From Awareness to Action: Women’s Self-Care Strategies and Clinical Behaviors in Recurrent Urinary Tract Infections

Round 1

Reviewer 1 Report

Comments and Suggestions for Authors

This paper reports a single-centre cross-sectional survey of 36 Polish women with recurrent UTI, describing gaps between patient knowledge/self-care and guideline-concordant clinical management, particularly low urine-culture use, high fluoroquinolone exposure, and underuse of non-antibiotic prevention such as vaginal estrogen and methenamine hippurate. 

1. Given the very small sample (N=36) and recruitment from a single specialist urology clinic, how do the authors justify the representativeness of the findings for broader rUTI populations?
2. Since diagnostics, antibiotic exposure, triggers, episode counts, and satisfaction are self-reported, what steps were taken to reduce recall bias (e.g., cross-checking with medical records), and how might recall bias affect estimates like ciprofloxacin exposure (55.6%) or culture use (38.9%)?
3. The study uses age ≥50 as a proxy for postmenopause, yet vaginal estrogen underuse is a central conclusion. Why was menopausal status not recorded directly, and could misclassification distort the interpretation of estrogen underutilisation?
4. The paper concludes preventive care was misaligned with guidelines, but did it account for patient-level contraindications/preferences, prior failure of therapies, or physician reasoning (e.g., limited access to therapies, local prescribing constraints such as furazidin availability)? How can “gap” be separated from appropriate individualised decision-making?

Author Response

Comment 1

Given the very small sample (N = 36) and recruitment from a single specialist urology clinic, how do the authors justify the representativeness of the findings for broader rUTI populations?

Response
We agree that the small sample size and single-centre design limit external validity and statistical generalisability. We therefore revised the manuscript to frame our findings as hypothesis-generating and most applicable to a clinically relevant subgroup of women with rUTI who experience persistent symptoms and repeated healthcare encounters prior to specialist referral. 

Changes made in the manuscript
We strengthened the limitations statement in the Discussion, adding that the study is single-centre, cross-sectional, and based on a small specialist-referred cohort, and that these factors limit generalisability while supporting the role of the results for hypothesis generation and targeted quality-improvement initiatives.

Comment 2

Since diagnostics, antibiotic exposure, triggers, episode counts, and satisfaction are self-reported, what steps were taken to reduce recall bias (e.g., cross-checking with medical records), and how might recall bias affect estimates like ciprofloxacin exposure (55.6%) or culture use (38.9%)?

Response
All data were collected by patient self-report at the time of the visit. Because many participants were newly referred and had received prior care across multiple external settings, medical records were not consistently available; therefore, we did not perform systematic verification against clinical documentation. 

Changes made in the manuscript
We added a clear statement in the Methods (“no systematic verification against clinical documentation was performed”) and expanded the limitations in the Discussion to address potential recall bias and its implications for self-reported proportions.

Comment 3

The study uses age ≥50 as a proxy for postmenopause, yet vaginal estrogen underuse is a central conclusion. Why was menopausal status not recorded directly, and could misclassification distort the interpretation of estrogen underutilisation?

Response
We acknowledge that menopausal status was not recorded directly. In the questionnaire-based design, we used age ≥50 years as a pragmatic proxy for postmenopausal status to enable a conservative, clinically interpretable stratification. We agree that this may introduce misclassification (e.g., perimenopause, surgical menopause, or earlier/later natural menopause), which could affect the estimated proportion of women eligible for vaginal oestrogen. 

Changes made in the manuscript
We retained the Methods statement that age ≥50 was used as a pragmatic proxy and strengthened the limitations in the Discussion to highlight potential misclassification and its impact on interpretation.

Comment 4

The paper concludes preventive care was misaligned with guidelines, but did it account for patient-level contraindications/preferences, prior failure of therapies, or physician reasoning (e.g., limited access to therapies, local prescribing constraints such as furazidin availability)? How can “gap” be separated from appropriate individualised decision-making?

Response
We agree that a divergence from guideline recommendations may reflect either a true evidence–practice gap or appropriate individualised decision-making. Our survey did not systematically assess patient preferences, contraindications, prior treatment failures, or physician reasoning, and we therefore cannot distinguish between these explanations. 

Changes made in the manuscript
In the Discussion, we added an explicit limitation: physician reasoning, patient-level contraindications, and prior failures were not assessed, limiting our ability to separate guideline divergence from appropriate individualised care. We also refined language to avoid overstatement and to better align with shared decision-making principles highlighted in current guidelines.

Reviewer 2 Report

Comments and Suggestions for Authors

Weaknesses

1. Limited Sample Size and Generalizability: With only 36 participants from a single Polish urology clinic, the cohort is small and potentially biased toward more severe or refractory cases, as most had prior primary/secondary care consultations. This restricts broader applicability, particularly given regional prescribing patterns (e.g., furazidin dominance) that may not translate to settings where nitrofurantoin prevails. The exploratory analyses, while insightful, risk type I errors due to unadjusted multiple comparisons, and the use of age ≥50 as a menopause proxy overlooks hormonal variability.
2. Methodological Gaps in Questionnaire Validation and Data Handling: The questionnaire, though comprehensive, lacks validation details (e.g., pilot testing, reliability metrics like Cronbach's alpha), which are essential for patient-reported instruments. Free-text harmonization for antibiotics is mentioned but not exemplified, potentially introducing subjectivity. Diagnostic practices (e.g., urine culture rates) are reported descriptively without stratification by episode complexity or resistance patterns, limiting depth. Additionally, satisfaction is assessed on a simplistic three-level scale, missing nuanced tools like the Recurrent UTI Impact Questionnaire (RUTIIQ), which the authors reference but do not employ.
3. Incomplete Integration of Evidence and Underdeveloped Discussion: While guidelines are cited, the discussion occasionally lacks critical appraisal; for instance, the mixed evidence for postcoital voiding is noted, but not contrasted with stronger alternatives like patient-initiated therapy. The seasonality findings (55.6% no pattern) are presented without exploring confounders (e.g., hydration behaviors). Moreover, the manuscript underemphasizes emerging therapies, such as D-mannose (dismissed via one trial without acknowledging prior meta-analyses), and omits discussion of microbiome influences or long-term outcomes.
4. Presentation and Structural Issues: Figures are informative but could be optimized (e.g., Figure 1 lacks error bars; Figure 3 uses proportions without raw numbers). The abstract and conclusions reiterate findings without sufficiently emphasizing novel contributions.  

These deficiencies, if unaddressed, could diminish the manuscript's persuasiveness and alignment with evidence-based medicine principles.

 

Recommendations for Improvement

To bolster the manuscript's suitability for publication, I propose the following enhancements, ordered by priority for methodological and interpretive rigor.

1. Expand Sample and Enhance Generalizability: Recruit a larger, multicenter cohort to mitigate selection bias and enable subgroup analyses (e.g., by age or recurrence frequency). Validate the menopause proxy with self-reported status or hormonal assays in future iterations. Adjust for multiple comparisons in statistical tests (e.g., via Bonferroni correction) and stratify diagnostics by guideline criteria to strengthen causal inferences.
2. Refine Questionnaire and Analytical Methods: Provide validation evidence for the instrument, including content validity and test-retest reliability. Incorporate established tools like RUTIIQ for satisfaction and QoL assessment. Exemplify data harmonization in supplements and explore confounders (e.g., comorbidity indices) through multivariable regression, even if exploratory, to elucidate associations like seasonality or education-knowledge links.
3. Deepen Evidence Integration and Discussion: Critically appraise conflicting evidence (e.g., D-mannose trials) and expand on microbiome-gut-UTI interactions, citing recent reviews. Elaborate on implications for underserved populations (e.g., premenopausal vs. postmenopausal) and propose a conceptual framework (e.g., a figure linking knowledge gaps to care mismatches). Address potential biases in self-report and suggest biomarkers (e.g., urine NGS for pathogens) for validation studies.

Comments on the Quality of English Language

Polish Presentation and Structure:

Correct typographical errors.

Streamline the discussion to avoid repetition, using subheadings for clarity (e.g., "Stewardship Challenges," "Preventive Opportunities").

Author Response

We revised the manuscript as follows:

1. We updated the Introduction to critically reflect contemporary evidence on non-antibiotic strategies, explicitly incorporating the 2014 randomised trial on D-mannose and clarifying its implications for prophylaxis recommendations.

2. We expanded the Discussion by adding a microbiome-oriented interpretive framework linking rUTI recurrence, dysbiosis, and the potential unintended consequences of repeated antibiotic exposure, thereby contextualising stewardship and non-antibiotic approaches.

3. We strengthened subgroup-focused interpretation by discussing prevention and counselling implications across life-stage proxies (age ≥50 as a pragmatic proxy for postmenopausal status) and behavioural triggers (intercourse-related recurrence among sexually active participants).

4. We expanded the Limitations to more clearly acknowledge self-report bias, lack of record linkage, and the absence of systematic assessment of physician reasoning, contraindications/preferences, and prior therapy failures, and we highlighted the need for prospective validation using objective microbiology.

Round 2

Reviewer 2 Report

Comments and Suggestions for Authors

Weaknesses

Despite these enhancements, several limitations persist, tempering the manuscript's robustness. The sample size remains modest (N=36), drawn from a single Polish urology clinic, which introduces selection bias toward more refractory cases and restricts external validity. While the revisions note this as a "clinically relevant subgroup," the absence of multicenter data or broader demographic representation (e.g., limited urban dwellers at 22.2%) hinders applicability to diverse populations. Reliance on self-reported histories without corroboration from medical records heightens susceptibility to recall inaccuracies, particularly for antibiotic exposures and diagnostic practices, where under- or overestimation could skew proportions (e.g., 55.6% ciprofloxacin use).

Interpretative nuances require further caution: exploratory p-values (e.g., for education-knowledge correlations) are unadjusted for multiplicity, risking Type I errors in a small cohort. The proxy of age ≥50 for postmenopausal status, while pragmatic, overlooks individualized hormonal assessments, potentially underestimating vaginal estrogen candidacy. Regional emphases, such as furazidin's analogy to nitrofurantoin, are clarified but may confuse international readers unfamiliar with local formularies. Minor residual inconsistencies include redundant phrasing in the discussion (e.g., repeated emphasis on "gaps between evidence and practice") and incomplete exploration of why certain strategies (e.g., cranberry) exhibit high awareness but low uptake.

Suggestions for Improvement

To further strengthen the manuscript for publication, the following targeted revisions are recommended:

1. Enhance Sample and Design Considerations: Acknowledge the exploratory nature more explicitly by suggesting future multicenter, prospective studies with larger cohorts to validate findings. Incorporate sensitivity analyses for key proportions (e.g., antibiotic exposures) to assess potential bias magnitudes.
2. Incorporate Objective Validation: Where feasible, propose linking future iterations to electronic health records for verification of self-reports. This would mitigate recall bias and enable subgroup analyses by verified resistance patterns or treatment failures.
3. Refine Statistical and Interpretative Elements: Apply Bonferroni corrections or similar adjustments for multiple comparisons in exploratory tests. Elaborate on barriers to vaginal estrogen uptake (e.g., via qualitative insights) and clarify furazidin's role with a brief comparative note to nitrofurantoin's global evidence base.
4. Improve Presentation and Clarity: Streamline repetitive elements in the discussion for conciseness. Expand the supplementary materials to include the full questionnaire (in Polish and English) for reproducibility. Consider a table summarizing guideline alignments versus reported practices to visually underscore key gaps.
5. Broaden Implications: Discuss integration with microbiome research, as noted in new references, to frame rUTI as a dysbiosis-related condition amenable to targeted prophylaxis. This would elevate the manuscript's forward-looking value.

Author Response

We thank the reviewer for these insightful comments. We acknowledge that the small, single-centre cohort (N=36) represents a specialist-referred population, which may introduce selection bias and limit external validity, particularly given the limited urban representation. We have strengthened the Study Design and Limitations sections to explicitly state the exploratory nature of the study and the constraints of relying on self-reported data, which may be affected by recall bias for antibiotic exposure and diagnostic practices.

To address interpretative concerns, we clarified that all between-group analyses are exploratory and applied a Holm–Bonferroni adjustment to contextualize multiplicity-related Type I error risk. We also explicitly note the limitations of using age ≥50 years as a proxy for postmenopausal status and the potential underestimation of vaginal estrogen eligibility. To improve clarity for international readers, we refined the description of furazidin as a regional nitrofuran analogous to nitrofurantoin.

Finally, we streamlined repetitive discussion elements, expanded interpretation of the awareness–uptake gap for cranberry products, added a table summarizing guideline alignment versus reported practices, and broadened the implications by integrating a microbiome-oriented perspective.