Abstract
Introduction:
Urinary tract infection (UTI) is an infrequent complication of diagnostic cystoscopy, with an incidence below 5%. Current guidelines are unclear and remain silent regarding the necessity of performing a urine bacterial culture (UBC) before cystoscopy. The objective of this study was to evaluate whether asymptomatic bacteriuria before cystoscopy increases the risk of UTI afterward. In addition, we aimed to identify risk factors in patient characteristics.
Patients and Methods:
We included all patients who underwent diagnostic cystoscopy at our center between September 2022 and May 2023 and performed a UBC within the preceding 15 days. Patients were excluded if they had received antibiotic treatment or did not provide UBC. Data on the UTI within four weeks post-cystoscopy were collected via telephone interviews. UTIs were defined on the clinical criteria established by the U.S. Centers for Disease Control and Prevention.
Results:
Six hundred eighteen cystoscopies were analyzed. Twenty-nine cases of UTI were reported, accounting for 4.7% of the procedures. Among these, 24 patients had sterile or polymicrobial UBC, whereas 5 had asymptomatic bacteriuria prior to cystoscopy. There was no significant association between asymptomatic bacteriuria and the occurrence of UTI (OR: 1.385, 95% confidence interval [0.513; 3.739], p = 0.52). In multivariable analysis, age was a protective factor (p = 0.01). There was an increase of UTI when the cystoscopy was performed for a suspected tumor (OR: 4.981, 95% confidence interval [1.122; 22.108], p = 0.035) or for a retention (OR: 7.910, 95% confidence interval [1.329; 47.093], p = 0.023). More UTIs occurred when cystoscopy was performed for suspected tumors (p = 0.035) or urinary retention (p = 0.023). There was a significant correlation between asymptomatic bacteriuria and consultation for urinary symptoms after (p = 0.043).
Conclusion:
We found no correlation between asymptomatic bacteriuria prior to cystoscopy and the subsequent occurrence of UTI. Routine UBC before cystoscopy appears unnecessary. Eliminating this practice could prevent the rescheduling of cystoscopies, which often leads to diagnostic delays. Moreover, it would contribute to reducing antibiotic consumption and combating antibiotic resistance.
Every surgical procedure carries a risk of infection. Many procedures are subject to clear recommendations aimed at minimizing this risk. In urology, performing a urine bacterial culture (UBC) is often recommended for procedures such as ureteroscopy or transurethral resection of the prostate. 1 Post-endoscopy infections are frequently caused by pathogens already present in the urinary tract, making it logical to detect and treat them beforehand.
Cystoscopy presents a low risk of urinary tract infection (UTI), with reported infection rates generally below 5%. 2 This low rate may be attributed to the fact that cystoscopy is a diagnostic, minimally invasive, and short-duration procedure. Nonetheless, existing studies show considerable variability regarding cystoscopy indications, performance of a UBC prior to the procedure, thresholds for asymptomatic bacteriuria, and antibiotic prophylaxis. Diagnostic criteria for post-cystoscopy infections also vary widely, with some being influenced by transient bladder instability following the procedure.
Numerous studies examined the benefits of antibiotic prophylaxis for cystoscopy:3–5 they generally found that prophylactic antibiotics provide moderate protection against post-cystoscopy infections. However, they mostly compared different prophylaxis protocols, sometimes including placebo groups, without accounting for asymptomatic bacteriuria or antibiogram results.
Fewer studies have explored the benefits of UBC. Herr investigated the occurrence of UTIs in a cohort of patients undergoing cystoscopy for urothelial bladder tumor surveillance: 2 UBC was systematically performed prior to the procedure, and no patient received antibiotic prophylaxis, regardless of UBC results. Patients were divided into two groups: positive UBC (single-species culture >104 CFU/mL) and sterile UBC (no bacteria, polymicrobial, or single-species culture <104 CFU/mL). Among 3,108 patients included, the positive and sterile UBC groups consisted of 673 (22%) and 2,435 (78%) patients, respectively. Within one month, 59 patients (1.9%) developed febrile UTIs, corresponding to 3.7% in the positive UBC group and 1.4% in the sterile UBC group (p = 0.01). All cases were managed on an outpatient basis, with symptoms resolving within 48 h.
Currently, there are no formal recommendations for preventing infection risk during cystoscopies, despite them being one of the most common endo-urological procedures. The necessity of systematic UBC prior to cystoscopy remains debated, with practices varying across institutions. In France, the 2015 guidelines on the management and prevention of health-care-associated UTIs, developed in collaboration with the French Urological Association (AFU), state that “screening and treating or not screening for bacteriuria before diagnostic cystoscopy is optional.” 6 In contrast, the European Association of Urology (EAU) guidelines consider asymptomatic bacteriuria a risk factor for infections in procedures involving mucosal breaches, recommending antibiotic prophylaxis in such cases. 7 However, these guidelines mainly apply to surgical interventions and do not explicitly address cystoscopy, where mucosal breaches are uncommon. The EAU strongly discourages antibiotic prophylaxis for urethrocystoscopy (flexible or rigid) in well-resourced countries due to the low absolute risk of infection, the high number of procedures, and the potential contribution to antimicrobial resistance. Nevertheless, this recommendation does not clarify whether a UBC should be performed before or consider its impact. In North America, the Infectious Disease Society of America published guidelines about the need of UBC before endo-urological procedures. Although they consider diagnostic cystoscopy as low risk for infectious complications, they suggest that a urine culture be obtained prior to the procedure and antimicrobial therapy prescribed for patients who will undergo endoscopic urologic procedures, but without distinguishing each type of procedure. 8
The objective of this observational study was to characterize patients who developed UTIs following diagnostic cystoscopy and to determine whether pre-procedure asymptomatic bacteriuria represents a risk factor for UTIs. We hypothesized that there was no correlation between a pre-cystoscopy asymptomatic bacteriuria and the occurrence of a post-procedure UTI.
Patients and Methods
Population
We included all patients who underwent diagnostic cystoscopy at our center between September 2022 and May 2023 and had a UBC performed in the 15 days before the procedure. Patients who received antibiotics before the procedure or had a ureteral stent removed during cystoscopy were excluded. In line with our center’s protocol, rigid cystoscopes were used for women, whereas flexible fiberscopes were used for men.
This study was approved by the Ethics Committee of the AFU (n° CERU_2023-027).
Data collection
Data were collected from consultation records and telephone interviews. Patient characteristics included established risk factors for UTI as described in the literature: gender, age, diabetes, immunosuppression, renal failure (CKD <30 mL/min), post-void residual volume ≥100 mL, presence of indwelling or suprapubic catheters, and anatomical abnormalities of the urinary tract. Additional information was gathered on self-intermittent catheterization, recurrent UTIs (≥4 infections per year), and continuous antibiotic prophylaxis.
Immunosuppression was defined as severe organ insufficiency, primary immune deficiency, hematological malignant disease, autoimmune diseases, or immunosuppressive therapy. Anatomical abnormalities included congenital or acquired urinary tract defects and kidney transplants, excluding renal and ureteral malformations or surgeries.
UBC results were categorized into positive group (single bacterial species ≥10³ CFU/mL) and negative group (sterile results, polymicrobial cultures, or single species <10³ CFU/mL). Data on bacterial species, concentrations, and leukocyturia were also collected.
Cystoscopy-related data included indication, diagnosis, and presence of additional procedure. A “suspected tumor” diagnosis was defined as the identification of a reddish area leading to a change in planned management, such as a bladder biopsy or a new cystoscopy.
Endpoints
Primary endpoint: Occurrence of UTI within four weeks, defined according to the U.S. Centers for Disease Control and Pevention 9 criteria: fever >38°c, suprapubic tenderness, urinary frequency, urinary urgency, dysuria. Patients needed to exhibit at least one of these symptoms starting 24 hours post-procedure, alongside a positive repeat UBC or antibiotic therapy.
Note: Costovertebral angle tenderness, a marker of upper UTI, was excluded from the analysis.
Secondary endpoints were, within the same timeframe, prescription of antibiotics for urinary symptoms, need for physician consultation due to urinary symptoms, and hospitalization for urinary infections.
Statistical analysis
Endpoints were analyzed concerning UBC results and patient characteristics. We used the softwares p value.io and SPSS (v29.0.1.0). We used a statistical significance value p < 0.05. We used Chi-squared and Fisher exact tests, exact tests for qualitative variables, and Welch test for quantitative variables.
Logistic regression identified risk factors for post-cystoscopy UTIs. For categorical variables, the reference group for cystoscopy indication was “urothelial carcinoma monitoring” due to its prevalence. The reference group for diagnosis was “normal” cystoscopy results. Variables with p < 0.2 in univariate analysis were included in the multivariable analysis.
Results
Data collection
We performed 1,136 cystoscopies in our center between September 2022 and May 2023. After excluding ureteral stent removals (237 cases), cystoscopies performed without prior UBC (135 cases), and those performed with antibiotic prophylaxis adapted to the UBC (49 cases), 715 cystoscopies remained within our inclusion criteria. Thirteen patients died between the cystoscopy and the data collection and could not be included (causes of death were not related to the cystoscopies). Five patients refused to answer the telephone interview questions, and 79 could not be contacted. Finally, 618 cystoscopies were included (Fig. 1).
Data collection flowchart.
Characteristics of the population
The mean age was 68 years. Of the 618 cystoscopies, 82 patients (13%) had asymptomatic bacteriuria, and 536 (87%) had a negative UBC. The characteristics of the patients are summarized in Table 1. Most of the patients included were men (78%). The presence of an anatomical abnormality was significantly different between the groups (p < 0.01): 15% (78 cases) in the negative UBC group and 28% (23 cases) in the asymptomatic bacteriuria group. There were no other significant differences in population characteristics. Escherichia coli was the bacterial species most frequently found (35%).
Characteristics of Patients, UBC, and Cystoscopy
UBC = urine bacterial culture; UTI = urinary tract infection.
Characteristics of cystoscopy
The most common indication was the monitoring of urothelial carcinoma (54%), followed by investigation of lower urinary tract symptoms (18%) and hematuria (11%). Cystoscopy was normal in 438 (71%) of patients. Few patients underwent additional procedures: urethral dilatation for 21 patients (3.4%), bladder biopsy for 12 patients (1.9%), 1 stone extraction, and 1 suprapubic catheter placement.
Primary endpoint
Twenty-nine patients (4.7%) out of the 618 cystoscopies had a UTI within four weeks of the procedure. Among them, 24 patients (83%) had a sterile UBC, and 5 patients (17%) had asymptomatic bacteriuria prior to cystoscopy. There was no significant difference in the rate of infection between the different UBC groups. The bacteria species of the five patients who presented an infection with an asymptomatic bacteriuria prior to the cystoscopy were all different from one another: Citrobacter freundii, Enterococcus faecalis, Klebsiella pneumoniae, Morganella morganii, and Staphylococcus aureus.
The mean age was significantly younger in the infected patients (58.7 yrs) than in the uninfected patients (69.4 yrs) (p < 0.01). There was also a significant difference between the proportion of women in infected patients (12 patients or 41%) and non-infected patients (121 patients or 21%) (p = 0.01). Of the patients who underwent an additional procedure, two developed an infection: 1 patient who had a bladder biopsy and 1 patient who had a stone removal.
We performed a multivariable analysis with the variables age, gender, presence of kidney transplant, enterocystoplasty, recurrent UTI, and cystoscopy indication according to their univariate significance levels p < 0.2 (Table 2). We did not include the leukocyturia rate because it was considered dependent on the result of the UBC. We did not include continuous antibiotic prophylaxis either, as it was too closely correlated with recurrent UTI. Hosmer-Lemeshow test value was 0.772 for the multivariable analysis, and c-statistic AUC was 0.780; we considered this threshold acceptable.
Logistic Regression in Univariate and Multivariable Analysis Between Patient Characteristics and Risk of Post-Cystoscopy UTI. Characteristics for Which no Infection Was Observed Are Not Represented
UTI = urinary tract infection.
The risk of post-cystoscopy UTI decreased significantly with age (p = 0.01), but only moderately (OR: 0.971, 95% confidence interval [0.949; 0.993]). There was an increase of UTI when the cystoscopy was performed for a suspected tumor (OR: 4.981, 95% confidence interval [1.122; 22.108]) or for a retention (OR: 7.910, 95% confidence interval [1.329; 47.093]); their increase was significant (p = 0.035 and p = 0.023, respectively). No other characteristic had a substantial impact.
An analysis of the asymptomatic bacteriuria group was carried out for the variables “leukocyturia” and “power of the CFU/mL.” There was no significant difference for these two variables in the occurrence of UTI.
Secondary endpoints
There were significantly more consultations for urinary symptoms in the asymptomatic bacteriuria group (p = 0.043). In the consultations, 6 had asymptomatic bacteriuria, and 15 had a sterile UBC. Six patients (28.6%) consulted without a diagnosis of UTI because they did not have a positive UBC or had not received antibiotic therapy: they consulted for hematuria, acute urine retention, dysuria, or difficulty with self-catheterization.
The three patients who were admitted to hospital with infectious urinary problems had a sterile UBC prior to cystoscopy. Two patients with asymptomatic bacteriuria before cystoscopy received antibiotic therapy without being diagnosed with a UTI: one during a hemodialysis session and one for isolated asthenia.
Discussion
The incidence of UTIs following diagnostic cystoscopy in our study was 4.7%, which is consistent with the literature. No significant association was observed between the UBC before cystoscopy and subsequent UTI occurrence, but a significant association was found with consultation for urinary symptoms (p = 0.043).
In the multivariable analysis, age appeared to be a protective factor for the UTI after cystoscopy. The effect is 3% per year (OR: 0.971, [0.949; 0.993]). This is a difference between our study and Herr’s study: 2 he showed age as a risk factor (p < 0.05), but with low impact (OR: 1.1), and the IC including 1 (95% confidence interval [0.99; 1.2]). Our results might be due to sampling fluctuation.
We also found an association of occurrence of UTI with the indication of the cystoscopy: more events occurred when it was performed for a suspected tumor (p = 0.035) or for a retention (p = 0.023). But the confidence intervals are large (OR: 4.981 CI: [1.122; 22.108] and OR: 7.910 CI: [1.329; 47.093], respectively) and the populations were low (n = 23 and n = 12, respectively). To our knowledge, this association has not been found in literature. We recall that Herr’s study only concerned cystoscopy for monitoring for urothelial bladder tumor surveillance.
Among the patients for which an UTI occurred, five (17%) had recurrent UTI. Their infection in the four weeks after cystoscopy might be caused by their underlying risk factor of recurrent UTI. However, the multivariable analysis did not find an association. (OR: 2.527, CI: [0.529; 12.059], p = 0.245). We did not find any study about the relation between recurrent UTI and post-procedural UTI; these patients represent a special case, as the physicians would tend to prescribe an antibiotic prophylaxis more easily; they would deserve a dedicated study.
No significant association between an additional procedure and the occurrence of a UTI was found. But the low proportion (35 patients) and the different procedures performed make these data difficult to interpret and warrant further investigations. Also, we did not include cystoscopies for ureteral stent removal, as they represent a measurement bias: they are covered by a bacterial biofilm, they induce irritative urinary symptoms, and many upper UTIs occur when they are in place or when they are removed. The value of UBC prior to cystoscopy for their removal is currently the subject of a multicenter randomized trial (AblaJ, RCT 2020-A02518-31) involving our center.
Our study hase limitations. We conducted a retrospective study: data were collected between May 2023 and July 2023 for procedures that occurred between September 2022 and May 2023. This necessarily implies a recall bias, and a short duration of data collection of nine months was chosen to minimize this bias. We also conducted a single-center study, which could introduce selection bias: different asepsis protocols or using fiberscopes also for women in other centers.
UTI remains a rare event, and a retrospective study of this kind would not allow any consensual conclusions about their management. A recent study 10 prospectively involved patients before cystoscopy in two cohorts in a 2:1 ratio: cohort A (n = 316) with patients who did not provide a UBC before cystoscopy and cohort B (n = 145) with patients who were required to provide a sterile UBC. Patients in cohort B with asymptomatic bacteriuria received an antibiotic therapy according to the antibiograms and the recommendations of expert microbiologists. UTI rates were 2.5% (8 patients) in cohort A and 4.8% (7 patients) in cohort B, with no significant difference. The authors therefore concluded that there was no benefit in performing a UBC prior to cystoscopy.
Most current guidelines use the study conducted by Herr, 2 in which there was a significant difference in the occurrence of UTI whether patients had asymptomatic bacteriuria or sterile UBC (p < 0.01). This study only concerned febrile UTI and therefore excluded uncomplicated cystitis. In our study, we used the U.S. Centers for Disease Control and Prevention criteria, in which fever is optional. Of the 29 post-cystoscopy infections in our study, only 13 (44%) were febrile. In the study of Herr, the author found 1.9% febrile infections (39 patients out of the 2010 included). Another difference was that he only included patients being followed for bladder tumors. His conclusion was that antibiotic prophylaxis was unnecessary before cystoscopy due to the low rate of infection and their low severity.
The diagnosis of UTI after cystoscopy is delicate: many patients report a “bladder instability” after the procedure, which may lead to the mistaken diagnosis of an infection and prescription of antibiotic therapy (up to 32% of patients 11 ). We empirically considered that bladder instability appeared in the first hours after cystoscopy, whereas symptoms of UTI appeared at least 24 hours later. Our study confirms that these are common post-procedure symptoms and should not be wrongly attributed to asymptomatic bacteriuria, and that antibiotic therapy is therefore unnecessary.
A prospective study evaluated the occurrence of UTI after cystoscopy: 11 patients were divided into groups according to the UBC: sterile, contaminated, and positive (single bacteria species, concentration >103 CFU/mL). Authors found a significant association between positive UBC and irritative symptoms after cystoscopy (OR: 5.0, 95% confidence interval [1.4; 17.85], p = 0.013), but no significant association was found with febrile UTI. In our study, we did not investigate irritative symptoms, but we found a significant association between asymptomatic bacteriuria and referral for urinary symptoms (p = 0.043). We can argue that both studies confirmed that positive UBC is a risk factor for post-cystoscopy urinary bother.
Three cystoscopies were canceled due to untreated asymptomatic bacteriuria during study period. Two were rescheduled after antibiotherapy, but one patient was lost to follow-up. His cystoscopy was scheduled for hematuria. This could lead to an important delay for a diagnosis of urothelial carcinoma. Eliminating the need for a UBC prior to cystoscopy could have avoided such an event.
Among the 49 positive UBC treated during the period and not included in the analysis, most were treated by the urologist performing the cystoscopy. The others were treated by the referring physician without consulting the specialist. Carrying out any additional examination can be stressful for a patient, especially if they are unable to interpret the result correctly in context: a positive UBC prior to cystoscopy can wrongly worry both the patient and his referring physician and lead to abusive and unnecessary antibiotherapy. We assume that not carrying an UBC would have a positive impact on patient anxiety and the prescription of antibiotherapy. 10 This is important given the high number of cystoscopies performed. Systematic treatment of asymptomatic bacteriuria prior to cystoscopy is responsible for significant selection pressure on bacterial species in the urinary tract and is therefore responsible for the development of antibiotic resistance.
Seven hundred sixty-four cystoscopies were performed in our center in the data collection period with untreated UBC, excluding ureteral stent removals. The cost of a UBC in France ranges from €15 to €20, suggesting potential savings of €11,460 to €15,280. Extrapolating these data to national and international levels, given the high volume of cystoscopies performed each year, we can easily see the economic impact of not performing these UBC.
According to the EAU, urinary tract colonization should not be treated if asymptomatic, except for pregnant women and prior to endo-urological procedures. 7 They also recommend not using antibiotic prophylaxis before cystoscopy, given the low rate of UTI, their low severity, and the risk of antibiotic resistance. These guidelines are on the basis of several meta-analyses, which suggest a moderate benefit from antibioprophylaxis.4,5,12 But asymptomatic bacteriuria was often a non-inclusion criterion in these studies.
On the basis of the current literature and our findings, we believe that there is no need to perform a systematic UBC prior to diagnostic cystoscopy. Of course, in the presence of symptoms suggestive of infection, cystoscopy should be deferred and an UBC should be performed to guide appropriate antibiotic therapy.
Conclusion
In our study, a low rate of UTI was observed after cystoscopy. There did not appear to be any correlation between asymptomatic bacteriuria before cystoscopy and the occurrence of post-cystoscopy UTI. The only significant association for UBC was with referral to a physician for urinary tract symptoms. It would therefore be pointless to carry out a UBC before the procedure to prevent a potential infection. An end to systematic screening would make it possible to avoid delays in diagnosis, anxiety for the patient, and the prescription of unnecessary antibiotic therapy, which contributes to antibiotic resistance.
Further large-scale, randomized studies would be needed to determine a management of the UBC before cystoscopy. This would clarify the current recommendations on the role of this examination.
Authors’ Contributions
A.F.: Conceptualization, investigation, formal analysis, and writing. A.B.: Formal analysis, review, and editing. F.B.: Conceptualization, validation, review and editing.
Footnotes
Acknowledgment
Authors have no acknowledgment to report.
Author Disclosure Statement
The authors declare that they have no competing interest.
Funding Information
This research received no external funding.