Urinary tract infection (UTI) is common and is a significant contributor to the high cost of health care in the United States, with more than $5 billion in direct costs in 2010. As a result of anatomic and physiologic differences, women have higher rates of UTI compared with men. Half of all women experience at least one UTI in their lifetimes, and more than 10% of postmenopausal women experience recurrent UTI. Furthermore, pelvic surgery is associated with increased risk of UTI with reported rates up to 50% in women undergoing certain reconstructive procedures.1,2 All of these factors make UTI in women after gynecologic surgery an important subcategory of UTI that is associated with high utilization of ambulatory and inpatient care. Consequently, reduction in postoperative UTI after pelvic surgery can result in a significant reduction in health care costs.3
Postoperative UTI rate has been proposed as an indicator of surgical quality.4 In 2012, both the Centers for Medicare & Medicaid Services and the Joint Commission on Accreditation of Healthcare Organization required all Medicare providers to report the rates of surgical site infections including UTI in a public registry.5 In addition, it has been increasingly suggested that postoperative UTI be used as a specific quality marker for pelvic surgery and potentially linked to reimbursement.6,7 One issue that needs to be investigated is the high variation in reported incidences of postoperative UTI, which is highlighted by several publications.5,8–11 Another issue is the lack of published studies that report on postoperative UTI as their primary outcome and the lack of identified risk factors that can be used to estimate the probability of UTI.5,11
The aim of this study was to report the proportion of the women who develop UTI within 8 weeks after hysterectomy for benign conditions or hysterectomy combined with pelvic reconstructive surgery and to use regression models to identify independent risk factors of UTI in this cohort of women. Our hypothesis is that there are independent risk factors that are associated with UTI after hysterectomy for benign gynecologic conditions or as part of pelvic reconstructive surgeries.
MATERIALS AND METHODS
This is a retrospective cohort study designed to report the rates of UTI and identify independent variables associated with UTI after hysterectomy for benign conditions and those done as part of pelvic reconstructive surgery. The article was written following the STROBE guidelines for reporting observational studies.12 The protocol of this study was approved by the Mayo Clinic institutional review board (IRBe-14-005150).
The study population included women, 18 years of age and older, undergoing hysterectomy for benign gynecologic conditions or hysterectomy combined with pelvic reconstructive surgery at our institution from January 1, 2012, through June 30, 2014. The end date was decided to avoid capturing effects for a newly implemented UTI reduction protocol, which started in July 2014. To be included in the study, patients had to provide their consent for use of their medical records in research, which is obtained routinely for all patients in our practice. Exclusion criteria included a preoperative or intraoperative diagnosis of malignancy. We also excluded women who had a diagnosis of UTI within 7 days before surgery to avoid including patients with preoperative UTI. Eligible patients were identified using the International Classification of Diseases, 9th Revision, Clinical Modification (procedure codes for each of the included procedures and also using a word search software, Surgical Operative Note eXplorer) to get a complete list of all procedures performed during the study period. Surgical Operative Note eXplorer is a web-based application that provides the ability to search surgical operative reports using certain criteria and word searches. A data retrieval specialist was consulted to outline a search strategy and to optimize identification of patients in each of the “a priori” selected procedures. Those procedures were grouped based on the route of the hysterectomy (vaginal vs others), the use of cystoscopy, and having a combined procedure for treating pelvic organ prolapse (POP; sacral colpopexy, vaginal apical repair, colporrhaphy, or urinary incontinence [UI]). Cystoscopies are routinely done in all pelvic reconstructive procedures. In our institution, routine use of cystoscopy in robotic, open, or vaginal hysterectomies is not required. Final selections were verified through review of the surgical report of those identified patients. Two investigators screened the records of all identified patients to ascertain inclusion in the study.
Using REDCap 4.13.17-2014, we designed a database for the case report form including variables related to baseline characteristics and outcomes for this current study. A separate data collection instrument for ascertaining UTI outcome was also designed. Data were entered for individual patients by the following authors: S.A.E.-N., R.S., J.J.S., D.C.L., and C.A. After completing the design of the data collection sheet in REDCap, a run of 10 patients was done for all authors entering data and reviewed to make sure the data were entered correctly. Data collection included demographic data, clinical gynecologic data, medical comorbidities, examination data, preoperative medications, and preoperative urinary instrumentation data including cystoscopy, and urodynamic results and operative data.
The primary outcome was a diagnosis of a UTI within 8 weeks after the procedure. In our institution, patients typically present for follow-up approximately 6 weeks for their final postoperative visit; we used 8 weeks to capture any variation in the timing of this visit and reduce bias related to lost to follow-up. The Centers for Disease Control and Prevention definition of symptomatic UTI was modified to fit the retrospective nature of this study to include one of the following: 1) at least one symptom or sign with positive urine culture evidence of infection (greater than 105 colony-forming units); 2) symptomatic UTI with a physician diagnosis or decision to treat; or 3) symptomatic UTI with positive urine (positive dipstick), pyuria, positive Gram stain, or two cultures with greater than 102 colony-forming units.13,14 Timing of diagnosis was recorded. The reported organisms and culture sensitivities were also abstracted. Information on availability of relevant clinical follow-up within the 8-week postoperative period was collected to distinguish between patients without a documented postoperative UTI and those who were lost to follow-up.
For our first aim of reporting the rates of postoperative UTIs, the sample size was calculated to estimate the UTI rates with reasonable precision as determined by the half width of the 95% CI. Estimates for the anticipated UTI rates ranged from 2.5% to 20%, which were based on previously published studies.5,9,15,16 The 95% CI for a rate of 10% is 4.1–15.9% (half width of 5.9%) based on 100 patients compared with 5.8–14.2% (half width of 4.2%) based on 200 patients. For the second aim, which included development of a model, assuming we have a sample size of 1,000 patients and the overall UTI rate is 10%, we anticipated having 100 cases with a postoperative UTI and 900 without. This would allow us to fit a multivariable prediction model with a total number of covariates accounting for 10 degrees of freedom using the rule of thumb of 10 events per one variable.17
Descriptive results were summarized using standard descriptive statistics: frequency and percentage for categorical data and mean and SD or median and interquartile range for continuous or count data. The rate of postoperative UTI within 8 weeks of surgery was calculated separately for each of the a priori decided procedure groups; 95% CIs were constructed based on the normal approximation for the binomial distribution. Binomial CIs were based on Clopper and Pearson.18
Univariate and multivariable logistic regression models were fitted to evaluate associations with postoperative UTI. Associations were summarized by reporting odds ratios (ORs) and corresponding 95% CIs from the parameters estimated in the models. All the variables that have been previously reported or thought to be related to development of a UTI were evaluated in univariate analyses. We also compared UTI rates in women undergoing female pelvic reconstructive procedures with those who had hysterectomy for other benign indications. This was based on having surgery for a POP or incontinence, namely a sling procedure. To construct the predictor model, we first considered demographic, presenting symptoms, clinical examination data, and any lower urinary tract instrumentation before surgery. To identify the best fit, continuous variables were evaluated in a univariate analysis as nontransformed, log-transformed, or using restricted cubic splines. Other patient characteristics and preoperative variables were considered for inclusion in the final multivariate model if they had a P<.2 in the univariate analysis. Development of a parsimonious model was explored using stepwise and backward variable selection. All statistical tests were two-sided and P values <.05 were considered statistically significant. Predictive discrimination of the final multivariable model was assessed using the concordance (c-index), which is a measure of a model's overall predictive ability. The c-index varies from 0 to 1, and a value of 0.5 denotes no predictive discrimination. The c-index derived from the original model was expected to overly optimistic; therefore, a less biased estimate (internal validation) was derived using 300 bootstrap resamples of the same size as the original sample as outlined by Harrell et al.17 Calibration was assessed by examining how far the model-predicted UTI probabilities are from the actual observed UTI rates based on patients grouped into subgroups based on their UTI probability predicted by the final multivariable model. Statistical analysis was performed using JMP 12.0 and SAS 9.4.
A total of 10,137 women underwent gynecologic surgery at our institution during the study period. Of those, 1,174 (11.6%) had hysterectomies for benign or pelvic reconstructive indications and were considered for inclusion in the study. Of those, 18 were excluded as a result of a diagnosis of UTI before surgery. Finally, 1,156 met the selection criteria and were included in the study (Fig. 1).
The mean age was 49.9±12.7 years including 172 (14.8%) women older than 65 years. In the univariate regression analyses, the following demographic and presenting symptoms had significant associations with the UTI outcome: age 50 years or older, postmenopausal status, POP presentation, UI presentation, recurrent UTI, constipation, vaginal atrophy on pelvic examination, any objective evidence of POP on pelvic examination, and evidence of incomplete bladder emptying with postvoid residual of greater than 150 mL (Table 1).
Of 1,156 included in the study, 136 had a diagnosis of UTI after surgery with an overall rate of 11.8% (95% CI 10.0–13.8). The unadjusted rates are presented in Figure 2 for a priori selected subgroups of hysterectomy based on route and combined procedures with rates ranging from 0% to 26% (Fig. 2). Women who underwent hysterectomy for benign gynecologic conditions had an overall UTI rate of 7.3% (95% CI 5.6–9.3) compared with 21.7% (95% CI 17.6–26.4) in women who had hysterectomy combined with any pelvic reconstructive procedure. Similarly, women who had a midurethral sling procedure had a higher UTI rate of 19.1% (95% CI 12.7–28.1) compared with a rate of 9.5% (95% CI 4.9–16.4) in those who did not. In addition, women who had surgery for POP had a higher UTI rate of 22.9% (95% CI 15.2–31.4) compared with 7.2% (95% CI 3.5–13.9) in those who did not have surgery for POP (Table 2).
Preoperative instrumentation including office cystoscopy, urodynamic testing, catheterization, and self-intermittent catheterization teaching was associated with significantly higher UTI rates compared with women who did not have any instrumentation (Table 3). For operative data, having POP or UI as an indication for surgery was associated with a significant increase in the UTI rate with unadjusted ORs of 4.03 (95% CI 2.79–5.83) and 2.15 (95% CI 1.34–3.46), respectively. Vaginal route of surgery was associated with more postoperative UTIs as well. For specific combinations of surgeries, all types of vaginal POP repairs were associated with an increased risk for UTI as well as having a midurethral sling and intraoperative cystoscopy (Table 4). Having an indwelling catheter after surgery was associated with an increased UTI rate compared with no catheter with an unadjusted OR of 1.6 (95% CI 1.04–2.44). The need for prolonged bladder drainage after discharge was also significantly associated with increased UTI with an OR of 3.35 (95% CI 2.31–4.86). Patients who developed UTI after hysterectomy had longer operative time (greater than 2 hours), higher estimated blood loss (greater than 250 mL), and a higher incidence of intraoperative urinary injuries (Table 4).
The final model is reported in Table 5 and included premenopausal status with an adjusted OR of 1.80 (95% CI 1.11–2.99); anterior vaginal wall prolapse of any grade on examination with an adjusted OR of 4.39 (95% CI 2.77–6.97); and postvoid residual greater than 150 mL with an adjusted OR of 2.38 (95% CI 1.12–4.36). The final model was adjusted to hormone therapy (HT) use. As planned, we then adjusted for route of hysterectomy (vaginal, laparoscopic including robotic and abdominal), and we adjusted for the type of combined procedures. This adjustment did not significantly change the estimates for the three independent variables (premenopausal status, anterior vaginal wall prolapse diagnosed on pelvic examination, and postvoid residual greater than 150 mL) (Table 5). Only intraoperative cystoscopy as a combined procedure was added to the final model as an independent variable with an adjusted OR of 1.83 (95% CI 1.00–3.33). The overall predictive accuracy after adjusting for route and combined surgeries did not significantly changed the c-index (Table 5). Table 6 includes the predicted and observed UTI rates based on the final model. Based on the model, predicted UTI rates after hysterectomy range from 4.3% to 59.4%.
In this study, the overall rate of UTI after hysterectomy was 11.8%. Interestingly, the rate varied widely, between 5% and 26%, based on route of hysterectomy and whether concomitant procedures were performed for POP or UI treatment (Fig. 2). Women who underwent hysterectomy as part of a pelvic reconstructive surgery had an approximately threefold increase in UTI rates compared with women who underwent hysterectomy alone. Independent variables associated with developing UTI after hysterectomy included premenopausal status, anterior vaginal wall prolapse finding on pelvic examination, and documented postvoid residual greater than 150 mL before surgery. This final model was adjusted for HT use. Using this model, we were able to predict rates of UTI after hysterectomy and we developed a simple table to aid clinicians and their patients in identifying the risk of UTI based on data from preoperative evaluation (Table 6). This preoperative stratification for UTI risk after hysterectomy may be a helpful aid for discussing route of surgery at the time of preoperative patient counseling. We also believe this model would help in identifying higher risk groups that can be specifically targeted with strategies to prevent or reduce their risk of postoperative UTI after external validation in a different population.
In previously published literature, the reported UTI rate varies widely based on the definition of UTI, route of hysterectomy, and type of combined procedures. When most of the included patients underwent an abdominal or laparoscopic approach to their hysterectomy, the hysterectomy was done for benign conditions, and minimal or no urinary instrumentation, the UTI rates are lower at less than 5%.5,8 When only patients who underwent pelvic reconstructive surgery are included, the rates are much higher with rates ranging from 7% to 45%.9–11 In one study that only included women who underwent uterosacral suspension, the overall rate of UTI was 14% and in one group of women who needed postoperative extended bladder drainage and went home with indwelling transurethral catheter, the rate was greater than 50%.9 Similarly, in our study, women who had increased postvoid residual after surgery and required extended bladder drainage had a UTI rate that was three times higher compared with women who did not. In addition, in this study, the rate of vaginal hysterectomy was greater than 50% and in all vaginal hysterectomies, apical suspension and cystoscopy were done to resupport the vaginal apex after the hysterectomy. These variations in UTI rates based on the variations in the indication for hysterectomy, route of hysterectomy, and type of combined procedures can explain why the overall rate in this study of 11.9% is higher than what is previously reported in the literature.
There have been prior attempts to identify predictors of surgical site infections and UTI after hysterectomy. Lake et al in 2013 reported several independent variables associated with UTI after hysterectomy. Those predictors included history of cerebrovascular accident with neurologic deficit, current corticosteroid use, and operative time greater than the 75th percentile.5 A history of cerebrovascular accident with neurologic deficit was only present in three (0.3%) patients in our study. Current corticosteroid use was also rare in our patients reported by 25 (2.2%). Longer duration of surgery was associated with increased UTI rates in our data; however, in the multivariate analysis, it was not an independent risk for UTI. If we applied Lake's independent risk factors in our cohort, this will not accurately predict our observed UTI rates and will not provide the needed adjustment for patient risk. In women who underwent incontinence surgery, Nygaard et al11 reported data from The Stress Incontinence Surgical Treatment Efficiency Trial (SISTEr) and the Trial Of Mid-Urethral Slings (TOMUS). In the first 6 weeks, the following risk factors were associated with postoperative UTI: sling use compared with Burch, recurrent UTI at baseline, POP stage III or IV, and need for clean intermittent self-catheterization at discharge compared with self-voiding. Of those variables, only baseline recurrent UTI was an independent risk factor for postoperative UTI. In the current study, recurrent UTI was associated with postoperative UTI with a threefold increase in the unadjusted analysis. In the final model, recurrent UTI was not an independent variable associated with UTIs. This might be the result of the difference in the study population between the two trials including only women with UI and a higher preoperative recurrent UTI compared with our study, in which UI was a presenting symptom in only 25.4% of patients and recurrent UTI was only present in 2% of the population at baseline.
The four risk factors we identified in this study can help stratify patients into higher risk and lower risk groups for developing UTI after hysterectomy. It seems that the highest risk was observed in women who had anterior vaginal wall prolapse, which may be linked to incomplete bladder emptying or to the effect of the procedure used to treat anterior compartment prolapse, which in most cases was anterior colporrhaphy. The use of HT can be a marker of symptomatic lower estrogen levels and has been previously reported as a risk factor for recurrent UTIs.19 The interesting finding of this study was the change in the degree of association between premenopausal status and risk of UTI in the univariate and multivariate analysis. After adjusting for other confounders, specifically HT use, postmenopausal status became a protective factor suggestive of an interaction between these two variables; however, interaction testing was not significant.
Although we were able to identify independent variables associated with UTI after hysterectomy, the best approach to lower the risk of UTI in patients with higher risk is not yet clear. In the 1980s and 1990s, the concept of preoperative prophylactic antibiotic therapy was rigorously studied for primary prevention of sepsis and surgical site infection including UTI and there was an observed significant reduction when preoperative prophylactic antibiotics were compared with placebo.20–23 Extended antibiotic prophylaxis in women with impaired bladder drainage was previously studied in women after prolapse surgery. In 2004, Rogers et al24 reported significant reduction of postoperative UTIs in women who had a suprapubic catheter after POP surgery who were using nitrofurantoin compared with placebo. More recently, nitrofurantoin was not helpful when compared with placebo for short-term use of a transurethral catheter after pelvic reconstructive surgery.2 Another important issue is the choice and duration of bladder drainage method after hysterectomy. Our data show that no catheter is better than another. This matches prior reports suggesting that there is no need for indwelling catheterization in women undergoing hysterectomy with no other combined reconstructive surgery and, even in women with combined reconstructive surgery, the shorter catheter duration, the lower postoperative UTI rates.25,26 Both from our data and previous reports, the highest rate of posthysterectomy UTI is reported in women who underwent pelvic reconstructive procedures combined with hysterectomy and even higher in women who had incomplete bladder emptying and who require extended postoperative bladder drainage.2,9,11,27–31 For those patients, the choice of method for drainage and its duration has been the main question.25,26,29–31
A limitation of our study is the retrospective design, which increases the possibility of selection and measurement biases. To reduce the effects of selection bias, we included all patients who met the criteria in a certain period of time to avoid any sampling bias. This study is also limited by the specific procedures conducted at our institution, which does not include sacrospinous apical suspension.32–34 This may reduce the external validity of the results of the study. Nonetheless, previous data comparing UTI after sacrospinous suspension with uterosacral suspension reported comparable rates. A validity study of the model in a different study population is recommended. Another potential limitation of retrospective analysis is underestimation of the UTI rate as a result of missing follow-up. For this, we planned a sensitivity analysis comparing rates of UTI in the whole study group with those rates after exclusion of patients who did not have a full 8-week follow-up. In addition, in our practice, we schedule a 6-week postoperative check for all patients, a follow-up phone call is done 1 week after surgery, and another one if the patient did not keep the appointment for her 6-week postoperative check. This resulted in having complete data on all patients with at least a phone call by 8 weeks postoperatively and the sensitivity analysis was not needed. One last limitation is excluding women with hysterectomy done for cancer. This was a decision during the design of the surgery and we believe that the rates of UTI would not be significantly different between hysterectomy done for benign gynecologic conditions and those with cancer. Nonetheless, given the difference based on the indication noted in this study, studying UTI after hysterectomy and other gynecologic oncology procedures might be a future focus of research. Despite those limitations, we believe that this study included an adequate sample size to test the hypothesis at hand and provide an accurate representation of UTI rates from a clinical practice at an academic center and provides a model that can aid in preoperative counseling and planning more measures to avoid UTI in patients with expected higher risk of UTI.
In summary, this study reported on UTI rates after hysterectomy for benign conditions and as part of pelvic reconstructive surgery. The results highlight the wide variation in the UTI rate based on the route and type of concomitant procedures. It also highlights the significant increase in UTI rates in women undergoing pelvic reconstructive surgery compared with those having hysterectomy for other benign gynecologic disorders. In this study, premenopausal status, anterior vaginal wall prolapse on pelvic examination, and postvoid residual greater than 150 mL were independent variables associated with UTI after hysterectomy. After external validation in a different population, this model may help counseling patients on their risk for postoperative UTI and provide a tool for risk adjustment of UTI when comparing UTI rates in women undergoing hysterectomy for benign conditions with those who are undergoing hysterectomy as a part of pelvic reconstructive surgery.
1. Maudsley RF, Robertson EM. Common complications of hysterectomy. Can Med Assoc J 1965;92:908–11.
2. Dieter AA, Amundsen CL, Edenfield AL, Kawasaki A, Levin PJ, Visco AG, et al. Oral antibiotics to prevent postoperative urinary tract infection: a randomized controlled trial [published erratum appears in Obstet Gynecol 2014;123:669]. Obstet Gynecol 2014;123:96–103.
3. Andy UU, Harvie HS, Ackenbom MF, Arya LA. Single versus multi-dose antibiotic prophylaxis for pelvic organ prolapse surgery with graft/mesh. Eur J Obstet Gynecol Reprod Biol 2014;181:37–40.
4. Enomoto LM, Hollenbeak CS, Bhayani NH, Dillon PW, Gusani NJ. Measuring surgical quality: a national clinical registry versus administrative claims data. J Gastrointest Surg 2014;18:1416–22.
5. Lake AG, McPencow AM, Dick-Biascoechea MA, Martin DK, Erekson EA. Surgical site infection after hysterectomy. Am J Obstet Gynecol 2013;209:490.e1–9.
6. Zielinski MD, Thomsen KM, Polites SF, Khasawneh MA, Jenkins DH, Habermann EB. Is the Centers for Medicare and Medicaid Service's lack of reimbursement for postoperative urinary tract infections in elderly emergency surgery patients justified? Surgery 2014;156:1009–15.
7. Wald HL, Kramer AM. Nonpayment for harms resulting from medical care: catheter-associated urinary tract infections. JAMA 2007;298:2782–4.
8. Mahdi H, Goodrich S, Lockhart D, DeBernardo R, Moslemi-Kebria M. Predictors of surgical site infection in women undergoing hysterectomy for benign gynecologic disease: a multicenter analysis using the national surgical quality improvement program data. J Minim Invasive Gynecol 2014;21:901–9.
9. Chung CP, Kuehl TJ, Harris SK, McBride MM, Larsen WI, Yandell PM, et al. Incidence and risk factors of postoperative urinary tract infection after uterosacral ligament suspension. Int Urogynecol J 2012;23:947–50.
10. Sutkin G, Alperin M, Meyn L, Wiesenfeld HC, Ellison R, Zyczynski HM. Symptomatic urinary tract infections after surgery for prolapse and/or incontinence. Int Urogynecol J 2010;21:955–61.
11. Nygaard I, Brubaker L, Chai TC, Markland AD, Menefee SA, Sirls L, et al. Risk factors for urinary tract infection following incontinence surgery. Int Urogynecol J 2011;22:1255–65.
12. von Elm E, Altman DG, Egger M, Pocock SJ, Gotzsche PC, Vandenbroucke JP; STROBE Initiative. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Lancet 2007;370:1453–7.
13. Gould CV, Umscheid CA, Agarwal RK, Kuntz G, Pegues DA; Healthcare Infection Control Practices Advisory Committee. Guideline for prevention of catheter-associated urinary tract infections 2009. Infect Control Hosp Epidemiol 2010;31:319–26.
14. Horan TC, Andrus M, Dudeck MA. CDC/NHSN surveillance definition of health care-associated infection and criteria for specific types of infections in the acute care setting. Am J Infect Control 2008;36:309–32.
15. Heisler CA, Casiano ER, Gebhart JB. Hysterectomy and perioperative morbidity in women who have undergone renal transplantation. Am J Obstet Gynecol 2010;202:314.e1–4.
16. Anand M, Woelk JL, Weaver AL, Trabuco EC, Klingele CJ, Gebhart JB. Perioperative complications of robotic sacrocolpopexy for post-hysterectomy vaginal vault prolapse. Int Urogynecol J 2014;25:1193–200.
17. Harrell FE Jr, Lee KL, Mark DB. Multivariable prognostic models: issues in developing models, evaluating assumptions and adequacy, and measuring and reducing errors. Stat Med 1996;15:361–87.
18. Clopper C, Pearson E. The use of confidence or fiducial limits illustrated in the case of the binomial. Biometrika 1934;26:404–13.
19. Dwyer PL, O'Reilly M. Recurrent urinary tract infection in the female. Curr Opin Obstet Gynecol 2002;14:537–43.
20. Kocak I, Ustün C, Gürkan N. Prophylactic antibiotics in elective abdominal hysterectomy. Int J Gynaecol Obstet 2005;90:157–8.
21. Tanos V, Rojansky N. Prophylactic antibiotics in abdominal hysterectomy. J Am Coll Surg 1994;179:593–600.
22. Tchabo JG, Cutting ME, Butler C. Prophylactic antibiotics in patients undergoing total vaginal or abdominal hysterectomy. Int Surg 1985;70:349–52.
23. Hamod KA, Spence MR, King TM. Prophylactic antibiotics in vaginal hysterectomy: a review. Obstet Gynecol Surv 1982;37:207–16.
24. Rogers RG, Kammerer-Doak D, Olsen A, Thompson PK, Walters MD, Lukacz ES, et al. A randomized, double-blind, placebo-controlled comparison of the effect of nitrofurantoin monohydrate macrocrystals on the development of urinary tract infections after surgery for pelvic organ prolapse and/or stress urinary incontinence with suprapubic catheterization. Am J Obstet Gynecol 2004;191:182–7.
25. Dunn TS, Shlay J, Forshner D. Are in-dwelling catheters necessary for 24 hours after hysterectomy? Am J Obstet Gynecol 2003;189:435–7.
26. Ahmed MR, Sayed Ahmed WA, Atwa KA, Metwally L. Timing of urinary catheter removal after uncomplicated total abdominal hysterectomy: a prospective randomized trial. Eur J Obstet Gynecol Reprod Biol 2014;176:60–3.
27. Bai SW, Jung HJ, Jeon MJ, Chung DJ, Kim SK, Kim JW. Surgical repair of anterior wall vaginal defects. Int J Gynaecol Obstet 2007;98:147–50.
28. Liang CC, Lee CL, Chang TC, Chang YL, Wang CJ, Soong YK. Postoperative urinary outcomes in catheterized and non-catheterized patients undergoing laparoscopic-assisted vaginal hysterectomy—a randomized controlled trial. Int Urogynecol J Pelvic Floor Dysfunct 2009;20:295–300.
29. Dixon L, Dolan LM, Brown K, Hilton P. RCT of urethral versus suprapubic catheterization. Br J Nurs 2010;19:S7–13.
30. Hofmeister FJ, Martens WE, Strebel RL. Foley catheter or suprapubic tube? Am J Obstet Gynecol 1970;107:767–79.
31. Dunn TS, Figge J, Wolf D. A comparison of outcomes of transurethral versus suprapubic catheterization after Burch cystourethropexy. Int Urogynecol J Pelvic Floor Dysfunct 2005;16:60–2.
32. Unger CA, Barber MD, Walters MD, Paraiso MFR, Ridgeway B, Jelovsek JE. Long-term effectiveness of uterosacral colpopexy and minimally invasive sacral colpopexy for treatment of pelvic organ prolapse. Female Pelvic Med Reconstr Surg 2017;23:188–94.
33. Barber MD. Pelvic organ prolapse. BMJ 2016;354:i3853.
34. Hill AJ, Barber MD. Apical prolapse repair: weighing the risks and benefits. Curr Opin Obstet Gynecol 2015;27:373–9.