furthermore, the staging in the existing classification systems have not been shown to be correlated with either repair complexity, surgical outcomes, patient symptom severity, or impact on quality of life. The World Health Organization (WHO) classification does provide guidelines on anticipated difficulty at the time of repair (Table 1a; Lewis & de Bernis, 2006).
Table 1a. Criteria based on the Degree of Anticipated Difficulty of Repair
Number of fistula
(RVF) mixed VVF/RVF: cervix
Involvement of the urethra/continence mechanism
Scarring of vaginal tissue
Degree of tissue loss
Ureters not draining into the vagina
Ureters draining in vagina, stones
Number of attempts at repair
No previous attempt
Failed previous attempts at repair
Note. Lewis and de Bernis (2006).
In general, the various classification systems take into account the fistula size, location relative to anatomic landmarks (urethra, ureters, cervix), involved viscera, and degree of scarring and other epithelial changes (e.g., radiation, inflammation). Although we would recommend describing a fistula with these details instead of choosing one classification system over another, the two more commonly used classifications systems were developed by Waaldijk (Table 1b) and Goh (Table 1c) (Goh, 2004; Waaldijk, 1995).
Table 1b. VVF
Type I: Does not involve the urethral closing mechanism
Type IIAa: Involves the urethral closing mechanism, without (sub)total urethral involvement and without circumferential defect
Type IIAb: Involves the urethral closing mechanism, without (sub)total urethral involvement and with circumferential defect
Type IIBa: Involves the urethral closing mechanism, with (sub)total urethral involvement and without circumferential defect
Type IIBb: Involves the urethral closing mechanism, with (sub)total urethral involvement and with circumferential defect
Type III: Involves the ureter; other exceptional fistulas
Note. Waaldijk (1995).
Table 1c. VVF
Type 1: Distal edge >3.5 cm from external urinary meatus
Type 2: Distal edge 2.5–3.5 cm from external urinary meatus
Type 3: Distal edge 1.5 to b2.5 cm from external urinary meatus
Type 4: Distal edge < 1.5 cm from external urinary meatus
(a) Size < 1.5 cm in the largest diameter
(b) Size 1.5–3 cm in the largest diameter
(c) Size > 3 cm in the largest diameter
i. None or only mild fibrosis (around the fistula and/or vagina), and/or vaginal length > 6 cm with normal capacity
ii. Moderate or severe fibrosis (around the fistula and/or vagina), and/or reduced vaginal length and/or capacity
iii. Special consideration, e.g., post-radiation, ureteric involvement, circumferential fistula, or previous repair
Note. Goh (2004).
Most of the literature on surgical outcomes consists of large retrospective case series with short follow-up periods (weeks to months). Rates of success vary according to population and technique but are generally high (75–95%), with highest repair success seen after the initial attempt (Hilton, 1998). Patient-specific characteristics that may contribute to repair failure include poor general health, significant vaginal scarring, small residual bladder size, loss of urethra, and circumferential fistula (the fistula involves the entire bladder so that the entire proximal bladder is detached from the distal bladder/urethra) (Nardos, Browning, & Chen, 2009; see Figure 9). Other factors also reported to affect repair outcomes include the surgeon’s experience, number of attempts, and availability of health facilities (Barone et al., 2012).
Although most GUF will require surgical repair, it is reasonable to proceed with a trial of conservative management, especially if the patient presents within days or a few weeks of the onset of symptoms, the fistula tract is not well epithelialized, and the fistula is small in size (<5–10mm), not related to malignancy, and located supratrigonally. Conservative management consists of prolonged bladder drainage with a transurethral Foley catheter. Typically, the transurethral catheter is left for four to six weeks. If, at the time of presentation, there is significant tissue edema surrounding the fistula preventing immediate repair, a catheter trial may also be reasonable. In a retrospective review of 1,716 women with obstetric fistula, catheter placement resulted in fistula closure in 15% of patients (Waaldijk, 2004). Similarly, in patients with ureteral fistula, ureteral stent placement for six to eight weeks should be attempted when there is ureteral patency. In these specific patients, after stent removal, imaging (e.g., intravenous pyelogram/retrograde pyelogram, voiding cystogram, or CT urography) should be done to confirm fistula resolution and to evaluate for ureteral stenosis. Other minimally invasive options that have been reported in the literature, including methods of de-epithelializing the fistula tract (e.g., curetting, electrofulguration, etc.) and adding substances (e.g., fibrin glue, cyanoacrylic glue, etc.) to promote healing, have not been well studied and are mostly used in the context of small fistula from iatrogenic causes (Rogers & Jeppson, 2016).
In addition to maximizing preoperative considerations including adequate hydration (fistula patients may limit fluid intake to minimize leakage) and nutrition, operative risks should be thoroughly reviewed with the patient, including the possibility of repair failure and other lower urinary tract symptoms such as persistent incontinence, overactive bladder, and voiding dysfunction even with repair success.
Perioperative factors to consider while planning for surgery include timing of repair, route of repair, specific surgical techniques, and duration of postoperative transurethral catheter placement. While most expert fistula surgeons will agree that these are important considerations, there is no universal consensus or definitive evidence on the optimal clinical practice. This deficiency, coupled with evidence that the highest success rate is usually after the initial repair (81.2% success after the first attempt versus 65% after two or more attempts), suggests that fistula repair should be undertaken by specialists or surgeons specifically trained in the care of fistula patients (Hilton, 1998).
In planning surgery, the first consideration is when to perform surgery. While there is a lack of consensus, most experts would consider proceeding with surgery if the fistula occurred within 72 hours of presentation, there is no significant tissue inflammation, and the patient has no contraindication to proceeding with surgery. If there is still evidence of tissue remodeling and necrosis from the initial obstructed labor and ischemic injury, serial exams should be performed and surgery should not commence until tissue inflammation and necrosis has resolved.
A traditional principle of fistula repair is excision of the fistula tract to create fresh edges for tissue reapproximation. However, in a randomized controlled trial where 64 women with obstetric GUFs were randomized to tract excision or no excision at the time of vaginal repair augmented with Martius graft, the authors found similar success rates (68% in the trimmed group vs. 75% in the not trimmed group; see Shaker, Saafan, Yassin, Idrissa, & Mourad, 2011). As with any surgical consideration, decisions regarding tract excision should be individualized. Any area of significant fibrosis should be trimmed.
Besides making sure the tissue surrounding the fistula tract are well vascularized, other surgical principles include locating and protecting the ureteral orifice, mobilizing enough of the surrounding tissue around the fistula to allow for tension-free closure, confirming the integrity of the repair after closure of the first layer, performing a second layer closure if possible, reapproximating the vaginal epithelium, and draining the bladder continuously after repair (Figures 10–14). Furthermore, some experts have also advocated for placement of the sutures extramucosally. However, there is no evidence to support this, and in actual practice this may not be possible. Absorbable sutures are usually utilized to minimize any potential complications associated with the use of permanent sutures, including stone formation. There is no definitive literature supporting the choice of monofilament or braided sutures or the choice of suture placement in an interrupted or running manner. Most surgeons will use a fine caliber suture such as 3-0 of 4-0 polyglactin 910 for the first layer closure closest to the mucosa.
Note. Weighted speculum is seen in the posterior vagina. The photograph shows a surgical instrument placed through the external urethral meatus into the bladder. The bladder defect is shown by the exposure of the instrument. Essential principles of fistula repair include locating and protecting the ureters, mobilizing surrounding tissue to allow for tension-free closure of the fistula, two-layer fistula repair if possible (only one layer shown), retrograde filling of the bladder to check for watertight closure after the first layer closure, closure of the vaginal epithelium, and placement of a Foley catheter to allow for continuous drainage of the bladder to facilitate healing.
One of the most important surgical deliberations is proceeding with surgical repair vaginally or abdominally. Although there are only retrospective series addressing this topic and no evidence to support higher repair success rates with one route over another, the vaginal approach has been shown to offer the expected benefits of minimally invasive surgery, including faster recovery, shorter hospital stays, and decreased cost (Blaivas, Heritz, & Romanzi, 1995; Gedik, Deliktas, Celik, Kayan, & Bircan, 2015; Warner et al., 2019; Zambon et al., 2010). The vaginal approach should be considered if the fistula is accessible vaginally. This may be especially pertinent in lower resource settings as patients may better tolerate vaginal surgery under regional anesthesia. However, other considerations that must also be taken into account include history and route of previous repair, surgeon expertise, and if any concomitant abdominal surgeries, such as hysterectomy or ureter reimplantation, are needed.
After surgical repair, an important consideration is duration of catheter drainage of the bladder to allow for bladder decompression and healing. There have been two randomized controlled trials comparing 10 days of catheter drainage versus 14 days and 7 days versus 14 days (Barone et al., 2015; Nardos, Menber, & Browning, 2012). Both trials demonstrated similar repair success rates. In 2018 the WHO issued a guideline recommending 7–10 days of bladder drainage after simple fistula repair.
As patients with short-term use of bladder catheters may be at increased risk for urinary tract infections, there is limited evidence supporting the use of antibiotics during catheter drainage in any post-surgical patients (Lusardi, Lipp, & Shaw, 2013). However, in the obstetric fistula population, there was a large randomized controlled trial that involved over 700 women comparing one dose of antibiotics given at the time of surgery versus one dose of intraoperative antibiotics along with extended use of antibiotics administered after surgery (Muleta, Tafesse, & Aytenfisu, 2010). This study demonstrated similar outcomes in both groups, including repair success rates and rates of infections.
Interposing healthy tissue or a flap at the time of fistula repair may eliminate a potential dead space and increase vascularity to the repair. Most fistula surgeons will only use a flap to augment fistula repair in cases that are recurrent/persistent or if the adjacent tissue is significantly scarred with impaired vascularity. Commonly used flaps include labial flap (Martius), which can be used during vaginal repair, and omentum or sigmoid epiploica, both of which can be used during abdominal repair (Lee, Dillon, & Zimmern, 2013). Use of xenografts such as porcine dermis or submucosa to augment repairs have also been described (Mellano & Tarnay, 2014).
Urinary diversions are usually reserved for patients with inoperative fistula or those who have undergone multiple failed repairs as the bladder is frequently non-compliant due to size or extensive fibrosis (Arrowsmith, 2007; Walker, Ambauen-Berger, Saha, & Akhter, 2018). These urinary conduits can be incontinent or continent and are usually fashioned from small or large bowel; however, these types of surgeries are complex and thorough consideration must be given to the long-term follow-up needed for patients upon which such surgeries are practiced. In addition to perioperative complications, these patients are at risk of long-term complications such as infection, electrolyte abnormalities, and renal damage. Further consideration must also be given to matters such as ostomy supplies.
Any patients who present with urinary incontinence symptoms after successful fistula repair must first undergo an evaluation for recurrent fistula. After confirming the lack of fistula recurrence, these patients may undergo an office simple cystometry or full urodynamic testing to further evaluate bladder symptoms. In simple cystometry, the patient is first ask to void and then a postvoid residual is checked with a catheter. The patient’s bladder can then be filled to determine bladder characteristics such as bladder capacity and evidence of detrusor contractions. After the catheter is removed and with a full bladder, Valsalva maneuvers can be performed with the patient in different positions (supine, squatting, standing) to determine if there is any leakage of urine. The patient is then asked to void again to determine the amount voided. Both the simple cystometry and the urodynamic testing will determine how well the bladder holds and releases urine, presence of involuntary bladder contractions, and evidence of incontinence. The difference is that simple cystometry can be performed without specialized equipment while urodynamics may require specialized pressure monitors that allow for the determination of intraabdominal and intravesical pressures, which may be useful in determining an optimal treatment plan.
Persistent incontinence symptoms have been reported in 10–55% of women who underwent successful obstetric fistula closure, with reported risk factors including urethral involvement (most experts believe 2–3 cm of urethra is required for continence), fistulas that have required multiple repairs, significant vaginal scarring, and small contracted bladder (Browning, 2004; Hilton, 1998; Murray, 2002; Nielsen et al., 2009). Urodynamic findings include stress incontinence (31–56%), detrusor instability (37–41%), and voiding dysfunction (4–13%). Treatment of persistent bladder symptoms, including stress urinary incontinence, in this population remains challenging as many effective treatments, such as the use of a midurethral or pubovaginal sling to treat stress urinary incontinence, may not be effective due to the extent of scarring and wide damage that resulted from the obstructed labor. Although it is beyond the scope of this article to further discuss optimal management strategies for persistent incontinence in OF patients, suffice it to say that some of these accepted treatments, such as those that involve the placement of polypropylene mesh sling in midurethra, may be relatively contraindicated in these patients due to an increased risk of mesh erosion (20%) (Ascher-Walsh et al., 2010).
Ureterovaginal, Ureterouterine, and Vesicouterine Fistulas
Ureterovaginal, ureterouterine, and vesicouterine fistulas are usually iatrogenic (Wanjala, Mwangi, & Mabeya, 2018). They may be associated with obstetrical interventions to relieve obstructed labor, most commonly after a cesarean section (Tazi et al., 2000), or more rarely after an operative maneuver or a destructive procedure to deliver a stillbirth.
In cases of VUFs, where the communication is between the bladder and the uterus, cyclic hematuria may be present with or without continuous vaginal leakage (Tazi et al., 2000). The diagnosis is clinical but may require imaging, including hysterography or retrograde cystography. Cystoscopy may be useful but evaluation of the upper tract is mandatory to evaluate the status of the ureters. Ureteral involvement has been estimated to occur in 10% of iatrogenic GUF. If a ureter is involved, it is commonly in the distal third of the ureter and is usually amenable to repair and/or reimplantation.
Reimplantation is the procedure of choice in cases of ureterouterine fistula where the communication is between the ureter and the uterus with or without involvement of the bladder. The diagnosis relies on clinical features following a cesarean section including incontinence of urine with persistence of normal micturition (Pernin, Fanton, & Dufour, 1993); the leakage of urine is usually visualized from the cervix.At cystoscopy, the bladder appears normal lest an ureterovesicouterine fistula is also present. Indeed, ureterovesicocervical fistula has also been reported following cesarean section (Tsivian, Tsivian, Sidi, & Tsivian, 2012).
In ureterovaginal fistulas, where the communication is between the ureter and the vagina, the dominant clinical feature is likewise urinary vaginal incontinence with persistent micturition with or without the presence of back pain. Imaging is necessary for confirmation and evaluation of the upper tract. Bilateral ureteric involvement has been reported and a concomitant VVF may often be present (Benchekroun et al., 1998). The most important aim of any treatments for ureteral fistula is the preservation of kidney function and ureteral patency. If a patient with a ureteral fistula present early from the inciting injury, endoscopic management with ureteric stent placement is usually possible and may lead to resolution of the fistula (Upadhyay et al., 2018). Although primary percutaneous nephrostomy has been attempted to preserve the function of the kidney and allow for the healing of the fistula, it has not been as effective in resolving the fistula (Schmeller, Göttinger, Schüller, & Marx, 1983). When surgery is required, a ureteroneocystostomy (reimplanting the non-injured portion of the ureter into the bladder) remains the procedure of choice, either open or laparoscopic (Falandry, 1992; Kumar et al., 2011). Transvaginal repair with a Latzko procedure following ureteric stenting has also been attempted successfully in very selective circumstances (Chen, Yang, Yang, & Huang, 2007).
RVFs most commonly result from obstetric trauma in both higher and lower resource settings. Whereas in higher resource settings, these fistulas typically occur either after unrecognized severe perineal injury or from infection and wound breakdown of laceration repair, approximately 5% of fourth-degree perineal lacerations repaired at the time of delivery will result in infection and dehiscence, with a smaller percentage progressing to RVF. In a historical report of 24,000 patients, 1.7% suffered a fourth-degree tear and 0.5% sustained an RVF (Goldaber, Wendel, McIntire, & Wendel Jr., 1993). In a literature review, 0.1% of patients who had undergone an episiotomy suffered an RVF. Specifically, of the patients who received a median episiotomy, 0.05% developed a RVF and 1% of patients with third- to fourth-degree laceration developed a RVF (Homsi, Daikoku, Littlejohn, & Wheeless Jr., 1994). Risk factors for complex perineal lacerations include primiparity, midline episiotomy, larger fetal size, older maternal age, and use of operative vaginal delivery. In the United States, the overall incidence of RVFs following vaginal delivery is in the range of 0.1%. The age-adjusted rate of RVF repair has declined since the late 1970s (3 per 100,000 in 1979 to 2 per 100,000 women in 2006), paralleling the declining rate of episiotomies and operative vaginal delivery (Brown, Wang, Bunker, & Lowder, 2012). A large population study found that the rate of severe perineal lacerations in the United States is declining, from 6.35% in 1992 to 5.43% in 1997 (Handa, Danielsen, & Gilbert, 2001).
In low resource settings, RVFs are the result of obstructed labor and unrepaired severe perineal lacerations. The risk factors for RVF of obstetric origin are increased by instrumental and operative delivery, including midline episiotomy, primiparity, high birth weight, prolonged second stage, and older maternal age (Landy et al., 2011; Smith, Price, Simonite, & Burns, 2013). These usually involve the lower third of the rectovaginal septum, a number of them being rather anovaginal. Nonetheless, the rectovaginal septum may be injured at any level of the vaginal canal. With a high obstruction, an RVF is more frequently associated with a GUF, while with an outlet obstruction such an association is less common and if present often involves the urethra. Concomitant GUF with RVF has been found to occur in 5 to 15% of all women with OF and may be indicative of a more severe injury. Kelly and Kwast (1993b), reporting on data from the Addis Ababa Fistula Hospital, noted a 15.2% prevalence of combined fistulas, and a 6.8% prevalence of isolated RVFs. In a review of 447 patients with OF resulting from obstructed labor at Panzi Hospital in the Democratic Republic of the Congo (DRC) between 2005 and 2007, there were 17 combined RVF/GUF and six isolated RVF (Sjøveian, Vangen, Mukwege, & Onsrud, 2011). Browning made the observation that the presence of a foot drop is usually indicative of an associated RVF. A severe high obstruction may be responsible for the compression of the rectum against the sacrum and the pressure damage of the lumbo-sacral neural plexus. In a review of 1,057 patients suitable for analysis, the combination of GUF/RVF was present in 7.5% (79) of the patients and 4.3% (45) had isolated RVF, of which only four were due to obstructed labor. The combined cases had longer labor, more stillbirths, more vaginal scarring, and larger and lower GUF (Browning & Whiteside, 2015).
Patients with RVF usually complain of an abnormal vaginal discharge, vaginal irritation, dyspareunia, the passage of gas and feces through the vagina and fecal incontinence. The discharge is malodorous and irritating, and insensible fecal soiling of undergarments is commonly reported. The severity and amount of these losses is dependent on the level and size of the communication as well as the frequency and characteristics of the bowel movement, evidently more prominent with loose stools (Figure 15). Some small fistulas can be minimally symptomatic or even asymptomatic.
Note. The photograph shows a surgical instrument placed in the anus going through the RVF.
Diagnosis and Evaluation
The evaluation of a patient with RVF begins with inspection of the perineum and the introitus for fecal soiling and suspicious discharge and irritation. The presence of granulation tissue or a dimple helps direct attention to the possibility of infection or communication and a thinned-out perineum raises the probability of a defective anal sphincter (Figure 16).
Note. No intervening tissue/skin between the anus and the posterior vagina.
The speculum exam looks for the presence of an associated GUF anteriorly and that of a septal vaginal defect posteriorly at all levels of the vaginal canal, including the midline and lateral posterior fornix.
A gentle digital examination of the anorectum will easily recognize a sizable defect while a smaller communication requires the use of a variety of aids to visualize it.
Simply placing methylene blue-dyed gel at the tip of the digit while massaging the anterior anorectal wall will force the dye into the communication seen in the vaginal canal.
Alternatively, in the office, with the patient in mild Trendelenburg position and while the vagina is filled with saline, a 18–20 Fr Foley catheter can be introduced in the rectum and a 5cc balloon inflated to retain air introduced in the rectum via the Foley catheter. Bubbles of air may be seen on the vaginal end, suggesting the presence and level of the communicating defect. This could also be carried out at the time of proctoscopy if needed. Air could also be replaced with a blue-dyed saline solution for an enema that could stain a previously placed vaginal tampon, confirming and exposing the fistula.
Endoscopic evaluation, including vaginoscopy and proctoscopy, have been used effectively, with the latter having the advantage of assessing the status of the anorectal mucosa for any inflammation or other pathology as well as the absence of obstructive strictures.
A bidigital exam, with one finger in the rectum and one in the vagina to evaluate the presence of induration as well as the size of the perineal body, and a bimanual examination complete the evaluation of the pelvic structures, including the external anal sphincter (EAS) and the levator ani muscle function. Indeed, it is important to evaluate the continent status before repair as preoperative fecal incontinence has been found in as many as 48% of patients and its persistence remains the most common cause of dissatisfaction (Tsang et al., 1998).
Imaging may be needed where available, particularly in high RVF and to assess the EAS. For the latter, endoanal ultrasound has proven most helpful, as an uncorrected defect will result in persistence of fecal incontinence, although the outcomes of sphincter repairs have not always been predictably encouraging. Endoanal ultrasound also helps assess the perineal body and internal and external sphincters (IAS, EAS, respectively). Barium enema (BE), CT, and magnetic resonance imaging (MRI) each may offer insights, with some limitations depending on the specifics of the RVF for the planning of treatment. These techniques are useful for the evaluation of non-traumatic OF, although MRI is more useful in the evaluation of the anorectum and the identification of the RVF tract, with the added advantage over ultrasound of not being operator dependent (Stoker, Rociu, Ruud Schouten, & Laméris, 2002): an ultrasound may be used in simple fistula while an MRI may be recommended for a complex one. In that instance, endoscopy also adds important information regarding the integrity of the lumen and the health of the mucosa. It is also critical to rule out the presence of other combined fistulas to the urinary and genital tracts and the perineum.
A number of classifications exist. In general, depending on anatomic location and degree of complexity, they are divided into high or low and simple or complex. A midlevel fistula is located in between, usually above the sphincter complex. Fistulas are also differentiated depending on the size of the communication: small if <2.5 cm, or large if above 2.5 cm. No consensus has been reached on their prognostic significance.
The importance of anatomical structures that may be involved in a defective rectovaginal septum and that are of critical consideration for reconstruction were captured in an anatomical classification of rectovaginal septal defects (Rosenshein, Genadry, & Woodruff, 1980). Five types were identified based on anatomic differences and the functional importance of the perineal body, including the EAS, with implication for surgical management (Table 2a).
Table 2a. Rosenshein RVF
I. Loss of the perineal body not associated with an identifiable fistulous tract
II. Loss of the perineal body associated with a fistulous tract involving the lower third of the vagina
III. Fistulas involving the lower third of the vagina with an intact or attenuated perineal body
IV. Fistulas involving the middle third of the vagina
V. Fistulas involving the upper third of the vagina
Table 2b. Waaldijk Rectovaginal Fistula
I. Proximal not involving the continence mechanism
a. without rectal stricture—transverse closure
b. with rectal stricture (common)—disruption of stricture
c. with circumferential defect (not common)—end-to-end anastomosis or combined A/V with colostomy
II. Distal fistulas involving the continence mechanism: without sphincter ani involvement—longitudinal closure; with sphincter ani involvement—layered reconstruction
III. Miscellaneous fistula i.e., ileo-uterine after instrumental abortion—as needed
Table 2c. Goh Rectovaginal Fistulas
Type 1: Distal edge of fistula >3.5 cm from hymen
Type 2: Distal edge of fistula 2.5–3.5 cm from hymen
Type 3: Distal edge of fistula 1.5–<2.5 cm from hymen
Type 4: Distal edge of fistula <1.5 cm from hymen
(a) Size < 1.5 cm in the largest diameter
(b) Size 1.5–3 cm in the largest diameter
(c) Size > 3 cm in the largest diameter
i. No or mild fibrosis around the fistula and/or vagina
ii. Moderate or severe fibrosis
iii. Special consideration, e.g., post-radiation, inflammatory disease, malignancy, or previous repair
The evidence remains based mainly on case series reporting on operative techniques with short-term and poor quality follow-up (Göttgens, Smeets, Stassen, Beets, & Breukink, 2014). Prospective collaborative studies are indeed urgently needed. A high fistula is usually located in the upper part of the vagina close to the cervix, proximal to the sphincter mechanism. It is more likely to require an abdominal approach due to its inaccessibility. A low one is located in the anorectum, just above or below the dentate line, distal to or through the sphincter. Depending on its characteristics, it is amenable to anal, perineal, or vaginal approaches. The management should be individualized in order to optimize closure and continence.
The only place for conservative management is in the small recently developed fistula or the matured minimally symptomatic lesion where bulking agents and alteration of bowel consistency may reduce its symptomatology. In reviewing favorable factors in RVF management over a 20-year period, Lo, Huang, Dass, Karim, and Uy-Patrimonio (2016) reported 14% favorable outcomes with conservative management in patients with RVF mainly from obstetrical complications.
A report of occasional closure of an early fistula following operative injury was managed with parenteral nutrition for 18 days along with a fluid diet and oral intake of daily loperamide for the first week followed by a fiber-rich regimen (Anaf, De Groote, Simon, & El Nakadi, 2001). The authors recommended close clinical and biological monitoring. Nonetheless, the chance of spontaneous closure remains low (Debeche-Adams & Bohl, 2010) but more substantial when the RVF is small and primary (Oakley et al., 2015). Fibrin glue has also been attempted in an effort to avoid more extensive surgery for small perineal fistulas, where success was reported in half of a small series (Hjortrup, Moesgaard, & Kjærgård, 1991).
Principles of Repair
The principles of surgical repair of RVF are the same as for GUF, namely wide dissection to allow a two-layer closure without tension and with gentle tissue handling and meticulous hemostasis. The first layer everts the edges of the excised tract. The second layer reduces the tension on the first layer. The vaginal side may be left open to drain. Avoiding placing sutures in the rectal mucosa is best unless circumstances dictate otherwise; at times a complete avulsion of both ends of the rectum may have occurred, in which case the posterior wall also needs closure and a stool diversion procedure such as colostomy may be needed. When damaged, the anal sphincter should be repaired. Interposing a vascular graft may be considered whenever severe scarring has reduced local blood supply. In the event a combined GUF/RVF are present, the decision to proceed with concomitant repair should be individualized. It is practically easier to close the GUF before addressing the RVF. In recurrent cases, it is best to close the RVF and carry out an interval GUF repair.
Lesions involving the vaginal apex that are inaccessible vaginally are best corrected abdominally. This could be done with either an open abdominal incision or with a minimally invasive approach including laparoscopically or robotically although these are not generally available in low resource countries. The principle of repair includes the separation of the involved organs, rectum, and vagina, and separate closure of each, with resection of the involved segment of bowel and reanastomosis. Interposition of peritoneum or omentum may decrease the risk of recurrence.
Lesions involving the lower vagina are best approached vaginally or rectally. The continence status and the relation to the continence mechanism are critical in the choice of approach. The continence mechanism of the anorectum relies on a compliant rectum and a functioning IAS, EAS, and levator ani muscles, along with a healthy mucosa providing a proper sensory and seal effect along the anal canal. The evaluation of the continence mechanism with ultrasound and manometry will help delineate the risk of residual fecal incontinence. Thus the relationship of the RVF to the EAS determines the best approach: proximal lesions involving the sphincter are best managed with creating a fourth degree and repairing it while a distal lesion can be managed with a Warren flap (vaginocutaneous flap) or an endorectal advancement flap, anocutaneous flap, or perineoproctotomy (Mazier, Senagore, & Schiesel, 1995). For an RVF located in mid vagina, a layer closure is most effective in the absence of stricture and a transverse rectal closure is preferred to avoid stricture.
Preoperative preparation: there is no consensus on the primacy of preoperative mechanical bowel preparation or the value of rectal lavage intraoperatively, although reducing the fecal load and reducing bacterial count may theoretically decrease the risk of infection and dehiscence. We recommend a mechanical bowel preparation and prophylactic antibiotics administered at the time of surgery. Proper hydration and a protein-rich diet are important for proper healing and recovery.
Postoperative care: there is no consensus regarding the optimal diet following repair except that avoiding the passage of hard stool through a fresh wound may be beneficial. We recommend liquid and low residue diet with laxatives on day 3–4 and stool softeners and gentle laxatives day 4–5, with advancement of diet. Stool softeners should be continued for one month. We also recommend pelvic rest, including no intercourse, for six to eight weeks. There are no definitive studies on this topic and there is a lack of consensus on the duration and the role of prophylactic antibiotics.
Timing of repair: it is generally agreed that repair should be delayed to allow for tissue healing, including resolution of tissue necrosis and inflammation, historically two to six months after injury. More current retrospective reviews and expert opinions recommend earlier repair in the absence of active inflammation.
A colostomy is normally not required in RVF repair, particularly following obstetric trauma in low and midlevel simple fistulas. If, in selected cases (e.g., severe perineal infections, complex colorectal repairs, failed previous attempts, circumferential fistula, etc.), a colostomy is required, one must make sure that the repair is performed two to three weeks after the colostomy is created and that the colostomy is reversed four to five weeks after successful RVF repair. Colostomy on its own is not a definitive solution to address RVF in lower resource settings where colostomy appliances are usually not available. More research is needed on the role of colostomy in RVF repair in this context.
There are many ways described to surgically correct an AVF and a RVF. In selecting an operative technique, the characteristics of the RVF and the status of the EAS are the most important factors for success of closure and achieving continence. The presence of additional fistulas and any previous repairs should be taken in consideration also.
Sphincteroplasty: when a defective sphincter is present, a sphincteroplasty has been shown to improve the outcome of the correction of the RVF (Tsang et al., 1998). For this, a curvilinear incision is made over the perineum and the dissection is carried out between the rectum and the vagina to identify the retracted ends of the EAS, usually found at 3–4 and 8–9 o’clock. Delayed absorbable sutures are used to provide an overlapping repair when feasible or an end to end approximation otherwise. It is critical to also repair the IAS with continuous or interrupted absorbable sutures as it is commonly involved in the defect. An additional reapproximation of the fascia of the Levatores Ani (LA) is an alternative to a levatorplasty to prevent subsequent introital pain and dyspareunia. Success with this procedure has been reported between 52% in repeat repair and 82% in primary cases.
Advancement flaps: where the sphincter is intact and enough tissue mobility is present, an advancement flap is effective. Endorectal flaps and anodermal flaps can be effective in small low fistulas; however, they lose some effectiveness with over two previously failed repairs. The flap is developed and used to cover the fistula.
Vaginal repair is effective particularly for traumatic OF. A Warren flap is useful in patients with cloacal introitus and those fistulas that are converted into such when a distal fistula involves the EAS that needs reconstruction. In midvaginal fistula, a layer closure is very effective. A high vaginal fistula involving the cul-de-sac of Douglas may require an abdominal or combined approach when sphincteric reconstruction is needed.
Fibrin glue and fistula plugs may be considered for residual or small recurrent communications although their effectiveness has been reportedly mixed in those circumstances (Abel, Chiu, Russell, & Volpe, 1993; Loungnarath et al, 2004). Novel techniques continue to be investigated (Knuttinen et al, 2018).
Where there is severe scarring and poor vascularization, particularly in recurrent cases, an interposing graft may be useful. These have included a Martius graft (Elkins, DeLancey, & McGuire, 1990; Songne et al., 2007) and gracilis muscle vaginally (Wexner et al., 2008), and an omental (Sapmaz, Celik, & Semerciöz, 2003), epiploical, or peritoneal graft abdominally.
The success rates of endorectal, transvaginal, or transperineal closures of the fistulous tract have ranged between 50% and 70% (Ommer et al., 2012). To optimize local healing conditions, patients with RVF are likely to benefit from fecal diversion if significant destruction of the anal canal has occurred, if the RVF is large, or, in some patients, if the RVF is recurrent. In a case series of 48 patients of whom 13 developed RVF of obstetric etiology, the repair was done through the transvaginal approach and only four required a temporary protective ileostomy due to size, severity of the lesion, or failed previous repair (Reisenauer, 2016). At the Addis Ababa Hospital, the incidence of colostomy for high RVF was reportedly 25% and 15% for low RVF (Kelly & Kwast, 1993b). When a colostomy is carried out, the repair can be done no earlier than two to three weeks later to optimize healing and reduce inflammation and edema. It could be reversed within five to six weeks. Two patients that became pregnant following repair were delivered by elective cesarean section. The decision to perform a cesarean section should be individualized as shown in a series by Rahman, Al-Suleiman, El-Yahia, and Rahman (2003): eight of 52 patients became pregnant following successful repair. In two patients who had had a large RVF, delivery was by cesarean section, while the remaining six patients with low and smaller fistula were delivered vaginally over a generous episiotomy.
Besides persistent fistula, complications associated with surgical correction of RVF include local infection and wound dehiscence. Additionally, dyspareunia resulting from scarring or vaginal stenosis may arise in 25% of sexually active patients (El-Gazzaz et al., 2010; Tunuguntla & Gousse, 2006).
OF remains an important clinical and public health issue that affects primarily women in lower resource settings. In addition to physical symptoms, many women suffering from these conditions may experience psychological and social consequences, including depression, suicidal ideation, divorce, and ostracism from immediate family and the local community. Although most OF can be surgically repaired, other conditions, including urinary and fecal incontinence, may persist. Most of the literature contributing to our understanding of these disorders consists of retrospective case series and expert opinions. Although the ultimate solution to OF is the establishment of universal emergency obstetric services, further work, such as longitudinal studies, randomized controlled trials, expert consensus focused on standardizing a classification system, establishing best practices in surgical and clinical care, and developing and implementing effective prevention strategies, is essential to guide public health officials and policymakers toward solutions to eradicate this condition.
- Breen, M. (2019). Manual of obstetric fistula surgery. Global Library of Women’s Medicine.
- Chen, C. C. G., & Genadry, R. Obstetric fistula. In Oxford research encyclopedia of global public health. Oxford, UK: Oxford University Press.
- Hancock, B. (2013). An introduction to obstetric fistula surgery. Sapiens.
- Waaldijk, K. (2018). Obstetric trauma surgery: Art and science.
- Wall, L. L. (2018). Tears for my sisters: The tragedy of obstetrical fistula. Baltimore, MD: Johns Hopkins University Press.
- Abel, M. E., Chiu, Y. S., Russell, T. R., & Volpe, P. A. (1993). Autologous fibrin glue in the treatment of rectovaginal and complex fistulas. Diseases of the Colon & Rectum, 36(5), 447–449.
- AbouZahr, C. (2003). Global burden of maternal death and disability. British Medical Bulletin, 67, 1–11.
- Abraham, W., & Berhan, Y. (2014). Predictors of labor abnormalities in university hospital: Unmatched case control study. BMC Pregnancy & Childbirth, 14, 256.
- Adler, A. J., Ronsmans, C., Calvert, C., & Filippi, V. (2013). Estimating the prevalence of obstetric fistula: A systematic review and meta-analysis. BMC Pregnancy & Childbirth, 13, 246.
- Ahmad, S., Nishtar, A., Hafeez, G. A., & Khan, Z. (2005). Management of vesico-vaginal fistulas in women. International Journal of Gynecology & Obstetrics, 88, 71–75.
- Ahmed, S., Genadry, R., Stanton, C., & Lalonde, A. B. (2007). Dead women walking: Neglected millions with obstetric fistula. International Journal of Gynecology & Obstetrics, 99(Suppl. 1), S1–S3.
- Anaf, V., De Groote, A., Simon, P., & El Nakadi, I. (2001). Conservative treatment of rectovaginal fistula after laparoscopic treatment of rectovaginal adenomyotic nodule. Gynaecological Endoscopy, 9(2), 123–127.
- Andargie, A., & Debu, A. (2017). Determinants of obstetric fistula in Ethiopia. African Health Sciences, 17, 671–680.
- Arrowsmith, S. D. (2007). Urinary diversion in the vesico-vaginal fistula patient: General considerations regarding feasibility, safety, and follow-up. International Journal of Gynecology & Obstetrics, 99(Suppl. 1), S65–S68.
- Arrowsmith, S. D., Hamlin, E. C., & Wall, L. L. (1996). Obstructed labor injury complex: Obstetric fistula formation and the multifaceted morbidity of maternal birth trauma in the developing world. Obstetrical & Gynecological Survey, 51, 568–574.
- Ascher-Walsh, C. J., Capes, T. L., Lo, Y., Idrissa, A., Wilkinson, J., Echols, K., … Genadry, R. (2010). Sling procedures after repair of obstetric vesicovaginal fistula in Niamey, Niger. International Urogynecology Journal, 21, 1385–1390.
- Barone, M. A., Frajzyngier, V., Ruminjo, J., Asiimwe, F., Barry, T. H., Bello, A., … Podder, D. C. (2012). Determinants of postoperative outcomes of female genital fistula repair surgery. Obstetrics & Gynecology, 120, 524–531.
- Barone, M. A., Widmer, M., Arrowsmith, S., Ruminjo, J., Seuc, A., Landry, E., … Gülmezoglu, A. M. (2015). Breakdown of simple female genital fistula repair after 7 day versus 14 day postoperative bladder catheterisation: A randomised, controlled, open-label, non-inferiority trial. Lancet, 386, 56–62.
- Benchekroun, A., Lachkar, A., Soumana, A., Farih, M. H., Belahnech, Z., Marzouk, M., & Faik, M. (1998). Uretero-vaginal fistulas: 45 cases. Annales d Urologie (Paris), 32, 295–299.
- Blaivas, J. G., Heritz, D. M., & Romanzi, L. J. (1995). Early versus late repair of vesicovaginal fistulas: Vaginal and abdominal approaches. Journal of Urology, 153, 1110–1112.
- Brown, H. W., Wang, L., Bunker, C. H., & Lowder, J. L. (2012). Lower reproductive tract fistula repairs in inpatient US women, 1979–2006. International Urogynecology Journal, 23(4), 403–410.
- Browning, A. (2004). Prevention of residual urinary incontinence following successful repair of obstetric vesico-vaginal fistula using a fibro-muscular sling. British Journal of Obstetrics and Gynaecology, 111, 357–361.
- Browning, A., Allsworth, J. E., & Wall, L. L. (2010). The relationship between female genital cutting and obstetric fistulas. Obstetrics & Gynecology, 115(3), 578–583.
- Browning, A., Lewis, A., & Whiteside, S. (2014). Predicting women at risk for developing obstetric fistula: A fistula index? An observational study comparison of two cohorts. British Journal of Obstetrics and Gynaecology, 121, 604–609.
- Browning, A., & Whiteside, S. (2015). Characteristics, management, and outcomes of repair of rectovaginal fistula among 1100 consecutive cases of female genital tract fistula in Ethiopia. International Journal of Gynecology and Obstetrics, 131, 70–73.
- Chen, S.-S., Yang, S.-H., Yang, J.-M., & Huang, W.-C. (2007). Transvaginal repair of ureterovaginal fistula by Latzko technique. International Urogynecology Journal, 18, 1381–1383.
- Creanga, A. A., & Genadry, R. R. (2007). Obstetric fistulas: A clinical review. International Journal of Gynecology & Obstetrics, 99(Suppl. 1), S40–S46.
- Cowgill, K., Bishop, J., Norgaard, A. K., Rubens, C. E., & Gravett, M. G. (2015). Obstetric fistula in low-resource countries: An under-valued and under-studied problem—systematic review of its incidence, prevalence, and association with stillbirth. BMC Pregnancy & Childbirth, 15, 193.
- Das, B., & Snyder, M. (2016). Rectovaginal fistulae. Clinics in Colon and Rectal Surgery, 29, 50–56.
- Debeche-Adams, T. H., & Bohl, J. L. (2010). Rectovaginal fistulas. Clinics in Colon and Rectal Surgery, 23(2), 99–103.
- Donnelly, K., Oliveras, E., Tilahun, Y., Belachew, M., & Asnake, M. (2015). Quality of life of Ethiopian women after fistula repair: Implications on rehabilitation and social reintegration policy and programming. Culture, Health and Sexuality, 17, 150–164.
- El Ayadi, A., Nalubwama, H., Barageine, J., Neilands, T. B., Obore, S., Byamugisha, J., … Miller, S. (2017). Development and preliminary validation of a post-fistula repair reintegration instrument among Ugandan women. Reproductive Health, 14, 109.
- El-Gazzaz, G., Hull, T. L., Mignanelli, E., Hammel, J., Gurland, B., & Zutshi, M. (2010). Obstetric and cryptoglandular rectovaginal fistulas: Long-term surgical outcome; quality of life; and sexual function. Journal of Gastrointestinal Surgery, 14(11), 1758–1763.
- Elkins, T. E., DeLancey, J. O., & McGuire, E. J. (1990). The use of modified Martius graft as an adjunctive technique in vesicovaginal and rectovaginal fistula repair. Obstetrics & Gynecology, 75(4), 727.
- Falandry, L. (1992). Ureterovaginal fistulas: Diagnosis and surgical tactic. Apropos of 19 personal cases. Journal of Urology (Paris), 98, 213–220.
- Fry, R. D., & Kodner, I. J. (1995). Rectovaginal fistula. Annals of Surgery, 27, 113–131.
- Gedik, A., Deliktas, H., Celik, N., Kayan, D., & Bircan, M. K. (2015). Which surgical technique should be preferred to repair benign, primary vesicovaginal fistulas? Journal of Urology, 12, 2422–2427.
- Goh, J. T. (2004). A new classification for female genital tract fistula. Australian and New Zealand Journal of Obstetrics & Gynecology, 44, 502–504.
- Goldaber, K. G., Wendel, P. J., McIntire, D. D., & Wendel, G. D., Jr. (1993). Postpartum perineal morbidity after fourth-degree perineal repair. American Journal of Obstetrics & Gynecology, 168, 489–493.
- Göttgens, K. W., Smeets, R. R., Stassen, L. P., Beets, G., & Breukink, S. O. (2014). The disappointing quality of published studies on operative techniques for rectovaginal fistulas: A blueprint for a prospective multi-institutional study. Diseases of the Colon & Rectum, 57(7), 888–898.
- Handa, V. L., Danielsen, B., & Gilbert, W. M. (2001). Obstetric anal sphincter lacerations. Obstetrics and Gynecology, 98(2), 225–230.
- Harrison, K. A. (1983). Obstetric fistula: One social calamity too many. British Journal of Obstetrics and Gynaecology, 90, 385–386.
- Hilton, P. (1998). Urodynamic findings in patients with urogenital fistulae. British Journal of Urology, 81, 539–542.
- Hilton, P. (2016). Trends in the aetiology of urogenital fistula: A case of “retrogressive evolution”? International Urogynecology Journal, 27(6), 831–837.
- Hilton, P., & Ward, A. (1998). Epidemiological and surgical aspects of urogenital fistulae: A review of 25 years’ experience in southeast Nigeria. International Urogynecology Journal and Pelvic Floor Dysfunction, 9, 189–194.
- Hjortrup, A., Moesgaard, F., & Kjærgård, J. (1991). Fibrin adhesive in the treatment of perineal fistulas. Diseases of the Colon & Rectum, 34(9), 752–764.
- Homsi, R., Daikoku, N. H., Littlejohn, J., & Wheeless, C. R., Jr. (1994). Episiotomy: Risks of dehiscence and rectovaginal fistula. Obstetrical & Gynecological Survey, 49(12), 803–808.
- Ibrahim, T., Sadiq, A. U., & Daniel, S. O. (2000). Characteristics of VVF patients as seen at the specialist hospital Sokoto, Nigeria. West African Journal of Medicine, 19, 59–63.
- Kelly, J. (1991). Fistulae of obstetric origin. Midwifery, 7, 71–73.
- Kelly, J., & Kwast, B. E. (1993a). Epidemiologic study of vesico-vaginal fistulas in Ethiopia. International Urogynecology Journal, 4, 278–281.
- Kelly, J., & Kwast, B. E. (1993b). Obstetric vesicovaginal fistulas: Evaluation of failed repairs. International Urogynecology Journal, 4, 271–273.
- Knuttinen, M. G., Yi, J., Magtibay, P., Miller, C. T., Alzubaidi, S., Naidu, S., … Mar, W. A. (2018). Colorectal-vaginal fistulas: Imaging and novel interventional treatment modalities. Journal of Clinical Medicine, 7(4), 87.
- Konje, J. C., & Ladipo, O. A. (2000). Nutrition and obstructed labor. American Journal of Clinical Nutrition, 72, Supplement 1, 291S–297S.
- Kumar, S., Barapatre, Y. R., Ganesamoni, R., Nanjappa, B., Barwal, K., & Singh, S. K. (2011). Laparoscopic management of a rare urogenital fistula. Journal of Endourology, 25(4), 603–606.
- Landy, H. J., Laughon, S. K., Bailit, J., Kominiarek, M. A., Gonzalez-Quintero, V. H., Ramirez, M., … Consortium on Safe Labor. (2011). Characteristics associated with severe perineal and cervical lacerations during vaginal delivery. Obstetrics & Gynecology, 117(3), 627–635.
- Lawson, J. B. (1989). Tropical obstetrics and gynaecology III. Vesico-vaginal fistula: A tropical disease. Transactions of the Royal Society of Tropical Medicine and Hygiene, 83, 454.
- Lee, D., Dillon, B. E., & Zimmern, P. E. (2013). Long-term morbidity of Martius labial fat pad graft in vaginal reconstruction surgery. Urology, 82, 1261–1266.
- Lewis, G., & de Bernis, L. (Eds.). (2006). Obstetric fistula: Guiding principles for clinical management and programme development. WHO.
- Lo, T. S., Huang, Y. H., Dass, A. K., Karim, N., & Uy-Patrimonio, C. M. (2016). Rectovaginal fistula: Twenty years of rectovaginal repair. Journal of Obstetrics and Gynaecology Research, 42(10), 1361–1368.
- Loungnarath, R., Dietz, D. W., Mutch, M. G., Birnbaum, E. H., Kodner, I. J., & Fleshman, J. W. (2004). Fibrin glue treatment of complex anal fistulas has low success rate. Diseases of the Colon & Rectum, 47(4), 432–436.
- Lufumpa, E., Doos, L., & Lindenmeyer, A. (2018). Barriers and facilitators to preventive interventions for the development of obstetric fistulas among women in sub-Saharan Africa: A systematic review. BMC Pregnancy & Childbirth, 18, 155.
- Lusardi, G., Lipp, A., & Shaw, C. (2013). Antibiotic prophylaxis for short-term catheter bladder drainage in adults. Cochrane Database of Systematic Reviews, 3, CD005428.
- Maheu-Giroux, M., Filippi, V., Maulet, N., Samadoulougou, S., Castro, M. C., Meda, N., … Kirakoya-Samadoulougou, F. (2016). Risk factors for vaginal fistula symptoms in sub-Saharan Africa: A pooled analysis of national household survey data. BMC Pregnancy & Childbirth, 16, 82.
- Maheu-Giroux, M., Filippi, V., Samadoulougou, S., Castro, M. C., Maulet, N., Meda, N., & Kirakoya-Samadoulougou, F. (2015). Prevalence of symptoms of vaginal fistula in 19 sub-Saharan Africa countries: A meta-analysis of national household survey data. Lancet Global Health, 3, e271–e278.
- Mazier, W. P., Senagore, A. J., & Schiesel, E. C. (1995). Operative repair of anovaginal and rectovaginal fistulas. Diseases of the Colon & Rectum, 38(1), 4–6.
- Mellano, E. M., & Tarnay, C. M. (2014). Management of genitourinary fistula. Current Opinion in Obstetrics & Gynecology, 26, 415–423.
- Mocumbi, S., Hanson, C., Hogberg, U., Boene, H., von Dadelszen, P., Bergström, A., … CLIP working group. (2017). Obstetric fistulae in southern Mozambique: Incidence, obstetric characteristics and treatment. Reproductive Health, 14(1), 147.
- Muleta, M., Tafesse, B., & Aytenfisu, H. G. (2010). Antibiotic use in obstetric fistula repair: Single blinded randomized clinical trial. Ethiopian Medical Journal, 48, 211–217.
- Murray, C., Goh, J. T., Fynes, M., & Craey, M. P. (2002). Urinary and faecal incontinence following delayed primary repair of obstetric genital fistula. British Journal of Obstetrics and Gynaecology, 109, 828–832.
- Nardos, R., Menber, B., & Browning, A. (2012). Outcome of obstetric fistula repair after 10-day versus 14-day Foley catheterization. International Journal of Gynecology & Obstetrics, 118, 21–23.
- Nardos, R., Browning, A., & Chen, C. C. (2009). Risk factors that predict failure after vaginal repair of obstetric vesicovaginal fistulae. American Journal of Obstetrics & Gynecology, 200, 578.e1–578.e 4.
- Nielsen, H. S., Lindberg, L., Nygaard, U., Aytenfisu, H., Johnston, O. L., Sørensen, B., … Duffy, S. A. (2009). Community-based long-term follow-up of women undergoing obstetric fistula repair in rural Ethiopia. British Journal of Obstetrics and Gynaecology, 116, 1258–1264.
- Oakley, S. H., Brown, H. D., Yurteri-Kaplan, L., Greer, J. A., Richardson, M. L., Adelowo, A., … Pauls, R. N. (2015). Practice patterns regarding management of rectovaginal fistulae: A multicenter review from the Fellows’ Pelvic Research Network. Female Pelvic Medicine & Reconstructive Surgery, 21(3), 123–128.
- Ojengbede, O. A., Baba, Y., Morhason-Bello, I. O., Armah, M., Dimiti, A., Buwa, D., & Kariom, M. (2014). Group psychological therapy in obstetric fistula care: A complementary recipe for the accompanying mental ill health morbidities? African Journal of Reproductive Health, 18, 155–159.
- Ommer, A., Herold, A., Berg, E., Fuerst, A., Schiedeck, T., & Sailer, M. (2012). German S3-guideline: Rectovaginal fistula. German Medical Science, 10, Doc15.
- Ouedraogo, I., McConley, R., Payne, P., Heller, A., & Wall, L. L. (2018). Gurya cutting and female genital fistulas in Niger: Ten cases. International Urogynecology Journal, 29, 363–368.
- Pernin, F., Fanton, Y., & Dufour, B. (1993). Uretero-uterine fistula: Diagnostic and therapeutic management. Revue française de gynécologie et d'obstétrique, 88, 599–602.
- Raassen, T. J. I. P., Ngongo, C. J., & Mahendeka, M. M. (2014). Iatrogenic genitourinary fistula: An 18-year retrospective review of 805 injuries. International Urogynecology Journal, 25, 1699–1706.
- Rahman, M. S., Al-Suleiman, S. A., El-Yahia, A. R., & Rahman, J. (2003). Surgical treatment of rectovaginal fistula of obstetric origin: A review of 15 years’ experience in a teaching hospital. Journal of Obstetrics and Gynaecology, 23, 607–610.
- Reisenauer, C. (2016). Presentation and management of rectovaginal fistulas after delivery. International Urogynecology Journal, 27, 859–864.
- Rogers, R. G., & Jeppson, P. C. (2016). Current diagnosis and management of pelvic fistulae in women. Obstetrics & Gynecology, 128, 635–650.
- Rosenshein, N. B., Genadry, R., & Woodruff, J. D. (1980). An anatomic classification of rectovaginal septal defect. American Journal of Obstetrics & Gynecology, 137, 4.
- Sapmaz, E., Celik, H., & Semerciöz, A. (2003). Omental graft use in Youssef syndrome. European Journal of Obstetrics & Gynecology and Reproductive Biology, 109, 92–96.
- Schmeller, N. T., Göttinger, H., Schüller, J., & Marx, F. J. (1983). Percutaneous nephrostomy as primary therapy of ureterovaginal fistula. Der Urologe Ausgabe A, 22, 108–112.
- Seim, A. R., Alassoum, Z., Bronzan, R. N., Mainassara, A. A., Jacobsen, J. L., & Gali, Y. A. (2014). Pilot community-mobilization program reduces maternal and perinatal mortality and prevents obstetric fistula in Niger. International Journal of Gynecology & Obstetrics, 127(3), 269–274.
- Shaker, H., Saafan, A., Yassin, M., Idrissa, A., & Mourad, M. S. (2011). Obstetric vesico-vaginal fistula repair: Should we trim the fistula edges? A randomized prospective study. Neurourology and Urodynamics, 30, 302–305.
- Sjøveian, S., Vangen, S., Mukwege, D., & Onsrud, M. (2011). Surgical outcome of obstetric fistula: A retrospective analysis of 595 patients. Acta Obstetricia et Gynecologica Scandinavica, 90, 753–760.
- Smith, L. A., Price, N., Simonite, V., & Burns, E. E. (2013). Incidence of and risk factors for perineal trauma: A prospective observational study. BMC Pregnancy & Childbirth, 13, 59.
- Songne, K., Scotté, M., Lubrano, J., Huet, E., Lefébure, B., Surlemont, Y., … Ténière, P. (2007). Treatment of anovaginal or rectovaginal fistulas with modified Martius graft. Colorectal Disease, 9, 653–656.
- Stoker, J., Rociu, E., Ruud Schouten, W., & Laméris, J. (2002). Anovaginal and rectovaginal fistulas endoluminal sonography versus endoluminal MR imaging. American Journal of Roentgenology, 178, 737–741.
- Tahzib, F. (1983). Epidemiological determinants of vesicovaginal fistulas. British Journal of Obstetrics & Gynecology, 90, 387–391.
- Tazi, K., el Fassi, J., Karmouni, T., Koutani, A., Ibn Attya, A. I., Hachimi, M., & Lakrissa, A. (2000). Vesico-uterine fistula: Report of 10 cases. Progrès en Urologie, 10, 1173–1176.
- Tebeu, P. M., Fomulu, J. N., Khaddaj, S., de Bernis, L., Delvaux, T., & Rochat, C. H. (2012). Risk factors for obstetric fistula: A clinical review. International Urogynecology Journal, 23(4), 387–394.
- Thaddeus, S., & Maine, D. (1994). Too far to walk: Maternal mortality in context. Social Science & Medicine, 38(8), 1091–1110.
- Tsang, C. B., Madoff, R. D., Wong, W. D., Rothenberger, D. A., Finne, C. O., Singer, D., & Lowry, A. C. (1998). Anal sphincter integrity and function influences outcome in rectovaginal fistula repair. Diseases of the Colon & Rectum, 41, 1141–1146.
- Tsivian, M., Tsivian, M., Sidi, A. A., & Tsivian, A. (2012). Uretero-vesico-cervical fistula following a caesarean section: A unique case report. International Urogynecology Journal, 23, 1639–1641.
- Tunçalp, O., Isah, A., Landry, E., & Stanton, C. K. (2014). Community-based screening for obstetric fistula in Nigeria: A novel approach. BMC Pregnancy & Childbirth, 14, 44.
- Tunuguntla, H. S., & Gousse, A. E. (2006). Female sexual dysfunction following vaginal surgery: A review. Journal of Urology, 175(2), 439–446.
- Upadhyay, A., Kunwar, A., Shrestha, S., Pradhan, H. K., Karki, A., & Dangal, G. (2018). Managing ureterovaginal fistulas following obstetric and gynecological surgeries. Journal of Nepal Health Research Council, 16, 233–238.
- Waaldijk, K. (1995). Surgical classification of obstetric fistulas. International Journal of Gynecology & Obstetrics, 49, 161–163.
- Waaldijk, K. (2004). The immediate management of fresh obstetric fistulas. American Journal of Obstetrics & Gynecology, 191, 795.
- Walker, S. H., Ambauen-Berger, B., Saha, S. L., & Akhter, S. (2018). Quality of life among women in Bangladesh following ileal conduit urinary diversion operations for irreparable vesicovaginal fistula and bladder exstrophy: Observational study. British Journal of Obstetrics and Gynaecology, 125, 616–622.
- Wall, L. L. (2006). Obstetric vesicovaginal fistula as an international public-health problem. Lancet, 368, 1201–1209.
- Wall, L. L., Karshima, J. A., Kirschner, C., & Arrowsmith, S. D. (2004). The obstetric vesicovaginal fistula: Characteristics of 899 patients from Jos, Nigeria. American Journal of Obstetrics & Gynecology, 190, 1011–1019.
- Wanjala, A., Mwangi, H., & Mabeya, H. (2018). Pattern of ureteric pathology presenting to a fistula centre in western Kenya. Advances in Urology, 2018, Article ID 5056049.
- Warner, R., Beardmore-Gray, A., Pakzad, M., Hamid, R., Ockrim, J., & Greenwell, T. (2019). The cost effectiveness of vaginal versus abdominal repair of vesicovaginal fistulae. International Urogynecology Journal, 1–7.
- Watt, M. H., Wilson, S. M., Joseph, M., Masenga, G., MacFarlane, J. C., Oneko, O., & Sikkema, K. J. (2014). Religious coping among women with obstetric fistula in Tanzania. Global Public Health, 9, 516–527.
- Watt, M. H., Wilson, S. M., Sikkema, K. J., Velloza, J., Mosha, M. V., Masenga, G. G., … Nyindo, P. M. (2015). Development of an intervention to improve mental health for obstetric fistula patients in Tanzania. Evaluation and Program Planning, 50, 1–9.
- Wexner, S. D., Ruiz, D. E., Genua, J., Nogueras, J. J., Weiss, E. G., & Zmora, O. (2008). Gracilis muscle interposition for the treatment of rectourethral, rectovaginal, and pouch-vaginal fistulas: Results in 53 patients. Annals of Surgery, 248(1), 39–43.
- Wilson, S. M., Sikkema, K. J., Watt, M. H., & Masenga, G. G. (2015). Psychological symptoms among obstetric fistula patients compared to gynecology outpatients in Tanzania. International Journal of Behavioral Medicine, 22, 605–613.
- Youssef, A. F. (1957). Menouria following lower segment cesarean section; a syndrome. American Journal of Obstetrics & Gynecology, 73, 759–767.
- Zambon, J. P., Batezini, N. S., Pinto, E. R., Skaff, M., Girotti, M. E., & Almeida, F. G. (2010). Do we need new surgical techniques to repair vesico-vaginal fistulas? International Urogynecology Journal, 21, 337–342.