By Robert W. Rebar, MD
Professor, Department of Obstetrics and Gynecology, Western Michigan University Homer Stryker M.D. School of Medicine, Kalamazoo
Dr. Rebar reports no financial relationships relevant to this field of study.
In a large population-based cohort study, genitourinary fistulas were increased significantly if ureteral and/or bladder injuries were not identified and treated at the time of hysterectomy.
Dallas KB, et al. Urologic injury and fistula after hysterectomy for benign indications. Obstet Gynecol 2019;134:241-249.
Despite the fact that more than 600,000 hysterectomies are performed in the United States each year, information about genitourinary injury after hysterectomy has been limited. We can now add knowledge gleaned from a population-based cohort study of urologic injuries suffered by all women (except for those seen in Veterans’ Administration hospitals) undergoing hysterectomy for benign indications. Using statewide data from California between the years of 2005 to 2011, information from this population aids in our understanding of the incidence and risk factors for the effect of urinary injuries on subsequent fistula formation.
The authors included 296,130 women (mean age 47.4 years) undergoing hysterectomy in the study. Open abdominal hysterectomy was performed in 36.9% of the women, laparoscopic hysterectomy performed in 26.1%, vaginal hysterectomy performed in 21.2%, and a laparoscopic-assisted vaginal approach was used in 12.3%. Fully 54.0% of the hysterectomies were performed for leiomyomas, 18.8% for endometriosis, and 29.3% for uterine bleeding, with 24% having multiple reasons for surgery. In addition to hysterectomy, 16.7% also underwent pelvic organ prolapse repair and 2.9% underwent a concomitant incontinence procedure.
Approximately 1.8% (5,455) had at least one genitourinary injury, with 2,817 (1.0%) ureteral injuries, 2,058 (0.7%) bladder injuries, and 834 (0.3%) genitourinary fistulas being identified. Overall, 86.2% of these injuries were identified (with 76.4% identified and repaired at time of surgery) and 13.8% were unidentified until they presented subsequently with a fistula. Fistula formation was markedly reduced if the injury was identified immediately for both ureteral (0.7% vs. 3.4%; odds ratio [OR], 0.28; 95% confidence interval [CI], 0.14-0.57) and bladder (2.5% vs. 6.5%; OR, 0.37; 95% CI, 0.16-0.83) injuries. The rate of fistula formation for those suffering concomitant ureteral and bladder injuries was twice as high when they were identified immediately and more than nine times higher for those identified later (2.7% vs. 25.0%), but the numbers were quite small in this latter group. Indwelling stent placement as the only means of treating a ureteral injury identified at hysterectomy was successful in preventing any future operative repair in 99% of cases; if identified after surgery, stent placement was successful less than 40% of the time.
Several factors altered the risk of injury. Genitourinary injuries at the time of hysterectomy were slightly but significantly more common in those undergoing open abdominal hysterectomy (2.0%) compared with a vaginal (1.5%) or a laparoscopic (1.7%) approach. Injuries were also significantly more common in cases undergoing concomitant pelvic organ prolapse repair (2.1% vs. 1.8%), an incontinence procedure (2.7% vs. 1.8%), or mesh use for prolapse repair (3.1% vs. 1.8%). There was a higher rate of injury in women converted from another approach to laparotomy (5.2% vs. 1.8%). Genitourinary injuries were also more common in women with a diagnosis of endometriosis (2.3%) compared with leiomyomas (1.8%) or abnormal uterine bleeding (1.7%). Injuries were more common in facilities in the bottom quartile of overall hysterectomy surgical volume (2.6%) compared to the rest of the cohort (1.8%).
COMMENTARY
It has been estimated that as many as one-third of women in the United States undergo hysterectomy by the age of 60 years.1 Thus, as the authors noted in the introduction, the potential burden of genitourinary complications following hysterectomy is not trivial. It is for just this reason that I elected to review this study.
Most (76.4%), but not all, of the urinary injuries were identified and repaired at the time of surgery. However, 18.6% of the ureteral injuries and 5.5% of the bladder injuries were not identified until later. Immediate identification and repair significantly reduced the odds of developing a genitourinary fistula — and that is one of the major findings of this study. Thus, it is critical to identify such injuries, regardless of the surgical approach to the hysterectomy. The authors suggested that cystoscopy, although not 100% sensitive, should be considered as a part of every hysterectomy and seem to accept a study indicating that cystoscopy can reduce the incidence of urinary tract injuries.2 In contrast, the AAGL in its 2012 practice guideline concluded that the low level of evidence and the limited data existing precluded recommending that cystoscopy should be performed with every laparoscopic hysterectomy.3 The present cohort study reinforces the need for a large multicenter randomized trial to address the possible value of cystoscopy.
Dallas et al also found that simply placing ureteral stents at the time of hysterectomy significantly reduced the need for later ureteral repair surgery. This observation leads to the conclusion that stents should be used liberally whenever ureteral damage or even “kinking” is suspected. In contrast, later placement of stents is much less effective.
Not surprisingly, they also indicated the need to be particularly mindful of careful repair in women suffering both ureteral and bladder injuries at hysterectomy, as these women particularly are prone to develop later fistulas. The risk of fistula formation appeared increased if these injuries were not identified at the time of surgery, but here the numbers were quite small.
The authors also provided additional information about specific risk factors predisposing to fistula formation. Laparotomy was associated with increased risk compared to vaginal, laparoscopic, or laparoscopically assisted vaginal hysterectomy. Addition of an incontinence procedure or the use of mesh also increased the risk. With regard to the indication for hysterectomy, a diagnosis of endometriosis significantly increased the risk compared to hysterectomies performed for either leiomyomas or uterine bleeding. The authors also reported that urinary tract injuries were significantly increased in Asian women compared to white, black, Hispanic, and other women; they hypothesized that this may be because Asian women have the highest rate of hysterectomy performed for endometriosis.4 Perhaps not surprisingly, the risk of injury was increased in facilities where hysterectomies are performed only rarely.
Despite the significance of these risk factors, the authors offered a note of caution: Most of the effect sizes of the statistically significant associations reported in this large cohort study fall within the range of potential bias (i.e., an odds ratio of > 1.0 or < 2.0 or < 1.0 and > 0.3. We frequently forget that associations do not prove causality. Our liberal reliance on “big data” today can lead to incorrect conclusions and to the promulgation of false beliefs. Thus, these findings should be viewed cautiously until such time as they are confirmed by prospective studies.
Although the cautions raised by the authors themselves are legitimate, all surgeons need to consider the findings of this study as they perform hysterectomies. They constitute the best data we have to date. The authors believed their findings support the liberal use of cystoscopy at the time of hysterectomy, with a low threshold for ureteral stent placement in individuals with suspected ureteral injury to reduce the risk of subsequent fistula development. It is difficult to argue with those conclusions. Cystoscopy is no more difficult to perform than is hysteroscopy and arguably should be mastered by everyone completing a residency in obstetrics and gynecology. Whether cystoscopy is always necessary with hysterectomy can be debated, but the need for surgeons to be careful not to damage the urinary tract during hysterectomy cannot be. Erring on the side of caution is seldom wrong.
REFERENCES
- Wright JD, et al. Nationwide trends in the performance of inpatient hysterectomy in the United States. Obstet Gynecol 2013;122:233-241.
- Chi AM, et al. Universal cystoscopy after benign hysterectomy: examining the effects of an institutional policy. Obstet Gynecol 2016;127:369-375.
- AAGL Practice Report: Practice guidelines for intraoperative cystoscopy in laparoscopic hysterectomy. J Minim Invasive Gynecol 2012;19:407-411.
- Jacoby VL, et al. Racial and ethnic disparities in benign gynecologic conditions and associated surgeries. Am J Obstet Gynecol 2010;202:514-521.
