Ureteral injury is a serious complication of vaginal hysterectomy and other gynecologic surgery. Its incidence during vaginal surgery is estimated to be 0.9% to 1.7%. Although intraoperative cystoscopy can provide a timely diagnosis and expedite management, it is not routinely performed. Barriers to routine use include cost and training. As an alternative to cystoscopy, ureteral injury may be detected by symptoms, an increase in perioperative serum creatinine level, or both. The primary aim of this study was to determine the rate of ureteral injury and associated costs in patients undergoing selective cystoscopy during vaginal hysterectomy. Another aim was to determine whether changes in perioperative creatinine level could be used to reliably diagnose ureteral injury in the absence of intraoperative cystoscopy. A secondary analysis of a retrospective cohort study was performed. A total of 593 patients underwent vaginal hysterectomy for benign indications, with or without additional pelvic floor repairs, from January 2, 2004, through December 30, 2005, at Mayo Clinic (Rochester, Minn). A logistic regression model was constructed to determine the propensity for a surgeon to perform intraoperative cystoscopy, in order to allow adjustment for the likelihood of detecting ureteral injury in a given surgical case. The rate of ureteral injury and associated cost was determined in patients undergoing surgery both with and without cystoscopy. In cases in which cystoscopy was not performed, standard practice of postoperative creatinine level determination allowed the authors to investigate the feasibility of using perioperative change in creatinine level to detect ureteral injury. Of the 593 eligible patients included in the analysis, 230 (38.8%) underwent cystoscopy during the hysterectomy. Ureteral injuries occurred in 6 patients (2.6%) in the cystoscopy group and 5 (1.4%) in the no-cystoscopy group; the odds ratio was 1.92, with a 95% confidence interval (CI) of 0.58 to 6.36, P = 0.35. After adjusting for the propensity to perform cystoscopy, the association was further attenuated (odds ratio, 1.31; 95% CI, 0.19–9.09). Five of the 6 were detected intraoperatively; of those, 4 were successfully managed intraoperatively. Adjusted mean-predicted costs for patients undergoing and not undergoing cystoscopy were $10,686 (95% CI, $7500–$13,872) and $10,217 (95% CI, $6894–$13,540), respectively; P = 0.28. The 5 patients with ureteral injury in the no-cystoscopy group had a median increase in perioperative creatinine level of 0.2 mg/dL, whereas patients without injury had a median decrease in creatinine level of 0.1 mg/dL (P < 0.001). Overall, the likelihood of ureteral injury with and without cystoscopy is similar, and predictive costs do not significantly differ. The use of postoperative creatinine level to detect ureteral injury is highly sensitive, but limited by a low positive predictive value and variable range. Based on these data, the authors recommend routine universal screening cystoscopy for patients undergoing vaginal hysterectomy.
Read full abstract