This chapter discusses recognition and management of postsurgical fistulas in the developed world. Genitourinary fistulas in the developed world are primarily postsurgical, although the tenets of repair are also applicable for operative obstetrical injury. There is a separate chapter in this textbook which addresses extensive field-effect fistula secondary to prolonged obstructed obstetrical labor.

It is imperative that surgeons have a thorough understanding and appreciation of pelvic anatomy in order to reduce genitourinary injury. Anatomic relationships of the ureter, bladder, and urethra have been further elucidated in recent studies.¹,² Avoidance of fistulas following gynecologic surgery requires strict attention to tissue planes, hemostasis, tissue vascularization, and appropriate use of instruments.

The risks for genitourinary fistula are affected by the radicality and route of surgery. Hilton and Cromwell³ reported on the risks of vesicovaginal fistula following hysterectomy in a National Health Service cohort study. The highest rate occurred in patients undergoing surgery for cervical cancer, and the lowest incidence was found in patients undergoing vaginal prolapse surgery. In 286,053 women undergoing benign hysterectomy, 339 fistulas were later identified (0.12%). The risk for vesicovaginal fistula following vaginal prolapse surgery was estimated to be 1 in 3,861 (0.02%).

A 2019 California population-based cohort study assessing fistula after hysterectomy for benign indications reviewed 296,130 patients.⁴ A total of 5,455 (1.80%) patients experienced at least one genitourinary injury. There were 2,817 (1.0%) ureteral injuries, 2,058 (0.7%) bladder injuries, and 834 (0.3%) patients developed subsequent genitourinary fistulas. The number of genitourinary injuries recognized during surgery was 4,701 (86.2%), and the number of unrecognized injuries was 754 (13.8%). In the injury cohort, 174 women (3.2%) sustained a complex injury involving both the bladder and ureter. These combined injuries had a rate of fistula formation which was 2 times higher compared to single-site injury patients, even with immediate identification. The risk of fistula occurrence was 9 times higher if recognition of complex injury was not immediate (2.7% vs. 25.0%). The overall risk for developing a fistula in the total injury cohort despite immediate recognition and repair was 9.5%.

Repairing a recognized ureteral injury or cystotomy must be done in a way to optimize healing and reduce the risk of a subsequent fistula. Principles of assuring adequate blood supply, suture lines without tension, catheter drainage, and tissue interposition are important in minimizing the risk of fistula. In a study of recognized incidental cystotomy at the time of benign hysterectomy, 11.7% developed a subsequent fistula.⁵

There is a significant difference in morbidity when comparing recognized and unrecognized genitourinary injuries at the time of surgery. A National Surgical Quality Improvement Program (NSQIP)-based study reviewed the incidence and risk factors for genitourinary injury recognized in the postoperative period following hysterectomy for benign indications.⁶ The study assessed 45,139 patients undergoing benign hysterectomy with an incidence of lower urinary tract complication at 0.2%. The unrecognized ureteral obstruction rate was 0.1% with 0.07% of patients developing a ureteral fistula and 0.6% experiencing a bladder fistula. Unrecognized ureteral obstruction has significant morbidity and requirements for additional hospitalization and surgery. A review of patients with recognized and unrecognized ureteral injury was reported in a study using a California inpatient database reviewing patients with ureteral
injuries following hysterectomy.⁷ The study identified 1,753 ureteral injuries in 223,872 patients following hysterectomy, with a delayed recognition rate of 62.4% (1,094 patients). The unrecognized injuries increased the risk of acute renal insufficiency (aOR 23.8, 95% confidence interval [CI] 20.1 to 28.2) and death (a1.4, 95% CI 1.03 to 1.9, P = .0032).

Unrecognized bladder injury during hysterectomy frequently occurs during the dissection of the bladder base from the underlying vagina and cervix. This dissection should always be done sharply, with attention paid to the lateral bladder pillars and fine-tissue visual characteristics. Several centimeters of endopelvic fascial dissection prior to vaginal incision allow separation of tissue planes for subsequent cuff closure. If bladder wall thinning is observed, it should be reinforced by fine-gauge, delayed absorbable interrupted sutures with attention to any vascular compromise. A three-way Foley catheter can allow bladder filling during surgical dissection to assist in recognition of boundaries.

The ureters should be identified throughout their course during abdominal and laparoscopic surgery and assessed during vaginal surgery. It is incumbent on the surgeon to carefully examine all tissues and structures at the termination of the procedure to rule out any unrecognized injury. This would include use of cystoscopy to assess bladder integrity and bilateral efflux of urine at the termination of the procedure.⁸

Additional detail on avoidance of injury at pelvic surgery can be found in Chapter 28 of this textbook.

A high index of suspicion is required for recognition of delayed bladder and ureteral injuries. The postoperative patient who presents with fluid leaking per vagina warrants careful clinical assessment. This includes a thorough history and physical examination. A fistula may present days or weeks following a pelvic surgery. Late presentations are often related to delayed necrosis of tissues.

Patients with urinary incontinence have episodes of leakage between longer dry periods. Patients who have a fistula may experience almost constant drips of urine depending on body position. The vast majority of postsurgical fistulas are supratrigonal in location. The cuff should carefully be examined to rule out fluid coming from the peritoneum or bladder. Simple filling of the bladder in the office with repeat inspection of the vagina may show influx of fluid into the vagina along the cuff. The tampon test of Moir can often identify small pinpoint fistulas after placement of dye within the bladder fluid. Three gauze pledgets are placed into the vagina with subsequent placement of dyed saline into the bladder. The patient is asked to stand, cough, walk, and void prior to pledget removal. The pledget closest to the introitus will be stained from voiding. If the upper portion of the second pledget or the third pledget is stained, a fistula should be suspected. A ureterovaginal fistula will not show coloration unless dye was given intravenously or if the patient had previously received oral phenazopyridine. A computed tomography (CT) urogram or other imaging is mandatory to rule out ureteral or complex fistula (concomitant ureteral and bladder fistulas) in all patients requiring repair. A concomitant ureteral injury may be present in up to 12% of cases of vesicovaginal fistula.⁹ If a ureteral fistula is suspected, a retrograde pyelogram is beneficial for localization.

Some patients with vesicovaginal fistulas may have a negative CT urogram in the supine position but immediately leak urine via the fistula when standing. Cystoscopies can also fail to show an overt fistula if scarring and/or inflammation is present. This further illustrates the importance of a detailed history and high index of suspicion, as some patients have received unnecessary incontinence operations in the presence of an occult fistula.

The most important features of fistula management involve careful evaluation and assessment of the entire clinical picture prior to surgery. This includes careful assessment of tissue integrity, status of the ureters, bladder characteristics, and urethral sphincteric function. History should include past issues relative to the bladder and social considerations. The surgeon should be cognizant of overall bladder capacity, vaginal limitations, tissue inflammation, sphincteric function, presence of multiple fistulae, and ureteral integrity prior to planning the operative approach. The surgeon can then make thoughtful decisions regarding timing and route of surgery.

There is an ongoing debate regarding the timing of fistula repair following surgical injury. This is reflective of the body’s tendency to create deleterious inflammation following surgery. Early recognition of injury allows early repair (either at the time of surgery or within several days following operation). Patients who present greater than 10 days following injury should have assessment of the fistula by cystoscopy to determine the size and degree of reactive inflammation, swelling, and absence of infection. Historically, if significant inflammation is noted, recommendations have been to delay surgery for a period of 2 to 3 months from the original operative insult in order to allow resolution of inflammation and improve healing of suture lines. There are articles in the literature advocating early repair of posthysterectomy fistulas no
matter the time of discovery. The vast majority of these articles employ an abdominal or transvesical approach with larger excision of surrounding tissue.⁹ Those who advocate early repair of post-obstetrical fistulas in the developing world use a vaginal approach.¹⁰

The surgical goal is healing of the fistula with optimal subsequent bladder function. In order to reduce the risk of persistent fistula, surgical principles of precise sharp tissue dissection, observation of vascularity, avoidance of tension along suture lines, watertight closure, placement of an interposition flap, and postoperative drainage are followed.

The length of postoperative bladder drainage may depend on characteristics of the fistula and surrounding tissue. A suprapubic catheter placed for duration of 2 weeks, followed by a cystogram prior to removal, is a reasonable clinical approach. There are ongoing studies assessing optimal length of postoperative bladder drainage in large obstetrical fistulas secondary to tissue necrosis.¹¹

Studies assessing use of interposition flaps have shown mixed results depending on the type of fistula and causation. It is clinically recommended to protect suture lines with an interposition flap because it is not time-consuming relative to the patient’s bother and may decrease the risk of persistent fistula. It should always be recognized that an interposition flap will not compensate for an inadequate bladder closure. Omental and Martius flaps can be useful when increased blood supply is beneficial for tissue healing.¹²

A systematic review and meta-analysis in the management of the vesicovaginal fistulas following benign gynecologic surgery was reported in 2017.¹³ Following review, 124 articles were included for assessment involving 1,379 patients. The fistula closure rate in this group was 97.98%. The transvaginal approach was performed in the majority of patients (39%), followed by the transabdominal/transvesical route (36%), and a laparoscopic/robotic approach (15%). Prolonged catheter drainage was initially used in 239 patients, with 19 (8%) achieving success without surgery. The remainder of patients not responding to conservative catheter drainage underwent surgical repair.

An analysis of fistula characteristics, treatments, and complications of surgical repair were reported using the American College of Surgeons (ACS) NSQIP database. Two hundred patients were reported of which 65% were repaired vaginally. Compared to the vaginal approach, the abdominal approach had higher overall morbidity (22% vs. 7%, P = .017); longer hospital stay; and were more likely to be associated with sepsis, blood transfusion, and readmission.¹⁴ This difference in morbidity may be reduced in the future secondary to minimally invasive surgery and limitations of dissection, including unnecessary bivalving of the bladder.