Influence of prophylactic antibiotics on incidence of urinary tract infections in acute-to-subacute patients with stroke and asymptomatic bacteriuria

Abstract

Urinary tract infections (UTIs) are a major complication in patients with acute-to-subacute stroke. Asymptomatic bacteriuria is prevalent in this population, and the role of prophylactic antibiotics remains unclear. This study evaluated whether prophylactic antibiotics reduces the incidence of symptomatic UTI in patients with acute-to-subacute stroke with asymptomatic bacteriuria. This retrospective cohort study analyzed 111 patients with acute-to-subacute stroke and asymptomatic bacteriuria at a Taiwanese medical center. Participants were stratified into intervention (n = 38, receiving oral prophylactic antibiotics) and control (n = 73, no antibiotics) groups. The study compared the incidence of urinary tract infections between two groups. A subgroup analysis of subacute patients with stroke was performed to confirm the antibiotics’ protective effect. UTI incidence in the intervention and control groups were 2.7% and 32.73%, respectively (p = 0.003). Multivariate logistic regression revealed a 93% reduced risk of symptomatic UTI in the intervention versus the control group (p = 0.025). Subgroup analysis demonstrated the persistent protective effect of oral prophylactic antibiotics (p = 0.036). Prophylactic antibiotic treatment effectively reduced symptomatic UTI incidence among patients with acute-to-subacute stroke with asymptomatic bacteriuria. Future multicenter randomized controlled trials are warranted to standardize treatment protocols and evaluate long-term outcomes in this vulnerable population.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-025-17932-z.

Introduction

Infection is a highly common complication in the acute-to-subacute phase poststroke, with urinary tract infection (UTI) being particularly prevalent¹. Stroke may lead to urinary voiding dysfunction and urinary retention, potentially promoting bacterial growth in urine². Up to 24% of patients may develop a UTI within the first month poststroke, with an incidence of 22–23% in the subsequent 6–30 months³. UTIs are more common in hospitalized patients and among individuals with severe cerebral stroke⁴leading to increased morbidity, prolonged hospital stays, and higher healthcare costs^1,2. Poststroke UTIs also increase the likelihood of future institutional care and mortality in patients with severe stroke⁵.

Asymptomatic bacteriuria, defined as the presence of bacteria in urine without UTI symptoms, is common in patients with stroke. Stroke may cause neurogenic lower urinary tract dysfunction and increased amount of postvoid residual urine⁶ increasing the risk of asymptomatic bacteriuria. Among older adults with stroke, asymptomatic bacteriuria has a baseline incidence of approximately 20%² and is associated with a high likelihood of symptomatic UTI during follow-up⁷. The severity of stroke, poststroke urinary status, and functional impairment are closely related to bacterial colonization in the urinary system of patients with stroke⁸.

Stroke-induced immunosuppression increases the risk of UTI in these patients⁹. Additionally, neurogenic bladder dysfunction and frequent use of indwelling catheters further increase the risk of UTI in this population^10,11. Poststroke asymptomatic bacteriuria may be associated with an increased risk of urinary system infections². Whether antibiotics for asymptomatic bacteriuria in patients with stroke reduces the likelihood of subsequent UTI is unclear. A 2009 systematic review suggested that prophylactic antibiotics reduce overall infection rates in patients with stroke¹¹. A Cochrane review in 2018 also revealed that prophylactic antibiotics significantly reduce the risk of UTIs by up to 2.5 times in patients with stroke, regardless of bacteriuria status¹². Because patients with asymptomatic bacteriuria poststroke are at high risk of subsequent infection, our research team hypothesizes that prophylactic antibiotics can reduce the likelihood of symptomatic UTI in this subpopulation of patients with acute-to-subacute stroke.

Methods

Study design

This retrospective cohort study was conducted at Shuang Ho Hospital, New Taipei City, Taiwan. The study included patients with a diagnosis of acute or subacute stroke (within 3 months of stroke onset) and asymptomatic bacteriuria during their hospitalization.

Participants

Eligible patients were identified using inpatient records in the electronic medical record system of the Departments of Rehabilitation Medicine, Neurology, and Neurosurgery at Shuang Ho Hospital. The records spanned January 1, 2018, to April 30, 2023. Discharge diagnoses had to meet the criteria for hemorrhagic or ischemic stroke in the ICD-10 codes I60 to I69. Patients with at least one urine test showing white blood cells (WBC) > 10/high power field (HPF) during hospitalization were screened out through the electronic medical record system. Given this study’s retrospective design, the risks of selection and misclassification biases were minimized using strict inclusion criteria and standardized definitions for exposures and outcomes.

Patients meeting the following criteria were included in the study:

1. Being aged 18 years or older.

2. Receiving a diagnosis of ischemic stroke or hemorrhagic stroke within 3 months of admission.

3. Having pyuria, defined by > 10 WBC/HPF in unspun urine.

Patients were excluded if they met the following criteria:

1. Having a known history of urinary tract abnormalities.

2. Having symptoms of UTI at admission.

3. Developing other infections requiring antibiotic treatment.

4. Having a hospital follow-up period of less than 7 days (the mean onset duration of poststroke UTI is 7 days)⁴.

5. Having undergone invasive urological procedures during hospitalization.

Ethical approval and informed consent

The study protocol was reviewed and approved by the TMU-Joint Institutional Review Board (approval number: N202401035). Owing to the retrospective design and the use of fully de-identified data, the requirement for written informed consent was formally waived by the TMU-Joint Institutional Review Board. All methods were carried out in accordance with relevant guidelines and regulations, including ethical approval from the institutional review boards.

Study groups

Patients were divided into two groups on the basis of whether they received preventive oral antibiotic treatment for asymptomatic bacteriuria. The choice of antibiotic agent, dosage, and duration of treatment in the intervention group was based on the treating physician’s discretion and institutional guidelines.

Sample size calculation

Given the absence of studies with similar designs, the minimum required sample size was estimated on the basis of effect sizes observed in a 2008 randomized controlled trial (RCT) investigating prophylactic antibiotics for infection prevention in patients surviving a stroke¹³.

The minimum required sample size was 108 patients (36 intervention, 72 control); this was estimated in G*Power software (α = 0.05, power = 0.8, allocation ratio 2:1) for a medium effect size.

Data collection

Data for the following demographic and clinical characteristics were collected from the patients’ electronic medicalrecords.

• Group (intervention or control).

• Age.

• Sex.

• Diabetes mellitus status.

• Duration from onset: Duration from stroke onset to urine sample collected.

• Length of hospital stay during the study period.

• Stroke type (ischemic or hemorrhagic).

• Stroke location: anterior or posterior circulation¹⁴.

• Urine WBC count at inclusion.

• Baseline blood WBC count and neutrophil percentage at inclusion.

• Whether symptomatic UTI occurred during hospitalization.

• Whether the UTI was catheter-related, defined by the occurrence of a UTI either when a catheter was in place or within 72 h from the removal of a catheter¹⁵.

• Type and duration of antibiotics used (for the intervention group).

• Indwelling urinary catheter duration (days).

• When available, bacterial identification and antimicrobial susceptibility testing, according to Clinical and Laboratory Standards Institute (CLSI) guidelines, were recorded¹⁶.

The primary outcome measure was the occurrence of local or systemic symptomatic UTI during hospitalization. The European Association of Urology classification scheme classifies symptomatic UTI as local or systemic and defines these classifications as follows:¹⁷.

1. Local UTI: defined by dysuria, urgency, frequency, incontinence, urethral discharge, and suprapubic pressure or spasm.

2. Systemic UTI: defined by fever or hypothermia, chills, delirium, hypotension, tachycardia, and costovertebral angle pain/tenderness.

Statistical analysis

Continuous variables are summarized in terms of the mean and standard deviation or the median and interquartile range, as appropriate. Categorical variables are summarized in terms of the frequency and percentage.

Primary analysis

The effectiveness of preventive antibiotic treatment was tested by comparing the incidence of symptomatic UTI between the intervention and control groups using a chi-square test.

Logistic regression analysis

A multivariate logistic regression analysis was performed to adjust for potential confounders with significant intergroup differences at baseline. Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated to determine the strength and significance of the associations.

Kaplan–Meier survival analysis

A Kaplan–Meier survival analysis was further conducted to evaluate the time to the first occurrence of symptomatic UTI. The variation of UTI free probability for the intervention and control groups over time was compared using a log-rank test.

Subgroup analysis

To determine whether heterogeneity between the groups affected the results, we conducted a subgroup analysis by subacute stroke status, defined by a period of more than 7 days from the onset of stroke¹⁸. Chi-square and logistic regression analyses were also performed.

Statistical software

All statistical analyses were performed using SPSS software (version 18, IBM, Armonk, NY, USA). A two-sided p value of < 0.05 indicated statistical significance.

Additional considerations

In patients recovering from stroke, bacteriuria may result from UTI, colonization, or specimen contamination. Pyuria (> 10 WBC/HPF) is a reliable marker indicating infection¹⁹, with high sensitivity and predictive value for positive cultures⁵. Considering the high risk of specimen contamination and bacterial colonization in patients recovering from stroke, this study included patients with asymptomatic bacteriuria only if they had a WBC/HPF test result of > 10. This criterion was set to exclude contamination or colonizationand ensures high sensitivity for bacterial cultures.

Results

We first included 435 patients in an initial sample on the basis of their discharge diagnoses and ICD-10 codes (I60 to I69) and the presence of at least one urine test with > 10 WBC/HPF during hospitalization. After the inclusion and exclusion criteria were applied, 111 patients were included in a final sample; 73 and 38 patients formed the control and intervention groups, respectively. The reasons for exclusion are detailed in Fig. 1.

Fig. 1.

Patient selection and group allocation process.

Demographic characteristics

Compared with the control group, the intervention group had significantly fewer men (37% vs. 58%, p = 0.039), a higher prevalence of diabetes (56% vs. 26%, p = 0.002), longer time from stroke onset to inclusion (33.8 vs. 8.9 days, p < 0.001), and greater baseline pyuria severity (5.97 vs. 3.78 WBC/HPF, p = 0.001). The intervention group also had significantly longer indwelling catheter duration (15.29 ± 9.06 vs. 10.30 ± 9.18 days, p = 0.001). Both groups did not significantly differ in age, stroke type, or other clinical parameters. Baseline demographic and clinical characteristics of the included patients were shown in Table 1.

Table 1.

Baseline demographic and clinical characteristics.

	Intervention	Control	P value
Number of patients	38	73
Age	69.18 (13.5)	63.84 (14.96)	0.07
Sex (male)	14 (37%)	42 (58%)	0.039
DM	21 (56%)	19 (26%)	0.002
Duration from onset (day)	33.84 (29.05)	8.9 (19.59)	0.001
Stroke type (hemorrhagic)	17 (48%)	24 (33%)	0.219
Stroke area (posterior)	16 (42%)	20 (27%)	0.084
Baseline pyuria (/HPF)	5.97 (3.51)	3.78 (3.13)	0.001
Baseline bacteriuria (/uL)	7463 (19,690)	2789 (12,321)	0.132
Baseline WBC count (103/uL)	8.6 (3.08)	9.496 (3.959)	0.615
Baseline neutrophil percentage	70.8 (11.09)	69.6 (11.4)	0.595
Indwelling catheter duration, mean±SD, (day)	15.29 ± 9.06	10.30 ± 9.18	0.001
Antibiotics duration (day)	8.18 (3.4)	0

Data are expressed as mean (standard deviation) or frequency (percentage).

Patients in the intervention group received cephalosporins, fluoroquinolones, or other antibiotics; cephalosporins were the most used (76.3%), followed by fluoroquinolones (28.9%). Most patients (89.4%) received antibiotics for ≥ 7 days (mean duration: 8.2 days). Furthermore, among the 19 patients who developed symptomatic UTI during hospitalization, 18 (94.7%) had catheter-related UTI. The demographic distribution of antibiotic classes used in the intervention group is summarized in Appendix 1.

Among the 38 patients who received prophylactic antibiotics, urine cultures were available for 30 (78.9%). The overall susceptibility rate was 80%. E. coli was the predominant pathogen (50%) with complete susceptibility to prophylactic antibiotics. Multi-drug resistance (MDR) organisms, defined as acquired non-susceptibility to at least one agent in three or more antimicrobial categories²⁰ were detected in 10% of cultures and showed complete resistance. (Appendix 3.)

Chi-square tests

The intervention group had a lower prevalence of symptomatic UTIs (1 out of 38) than did the control group (18 out of 73; Table 2). Preventive antibiotic treatment was associated with a lower likelihood of symptomatic UTI (OR = 12.12, p = 0.003) in a Pearson chi-square test.

Table 2.

Relationship between prophylactic antibiotic treatment and symptomatic UTI Incidence.

Chi square	UTI	No UTI	UTI rate
Primary analysis
Control	18	55	32.73%
Intervention	1	37	2.7%
	Pearson’s Chi square p: 0.003
Subgroup analysis (> 7 days poststroke)
Control	4	11	26.67%
Intervention	1	30	3.2%
	Pearson’s Chi square p: 0.033

Logistic regression analysis

The multivariate logistic regression analysis had adjustments for sex, diabetes mellitus status, time from stroke onset to pyuria detection, baseline pyuria levels and indwelling catheter duration; the findings revealed a significant protective effect of prophylactic antibiotics against symptomatic UTI (OR = 14.033, 95% CI = 1.391–141.575, p = 0.025; Table 3). This indicated a 93% reduction in UTI risk from prophylactic antibiotics. Indwelling catheter was also an independent risk factor for symptomatic UTI(P = 0.002).

Table 3.

Multivariate logistic regression analysis of factors associated with symptomatic UTI Incidence.

Logistic regression	Odds ratio	95% CI	P value
Primary analysis
Antibiotics vs. Control	14.033	1.391-141.575	0.025
Sex	1.612	0.489–5.307	0.433
DM	2.081	0.496–8.732	0.316
Duration to onset	0.967	0.927–1.008	0.116
Baseline pyuria	0.998	0.825–1.245	0.987
Indwelling catheter duration	1.145	1.052–1.245	0.002
Subgroups analysis (> 7 Days Poststroke)
Antibiotics vs. Control	45.286	1.288–1592.1	0.036
Sex	10.646	0.387–292.9	0.162
DM	0.114	0.006–2.227	0.152
Duration to onset	1.025	0.977–1.075	0.317
Baseline pyuria	1.001	0.715-1.400	0.996
Indwelling catheter duration	1.036	0.948–1.133	0.437

Kaplan–Meier analysis

As presented in Fig. 2, the Kaplan–Meier survival curves indicated that the intervention group had a significantly higher UTI-free probability (i.e., lower incidence of symptomatic UTI) compared with the control group throughout the study period. The log-rank test revealed that the UTI-free probability significantly differed between the groups (p = 0.002).

Fig. 2.

Kaplan–Meier survival curves for UTI-free probability in intervention versus control groups.

Subgroup analysis

Among patients for whom more than 7 days had elapsed after a stroke, 31 and 15 belonged to an intervention and control group, respectively. These two groups of patients did not significantly differ in any baseline or demographic characteristics (Appendix 2).

UTI incidence was significantly lower in the intervention group (3.2% vs. 26.7%, p = 0.033), and an adjusted analysis indicated a 45.3-fold increased risk of UTI in the control group (OR = 45.3, p = 0.036).

Discussion

This is the first study to evaluate prophylactic antibiotics in patients with acute-to-subacute stroke and asymptomatic bacteriuria. Our findings demonstrate that prophylactic antibiotics reduced UTI risk by 93% (OR = 14.033, p = 0.025) and delayed UTI onset (log-rank p = 0.002). These results were corroborated in a subgroup analysis by subacute stroke status.

Traditionally, the management of asymptomatic bacteriuria has been largely restricted to pregnant women and patients undergoing urological procedures^21,22. The use of antibiotic prophylaxis in patients with asymptomatic bacteriuria who are recovering from stroke remain controversial and nonroutine in clinical practice. Our findings align with a 2009 systematic review and 2018 Cochrane review demonstrating that prophylactic antibiotics reduce overall infection rates in patients with stroke^11,12. By focusing on a high-risk subgroup—those with asymptomatic bacteriuria—our study provides critical evidence that targeted antibiotic use in this specific population yields substantial clinical benefits while ensuring responsible antibiotic use.

During the acute-to-subacute phase poststroke, patients often develop neurogenic bladder dysfunction, characterized by impaired detrusor contraction and urinary retention^23,24, which promotes bacterial proliferation in the urinary tract^3,25. Bacteria in urine activate the immune system, similar to the mechanism underlying Bacillus Calmette–Guérin (BCG) therapy for bladder cancer²⁶. However, a study observed that in patients with persistent bacteriuria, significant B cell infiltration was present in the submucosa, including lymphoid nodules²⁷, and caused chronic damage to the bladder wall because of persistent immune responses triggered by these pathogens. Prophylactic antibiotics administered during the acute-to-subacute poststroke period may reduce early bladder submucosal inflammation caused by asymptomatic bacteriuria, thereby preventing long-term and irreversible bladder damage.

Furthermore, stroke-induced immunosuppression and frequent indwelling urinary catheterization elevates UTI risk^28,29. These risk factors typically resolve as stroke recovery progresses⁶. Prophylactic antibiotic use during this high-risk transitional period may reduce overall UTI incidence by mitigating vulnerability to pathogens.

The 80% overall susceptibility rate provide microbiological evidence for the observed prophylactic efficacy, with no symptomatic UTIs developing among patients harboring susceptible isolates. However, the detection of MDR organisms in 10% of cultures²⁰ emphasizes the continued importance of individual bacterial culture and susceptibility testing when implementing prophylactic protocols.

According to our study, a treatment course of cephalosporins and/or fluoroquinolones over 8 days could prevent symptomatic UTI in most survivors of stroke with asymptomatic bacteriuria. Most cases of symptomatic UTIs in this study were of catheter-related UTIs (94.7%). Indwelling urinary catheterization represents an independent risk factor for UTI development in the acute post-stroke period³⁰. In conjunction with our main findings, these results underscore the critical importance of implementing early catheter removal protocols in acute post-stroke care while advocating for the judicious use of prophylactic antibiotics in patients who require prolonged catheterization.

This study may have underestimated the effect size due to its retrospective design and relatively small and imbalanced sample³¹. The need to control for multiple covariates in the logistic regression analysis may have led to a further reduction in the observed effect size. Crucially, the intervention group had characteristics that could potentially bias it toward a higher risk of symptomatic UTI, including a higher mean age, greater prevalence of diabetes mellitus, higher baseline pyuria levels and longer indwelling catheter duration^32,33.

This study’s limitations include its retrospective design, single-center setting, modest sample size (n = 111), and imbalance between groups at baseline; these give rise to potential selection bias³⁴, preclude causal inference, and limit generalizability. Our retrospective design also precluded serial bacterial culture for resistance emergence or bacterial species succession. Furthermore, an optimal regimen could not be determined due to the absence of standardized antibiotic protocols. RCTs with larger sample sizes and long-term follow-up should be conducted; these RCTs should focus on functional outcomes and resistant organisms.

Conclusion

Prophylactic antibiotic treatment significantly reduces symptomatic UTI incidence in patients with acute-to-subacute stroke and asymptomatic bacteriuria. This finding was corroborated in a subgroup analysis by subacute stroke status. Further RCTs were needed to optimize treatment protocols and address the potential risks of antibiotic resistance and other adverse effects.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contributions

H.-C.C. conceived and designed the study. F.-S.Y. acquired the data, performed the statistical analysis, and contributed to data analysis and interpretation. F.-S.Y. also drafted the manuscript. C.-W.W. critically revised the manuscript for important intellectual content. K.-C.C. and M.-H.H. provided administrative, technical, or material support. Y.-J.L. supervised the project. All authors reviewed and approved the final manuscript.

Funding

This study received no external funding.

Data availability

The data generated and/or analyzed during the current study are not publicly available. However, they are available from the corresponding author upon reasonable request.

Declarations

Competing interests

The authors declare no competing interests.