Intravaginal boric acid treatment for recurrent bacterial vaginosis: short-term effects on vaginal health parameters and patient satisfaction

Abstract

Background

Recurrent bacterial vaginosis affects 15–30% of women with bacterial vaginosis, demonstrating recurrence rates exceeding 60% after conventional therapy modalities. Alternative treatments are needed when standard therapies fail.

Objective

To evaluate short-term effects of intravaginal boric acid on vaginal health parameters and patient satisfaction in women with recurrent bacterial vaginosis resistant to conventional treatment.

Methods

This retrospective observational study analyzed 52 women with recurrent bacterial vaginosis (≥ 3 episodes within 12 months) treated with intravaginal boric acid (600 mg daily for 14 days) between January 2022 and December 2023. All participants had documented failure or rapid recurrence following metronidazole or clindamycin therapy. Primary outcomes included Vaginal Health Index and Maturation Index. Secondary outcomes comprised validated patient-reported measures (Quality of Sexual Experience Scale, Female Sexual Distress Scale) and clinical symptoms. Assessments occurred at baseline and Day 30 (16 days post-treatment).

Results

Participants (mean age 34.7 ± 6.7 years, 4.2 ± 1.3 prior episodes) demonstrated significant improvements. Nugent scores decreased from 7.8 ± 0.8 to 3.1 ± 1.0 (p < 0.001), with 88.5% achieving scores < 4. Clue cells decreased from 68.4% to 8.3% (p < 0.001). Vaginal Health Index improved from 12.1 ± 1.9 to 20.5 ± 1.5 (p < 0.001, Cohen’s d = 4.9). Quality of Sexual Experience scores increased from 13.8 ± 2.3 to 38.7 ± 3.1, while Female Sexual Distress scores decreased from 38.9 ± 5.2 to 8.4 ± 2.9 (both p < 0.001). Vaginal odor prevalence decreased from 92.3% to 1.9%. Mild adverse events occurred in 7.8% of participants, with all participants completing the 14-day treatment course.

Conclusions

Intravaginal boric acid was associated with significant short-term improvements in microbiological, clinical, and patient-reported outcomes. However, the single-arm retrospective design and limited follow-up preclude definitive conclusions about treatment efficacy.

Introduction

Bacterial vaginosis represents the most common vaginal disorder in reproductive-aged women, with prevalence ranging from 23 to 29% in the general population [7, 14]. Recurrent bacterial vaginosis, defined as three or more episodes annually, affects approximately 15–30% of women with bacterial vaginosis and presents clinical management challenges [8, 21]. This condition impacts women’s physical health through increased susceptibility to sexually transmitted infections including HIV, elevated risk of pelvic inflammatory disease, and associations with preterm birth and postpartum endometritis in pregnant women [4, 14]. Psychosocial consequences include decreased sexual satisfaction, relationship difficulties, and reduced quality of life, with affected women reporting emotional distress associated with repeated treatment failures [4]. The economic burden encompasses healthcare costs from clinical visits, diagnostic testing, and prescription medications, indicating a need for alternative therapeutic approaches.

Standard treatment protocols utilize metronidazole or clindamycin as first-line therapies according to international guidelines, with initial efficacy rates of approximately 70–80% [15, 23]. However, recurrence rates exceed 60% within six months and approach 80% within twelve months following treatment completion [8, 20]. Treatment failure relates to the polymicrobial nature of bacterial vaginosis, which involves Gardnerella vaginalis, Atopobium vaginae, Prevotella species, and other anaerobic organisms that form microbial communities resistant to single-agent antimicrobial therapy [14]. These organisms establish biofilms on vaginal epithelium that reduce antimicrobial penetration and host immune response effectiveness [22]. Molecular studies indicate these biofilms, composed of G. vaginalis and A. vaginae, persist after clinical cure and may serve as reservoirs for recurrent infection [18, 22]. Repeated antibiotic exposure contributes to resistance development, disrupts vaginal microbiota balance, and can cause secondary vulvovaginal candidiasis.

Intravaginal boric acid has been used as an alternative treatment for vaginal infections for several decades, with recent literature examining its application in bacterial vaginosis management [9]. This weakly acidic compound (H₃BO₃) demonstrates antimicrobial activity against bacterial vaginosis-associated anaerobes, including organisms within biofilms that show resistance to conventional antibiotics [16]. The compound’s pH-modulating properties create conditions less favorable for pathogen growth while supporting lactobacilli recovery. A systematic review by Lærkeholm Müller and colleagues [9] documented boric acid use in cases where standard treatments had limited success [9]. Clinical studies have examined various treatment protocols, including monotherapy and combination approaches. Reichman et al. [17] reported that adding intravaginal boric acid to suppressive metronidazole therapy reduced recurrence rates compared to antibiotics alone [17]. Conversely, Marrazzo et al. [10] found variable efficacy when boric acid was used as monotherapy in a randomized controlled trial, suggesting treatment outcomes depend on patient selection and dosing regimens. The biofilm-disrupting properties of boric acid may address bacterial reservoirs that persist despite antibiotic treatment [10].

Current literature regarding boric acid effectiveness consists primarily of small observational studies with methodological limitations that restrict evidence-based practice recommendations. Most investigations have focused on microbiological outcomes without assessing vaginal health parameters, symptom resolution patterns, or quality-of-life measures relevant to patient care. Treatment duration, dosing regimens, and patient selection criteria for boric acid therapy remain undefined in the literature. While previous studies have examined boric acid efficacy primarily through microbiological outcomes, comprehensive data incorporating validated vaginal health indices and standardized patient-reported outcome measures remain limited. This study addresses this gap by evaluating intravaginal boric acid therapy using both objective vaginal health parameters (VHI, MI) and validated quality-of-life instruments (QSES, FSDS) in a well-characterized cohort of women with treatment-resistant recurrent bacterial vaginosis. We assessed short-term effects of a 14-day treatment protocol on microbiological parameters through Nugent scoring, vaginal health indices, patient-reported measures of sexual function and distress, and clinical symptom resolution. The study population consisted of women with documented recurrent bacterial vaginosis who experienced treatment failure or rapid recurrence following standard antibiotic therapy.

Materials and methods

Study design and setting

This retrospective observational study analyzed medical records of patients who received intravaginal boric acid treatment for recurrent bacterial vaginosis between January 2022 and December 2023 at the gynecology outpatient service of Haseki Training and Research Hospital. Data were collected retrospectively from patient medical records following institutional review board approval (Registry no 83-2024, dated October 31, 2024). The ethics approval was obtained retrospectively to analyze existing clinical data, as is standard practice for retrospective chart review studies. All patients had previously provided written informed consent for treatment and data utilization as part of routine clinical care. This single-arm observational study design without a control group represents a limitation that is acknowledged in our interpretation of results.

Participant selection

Sample size determination was based on detecting a clinically meaningful change in VHI scores of 5 points with 80% power and alpha of 0.05, requiring 52 participants accounting for 15% attrition.

Medical records were reviewed to identify women aged 18–49 years with recurrent bacterial vaginosis who received intravaginal boric acid treatment. Recurrent bacterial vaginosis was defined as three or more documented episodes within the preceding 12 months, confirmed by both Amsel criteria and Nugent scoring (≥ 7). The diagnosis and treatment decisions were made by attending physicians as part of routine clinical care, not as part of a prospective research protocol. Inclusion required documented failure or rapid recurrence (< 30 days) following conventional antibiotic therapy with metronidazole or clindamycin within the previous six months.

From a total of 78 women who received boric acid treatment for bacterial vaginosis during the study period, 52 met all inclusion criteria and had complete follow-up data available for analysis. Exclusions included: concurrent sexually transmitted infections confirmed by nucleic acid amplification testing (n = 8), pregnancy or lactation confirmed by serum beta-hCG or documented breastfeeding status (n = 5), incomplete follow-up data with missing outcome assessments (n = 7), immunocompromising conditions including HIV infection or immunosuppressive therapy (n = 3), and concurrent antifungal therapy for vulvovaginal candidiasis (n = 3). All participants were premenopausal, confirmed by regular menstrual cycles and absence of menopausal symptoms documented in medical records.

Diagnostic procedures

Bacterial vaginosis diagnosis was established using standardized Amsel criteria requiring presence of at least three of four clinical findings: thin, homogeneous vaginal discharge; vaginal pH exceeding 4.5 measured using pH indicator strips; positive whiff test with 10% potassium hydroxide application; and presence of clue cells on saline wet mount microscopy comprising more than 20% of epithelial cells observed. The percentage of clue cells was quantified and recorded at both baseline and follow-up assessments to evaluate treatment response.

Nugent scoring was performed on Gram-stained vaginal smears according to established methodology [12]. Scores range from 0 to 10, with bacterial vaginosis confirmed by Nugent scores of 7 or greater, scores of 4–6 indicating intermediate flora, and scores of 0–3 representing normal lactobacilli-dominant microbiota. All microscopic evaluations were performed by experienced pathologists blinded to treatment status, ensuring objective assessment throughout the diagnostic process.

Treatment protocol

Patients received commercially available intravaginal boric acid suppositories containing 600 mg of pharmaceutical-grade boric acid administered once daily at bedtime for 14 consecutive days. The Day 30 assessment timepoint (16 days post-treatment completion) was selected based on CDC guidelines for test-of-cure timing in bacterial vaginosis, allowing sufficient time for vaginal flora reestablishment while capturing early treatment response [23]. Standardized written and verbal instructions were provided regarding proper insertion techniques, hygiene measures, and potential adverse effects. Patients were explicitly warned about the potential toxicity of boric acid if ingested orally and instructed to store medication securely away from children, with this safety counseling documented in medical records.

During the treatment period, patients were advised to avoid sexual intercourse, douching, and use of additional vaginal preparations or medications that might interfere with treatment assessment. Concurrent systemic medications were permitted provided they did not include antibiotics or antifungal agents that might confound treatment evaluation. Patient education materials were provided in Turkish language with visual aids to ensure comprehension.

Outcome measures

Primary outcomes included objective measures of vaginal health assessed through validated scoring systems. The Vaginal Health Index (VHI) represents a composite scoring system evaluating five parameters: vaginal pH (1–5 points), epithelial integrity (1–5 points), fluid volume (1–5 points), epithelial thickness (1–5 points), and moisture levels (1–5 points), with total scores ranging from 5 to 25 and higher scores indicating superior vaginal health status [3]. Each VHI component was assessed independently to determine specific areas of improvement.

The Maturation Index (MI) assessment involved cytological evaluation of vaginal epithelial cells to determine estrogen effect and epithelial maturation patterns. The index represents the percentage distribution of parabasal, intermediate, and superficial cells observed in vaginal cytology specimens, with higher superficial cell percentages indicating greater epithelial maturation and improved vaginal health status.

Secondary outcomes focused on patient-reported measures using validated Turkish language versions of internationally established questionnaires. The Quality of Sexual Experience (QSE) Scale was originally developed by Sanders et al. [19] as a brief event-level measure to assess the quality of sexual experiences across diverse populations regardless of age, gender, sexual orientation, or relationship status [19]. The original validation study in a nationally representative sample of 1,466 women and 869 men in the United States demonstrated excellent psychometric properties with QSE scores significantly correlated with overall sexual quality ratings (r = 0.901, p < 0.001 for both genders). The scale comprises 11 items measuring various dimensions of sexual experience quality, with total scores ranging from 11 to 55 where higher scores indicate better sexual experiences. The Turkish adaptation and validation were conducted by Yassa et al. [24] following international translation and cultural adaptation guidelines, including forward-translation, synthesis, back-translation, expert committee review, pretesting, and cognitive debriefing [24]. The Turkish version was validated in Turkish women, demonstrating strong internal consistency (Cronbach’s alpha = 0.91), excellent test-retest reliability (r = 0.89, p < 0.001), and confirmed factor structure replicating the original unidimensional model. The scale successfully discriminated between women with and without sexual dysfunction (p < 0.001). In our study population, the QSE Scale demonstrated good internal consistency with Cronbach’s alpha of 0.85, supporting the reliability of this instrument in women with recurrent bacterial vaginosis.

The Female Sexual Distress Scale (FSDS) was originally developed by Derogatis et al. [6] to specifically measure sexually related personal distress, addressing the essential component of distress in female sexual dysfunction definitions [6]. The original validation study in approximately 500 women across three studies demonstrated a unidimensional factor structure with excellent internal consistency (Cronbach’s alpha = 0.93 for the 20-item version and 0.91 for the refined 12-item version) and high test-retest reliability (r = 0.91, p < 0.001). The scale showed strong discriminative ability between sexually functional and dysfunctional women and demonstrated sensitivity to treatment response. The scale consists of 12 items in its refined version, assessing distress, frustration, and concern related to sexual activity, with scores ranging from 0 to 52 where higher scores indicate greater sexual distress. The Turkish validation was performed by Aydın et al. [2] in a sample of 265 Turkish women, including both healthy controls and women with sexual dysfunction [2]. The Turkish version showed excellent psychometric properties with Cronbach’s alpha of 0.92, test-retest reliability coefficient of 0.88 (p < 0.001), and confirmed criterion validity through significant correlations with other validated sexual function measures. The scale successfully discriminated between women with and without sexual dysfunction (p < 0.001) with a recommended cutoff score of 11 for identifying clinically significant sexual distress in Turkish populations. In the current study, the FSDS demonstrated good reliability with Cronbach’s alpha of 0.87, confirming its utility for assessing sexual distress in our patient population.

Additional clinical outcome measures included systematic assessment of symptom resolution including vaginal discharge characteristics, odor presence, pruritus, dyspareunia, and vaginal pain. The percentage of clue cells on wet mount microscopy was recorded as a quantitative measure of bacterial vaginosis resolution.

Data collection

Baseline assessments were documented from medical records prior to treatment initiation, including comprehensive medical history, physical examination findings, laboratory test results, and completion of validated questionnaire instruments. Previous gynecological surgery types were categorized as cesarean section, vaginal delivery with episiotomy, hysterectomy, ovarian surgery, or other procedures as documented in surgical records. Follow-up evaluations were performed 30 days after treatment initiation (16 days post-treatment completion) and included repeat clinical assessments, laboratory testing, and questionnaire administration.

Adverse events were systematically assessed through a structured patient interview using a standardized checklist that included common side effects reported in previous boric acid studies: vaginal burning, irritation, discharge changes, and systemic symptoms. Patients were asked open-ended questions about any symptoms experienced during treatment, followed by specific queries about each potential adverse effect. All reported adverse events were graded using Common Terminology Criteria for Adverse Events (CTCAE) version 5.0.

Treatment compliance was verified through multiple methods: patient self-report using a daily diary card, suppository count at the follow-up visit, and specific questioning about missed doses or treatment interruption. Complete compliance was defined as use of all 14 suppositories without interruption. Patients who missed more than two doses were to be excluded from per-protocol analysis, though no participants met this exclusion criteria.

Patient characteristics collected included demographic information (age, ethnicity, education level), reproductive history (gravidity, parity, previous abortion), number of sexual partners in the past 12 months documented through standardized sexual history questionnaire, anthropometric measurements (body mass index calculated from measured height and weight), medical history (diabetes mellitus, hypothyroidism, other chronic conditions), current medications (defined as any prescription or over-the-counter medications used regularly for at least 30 days), contraceptive methods, and previous surgical procedures.

Statistical analysis

Statistical analysis was performed using IBM SPSS version 29.0 (IBM Corp., Armonk, NY, USA), with graphical presentations prepared using GraphPad Prism version 9.0 (GraphPad Software, San Diego, CA, USA). Results are presented as mean with standard deviation for normally distributed variables, median (interquartile range) for non-normal distributions, or n(%) for categorical variables. Confidence intervals (95% CI) were calculated for all primary outcomes using bootstrapping methods with 1000 iterations to account for non-normal distributions. Statistical significance was set at p < 0.05 for all comparisons. Normality was assessed using the Kolmogorov-Smirnov test with results showing non-normal distribution for most continuous variables (p < 0.05). For non-normally distributed paired data, the Wilcoxon signed-rank test was employed. Effect sizes were calculated using Cohen’s d for paired comparisons, with values of 0.2, 0.5, and 0.8 representing small, medium, and large effects respectively [5]. Categorical variables were analyzed using McNemar’s test for paired proportions. Exploratory subgroup analyses were performed to identify potential predictors of treatment response, including age (< 35 vs. ≥ 35 years), body mass index (< 30 vs. ≥ 30 kg/m²), baseline bacterial vaginosis severity (Nugent score 7–8 vs. 9–10), and number of previous recurrences (3–4 vs. ≥ 5).

Results

The study included 52 women with recurrent bacterial vaginosis who completed the full 14-day treatment protocol and 30-day follow-up assessment. Table 1 presents baseline demographic and clinical characteristics of the study population. The mean age was 34.7 ± 6.7 years (range 22–49), with all participants confirmed as premenopausal based on regular menstrual cycles documented in medical records. Participants experienced a mean of 4.2 ± 1.3 bacterial vaginosis episodes in the previous 12 months (range 3–8). The mean number of sexual partners in the past 12 months was 1.1 ± 0.3 (range 1–2), based on standardized sexual history questionnaires.

Table 1.

Baseline characteristics of the study population (n = 52)

Variable	n (%) or Mean ± SD	Median (Range)
Age (years)	34.7 ± 6.7	35.0 (22–49)
Body mass index (kg/m²)	27.8 ± 5.4	27.5 (18–40)
Premenopausal status	52 (100.0)	-
Ethnicity
Turkish	48 (92.3)	-
Arabic	4 (7.7)	-
Education
High school or less	4 (7.7)	-
Beyond high school	48 (92.3)	-
Gravidity	2.7 ± 1.7	3 (0–10)
Parity	2.4 ± 1.4	2 (0–6)
Previous abortion	0.4 ± 0.8	0 (0–4)
Number of BV recurrences (past 12 months)	4.2 ± 1.3	4 (3–8)
Number of sexual partners (past 12 months)	1.1 ± 0.3	1 (1–2)
Previous gynecological surgerya	22 (42.3)	-
Regular medication useb	13 (25.0)	-
Intrauterine device	17 (32.7)	-
Medical comorbidities
None	36 (69.2)	-
Diabetes mellitus	2 (3.8)	-
Hypothyroidism	11 (21.2)	-
Other	3 (5.8)	-

^aIncludes cesarean section (n = 15), vaginal delivery with episiotomy (n = 4), ovarian surgery (n = 2), other (n = 1)

^bDefined as prescription or over-the-counter medications used regularly for ≥ 30 days, excluding antibiotics and antifungals; primarily thyroid hormone replacement (n = 11)

Table 2 presents changes in clinical symptoms between baseline and the 30-day follow-up assessment (16 days post-treatment completion). Vaginal odor prevalence decreased from 92.3% to 1.9% of patients (p < 0.001). All participants presented with vaginal discharge at baseline, which was present in 15.4% at follow-up (p < 0.001). Vaginal itching decreased from 73.1% to 9.6% (p < 0.001), vaginal pain from 84.6% to 5.8% (p < 0.001), and dyspareunia from 80.8% to 11.5% (p < 0.001). Vaginal dryness changed from 30.8% to 21.2% (p = 0.063).

Table 2.

Clinical symptoms before treatment and at 30-day follow-up (16 days post-treatment)

Variable	Before Treatment n (%)	At 30-Day Follow-up n (%)	p valuea
Vaginal odor	48 (92.3)	1 (1.9)	< 0.001
Vaginal discharge	52 (100.0)	8 (15.4)	< 0.001
Vaginal itching	38 (73.1)	5 (9.6)	< 0.001
Vaginal pain	44 (84.6)	3 (5.8)	< 0.001
Dyspareunia	42 (80.8)	6 (11.5)	< 0.001
Vaginal dryness	16 (30.8)	11 (21.2)	0.063

^aMcNemar test for paired proportions

Table 3 presents changes in objective clinical measures and patient-reported outcomes. Nugent scores decreased from 7.8 ± 0.8 to 3.1 ± 1.0 (p < 0.001), with 46 patients (88.5%) achieving scores below 4. The percentage of clue cells on wet mount microscopy decreased from 68.4 ± 18.2% to 8.3 ± 6.1% (p < 0.001). The Vaginal Health Index improved from 12.1 ± 1.9 to 20.5 ± 1.5 (p < 0.001), while the Maturation Index increased from 41.3 ± 8.6 to 52.7 ± 5.6 (p < 0.001). Quality of Sexual Experience Scale scores increased from 13.8 ± 2.3 to 38.7 ± 3.1 (p < 0.001), and Female Sexual Distress Scale scores decreased from 38.9 ± 5.2 to 8.4 ± 2.9 (p < 0.001).

Table 3.

Comparison of Microbiological and vaginal health parameters

Variable	Before Treatment Mean ± SD	At 30-Day Follow-up Mean ± SD	Mean Difference (95% CI)a	Effect Size (Cohen’s d)	p valueb
Nugent score	7.8 ± 0.8	3.1 ± 1.0	4.7 (4.4-5.0)	5.2	< 0.001
Clue cells (%)	68.4 ± 18.2	8.3 ± 6.1	60.1 (54.7–65.5)	4.4	< 0.001
Vaginal Health Index	12.1 ± 1.9	20.5 ± 1.5	8.4 (7.8-9.0)	4.9	< 0.001
Maturation Index	41.3 ± 8.6	52.7 ± 5.6	11.4 (9.2–13.6)	1.6	< 0.001
QSES scorec	13.8 ± 2.3	38.7 ± 3.1	24.9 (23.9–25.9)	9.1	< 0.001
FSDS scored	38.9 ± 5.2	8.4 ± 2.9	30.5 (28.8–32.2)	7.2	< 0.001

^a95% Confidence intervals calculated using bootstrapping with 1000 iterations

^bWilcoxon signed-rank test used due to non-normal distribution (Kolmogorov-Smirnov test p < 0.05 for all variables)

^cQSES: Quality of Sexual Experience Scale (range 11–55, higher scores indicate better sexual experience)

^dFSDS: Female Sexual Distress Scale (range 0–52, higher scores indicate greater distress)

Individual components of the Vaginal Health Index are presented in Table 4. Vaginal pH scores increased from 2.1 ± 0.6 to 4.3 ± 0.5 (p < 0.001), with 45 patients (86.5%) achieving normal pH (≤ 4.5). Epithelial integrity scores improved from 2.3 ± 0.5 to 4.1 ± 0.4 (p < 0.001), with 41 patients (78.8%) showing normal or near-normal integrity. Fluid volume scores changed from 2.5 ± 0.6 to 3.9 ± 0.5 (p < 0.001), with normalization in 38 patients (73.1%). Epithelial thickness scores increased from 2.6 ± 0.5 to 4.2 ± 0.4 (p < 0.001), reaching normal range in 43 patients (82.7%). Moisture scores improved from 2.6 ± 0.7 to 4.0 ± 0.6 (p < 0.001), with improvement in 36 patients (69.2%).

Table 4.

Individual vaginal health index component scores

VHI Component	Before Treatment Mean ± SD	At 30-Day Follow-up Mean ± SD	p valuea
pH	2.1 ± 0.6	4.3 ± 0.5	< 0.001
Epithelial integrity	2.3 ± 0.5	4.1 ± 0.4	< 0.001
Fluid volume	2.5 ± 0.6	3.9 ± 0.5	< 0.001
Epithelial thickness	2.6 ± 0.5	4.2 ± 0.4	< 0.001
Moisture	2.6 ± 0.7	4.0 ± 0.6	< 0.001

^aWilcoxon signed-rank test; each component scored 1–5, with higher scores indicating better vaginal health

Figure 1 displays the distribution of Vaginal Health Index measurements before and after treatment using violin plots. Pre-treatment values showed a mean of 12.1±1.9 with a coefficient of variation of 15.7%. Post-treatment measurements showed a mean of 20.5±1.5 with a coefficient of variation of 7.3%.

Fig. 1.

Violin plot of VHI measurements before and after treatment with median and interquartile ranges. Wilcoxon signed-rank test p < 0.001

Figure 2 illustrates Maturation Index distributions. Pre-treatment values were 41.3±8.6 with a coefficient of variation of 20.8%. Post-treatment values were 52.7±5.6 with a coefficient of variation of 10.6%.

Fig. 2.

Violin plot of MI measurements before and after treatment with median and interquartile ranges. Wilcoxon signed-rank test p < 0.001

Exploratory subgroup analyses examined potential predictors of treatment response, defined as achieving Nugent score < 4. No statistically significant differences were observed based on age groups (< 35 years, n = 26 vs. ≥ 35 years, n = 26; p = 0.42), body mass index categories (< 30 kg/m², n = 35 vs. ≥ 30 kg/m², n = 17; p = 0.38), baseline Nugent score severity (7–8, n = 31 vs. 9–10, n = 21; p = 0.51), or number of previous recurrences (3–4 episodes, n = 29 vs. ≥ 5 episodes, n = 23; p = 0.67). The study was not powered for these subgroup comparisons.

No serious adverse events were reported. Minor adverse effects included mild transient vaginal burning sensation in 4 participants (7.7%) during the first three days of treatment, which resolved without intervention. Six participants (11.5%) reported increased vaginal discharge during treatment. All adverse events were Grade 1 (mild) according to CTCAE version 5.0. Treatment compliance was verified through patient diary cards, suppository count, and structured interviews. All 52 participants (100%) completed the 14-day treatment course without interruption.

Discussion

This retrospective observational study examined the short-term effects of intravaginal boric acid treatment in 52 women with recurrent bacterial vaginosis who had experienced treatment failure with conventional antibiotics. At the 30-day follow-up assessment, we observed statistically significant improvements in microbiological parameters, vaginal health indices, and patient-reported outcomes. The Nugent scores decreased from the bacterial vaginosis range (7.8 ± 0.8) to within normal limits (3.1 ± 1.0), with 88.5% of participants achieving scores below 4, indicating restoration of normal vaginal flora. The parallel reduction in clue cell percentage from 68.4% to 8.3% provides additional microbiological evidence of bacterial vaginosis resolution. These microbiological improvements were accompanied by substantial enhancements in vaginal tissue health, as evidenced by the 69.4% improvement in Vaginal Health Index scores and 27.6% increase in Maturation Index values. The magnitude of these changes, reflected in large effect sizes (Cohen’s d ranging from 1.6 to 9.1), suggests clinically meaningful improvements beyond statistical significance alone.

Assessment of individual VHI components showed varied improvement patterns across different parameters, with pH normalization occurring most consistently (86.5% of participants). Vaginal pH normalization occurred in 86.5% of participants, representing the most consistent improvement among VHI parameters, which aligns with boric acid’s known pH-modulating properties. Epithelial integrity and thickness showed substantial improvements in 78.8% and 82.7% of participants respectively, suggesting restoration of vaginal epithelial barrier function. The more variable response in moisture levels (69.2% improvement) may reflect individual differences in epithelial recovery rates or concurrent factors affecting vaginal lubrication. Patient-reported outcomes demonstrated particularly notable improvements, with Quality of Sexual Experience Scale scores increasing by 180% and Female Sexual Distress Scale scores decreasing by 78%. The observed improvements in sexual function scores suggest that effective bacterial vaginosis treatment may positively influence quality of life parameters. The resolution of cardinal symptoms, particularly the 90.4% point reduction in vaginal odor prevalence and 84.6% resolution of vaginal discharge, addresses the most distressing aspects of bacterial vaginosis that prompt women to seek repeated medical care.

The mechanism underlying boric acid’s therapeutic effects likely involves multiple complementary pathways that address the complex pathophysiology of bacterial vaginosis. Boric acid’s antimicrobial activity against anaerobic bacteria, including Gardnerella vaginalis and Atopobium vaginae, disrupts the polymicrobial communities characteristic of bacterial vaginosis [9]. The compound’s ability to penetrate and disrupt bacterial biofilms may be particularly important, as these structures contribute to antibiotic resistance and treatment failure in recurrent cases [22]. The restoration of acidic vaginal pH creates an environment favorable for lactobacilli recovery while inhibiting pathogenic bacterial growth [18]. Our observation of reduced variability in post-treatment measurements, as evidenced by decreased coefficients of variation for both VHI (15.7% to 7.3%) and MI (20.8% to 10.6%), suggests that boric acid produces consistent therapeutic effects across diverse patient populations, regardless of baseline disease severity or individual characteristics.

Our findings align with several previous investigations while highlighting important areas of divergence in the literature. Powell et al. [16] reported similar patterns of clinical improvement with boric acid in their clinical practice assessment, though their study lacked the comprehensive outcome measures employed in our investigation [16]. Reichman et al. [17] demonstrated that combining boric acid with suppressive metronidazole therapy achieved superior outcomes compared to antibiotics alone, with their 12-month follow-up showing recurrence rates of 28% in the combination group versus 48% with metronidazole monotherapy [17]. However, Marrazzo et al. [10] found less favorable results in their randomized controlled trial of TOL-463 (a boric acid-containing vaginal gel), with only 50% of participants achieving clinical cure at day 17 [10]. These divergent findings may reflect differences in formulation, dosing regimens, patient selection criteria, or outcome assessment methods, highlighting the need for standardized protocols in future research.

Recent evidence has increasingly supported combination approaches for managing recurrent bacterial vaginosis. Muzny and Sobel [11] advocate for combining oral metronidazole with intravaginal boric acid based on their clinical experience with treatment-resistant cases, reporting that this combination addresses both planktonic bacteria and biofilm-associated organisms more effectively than either agent alone [11]. The systematic review by Lærkeholm Müller et al. [9] analyzed 14 studies encompassing 1,019 patients treated with various boric acid regimens, finding cure rates ranging from 40 to 100% depending on treatment duration and follow-up periods [9]. Their analysis highlighted the heterogeneity in treatment protocols and outcome measures across studies, emphasizing the need for standardized approaches to evaluate boric acid efficacy. The restoration of acidic vaginal pH through boric acid’s pH-modulating properties creates an environment favorable for lactobacilli recovery, as BV-associated anaerobic bacteria grow poorly in acidic conditions while lactobacilli thrive at pH levels between 3.5 and 4.5 [1, 13].

The clinical implications of our findings must be considered within the context of current treatment challenges in recurrent bacterial vaginosis management. While our results suggest that boric acid may offer an alternative for women who have failed conventional antibiotic therapy, several important considerations temper enthusiasm for widespread adoption. The short follow-up period in our study provides no information about long-term efficacy or recurrence prevention, which represents the primary therapeutic goal in recurrent bacterial vaginosis. The absence of a control group prevents determination of treatment effect magnitude beyond placebo response or natural disease fluctuation, particularly given that bacterial vaginosis symptoms can vary with menstrual cycle phases and other factors [8]. Furthermore, the retrospective design introduces potential selection bias, as patients receiving boric acid may have differed systematically from those receiving standard care in ways not captured by our data collection.

Safety considerations warrant careful attention when considering boric acid therapy. While our study documented only mild adverse effects in 7.7% of participants, all of which resolved without intervention, the potential for serious toxicity if boric acid is ingested orally necessitates comprehensive patient education [9]. Healthcare providers must ensure patients understand proper handling and storage requirements, particularly the critical importance of keeping suppositories away from children and pets. The favorable safety profile observed in our study aligns with previous reports, though long-term safety data remain limited. The 100% treatment compliance rate in our study may not reflect real-world adherence, particularly given the 14-day treatment duration and need for nightly administration.

Strengths and limitations

This study possesses several methodological strengths that enhance the validity and clinical relevance of our findings. The comprehensive assessment approach utilizing multiple validated outcome measures provides a holistic evaluation of treatment effects across microbiological, clinical, and quality-of-life domains. The inclusion of both objective measures (Nugent scores, VHI, MI, clue cell percentage) and subjective patient-reported outcomes (QSES, FSDS) captures the multidimensional impact of bacterial vaginosis and its treatment. The use of validated Turkish versions of questionnaires with documented psychometric properties strengthens the reliability of patient-reported outcomes. The standardized diagnostic criteria employing both Amsel criteria and Nugent scoring ensure accurate bacterial vaginosis diagnosis, while the blinded microscopic evaluation minimizes assessment bias. The 100% follow-up completion rate eliminates attrition bias, and the detailed documentation of individual VHI components provides mechanistic insights into treatment effects. The study population represents a clinically relevant cohort of women with documented treatment-resistant recurrent bacterial vaginosis, addressing an important unmet clinical need in gynecological practice.

However, significant limitations constrain the interpretation and generalizability of our findings. The retrospective, single-arm observational design precludes causal inference and cannot account for confounding factors, placebo effects, or regression to the mean. The absence of a control group prevents determination of treatment effect magnitude relative to standard care or placebo response, which may be substantial in conditions with subjective symptom components such as bacterial vaginosis. The short follow-up period of 30 days (16 days post-treatment) provides no information about recurrence prevention, which represents the primary concern in recurrent bacterial vaginosis management. The single-center experience in a specific Turkish population may limit generalizability to other healthcare settings and patient populations. The lack of vaginal microbiome analysis using molecular methods limits mechanistic understanding of treatment effects on microbial community composition. The retrospective identification of patients may have introduced selection bias, as treatment allocation was based on clinical judgment rather than randomization. The absence of long-term safety data and information about potential effects on fertility or pregnancy outcomes represents an important knowledge gap. These limitations underscore the preliminary nature of our findings and the need for prospective controlled trials to establish definitive evidence for boric acid efficacy in recurrent bacterial vaginosis.

Conclusions

This study found that intravaginal boric acid treatment was associated with significant short-term improvements in women with recurrent bacterial vaginosis who had failed conventional antibiotic therapy, including microbiological cure in 88.5% of participants, substantial improvements in vaginal health indices, and marked enhancements in sexual function and quality of life measures. These findings suggest that boric acid may represent an alternative treatment option for selected patients with treatment-resistant disease. The observed safety profile, with mild adverse effects in 7.7% of participants, appears acceptable, though long-term safety data remain limited. However, clinicians should interpret these results cautiously due to the retrospective single-arm design and limited follow-up period, which preclude definitive conclusions about long-term efficacy or recurrence prevention. The treatment should be positioned as a second-line option for women with documented treatment failure following appropriate antibiotic therapy, with clear communication about the preliminary nature of current evidence. Future prospective controlled trials with extended follow-up would help establish the optimal role of boric acid in managing this challenging condition.

Authors’ contributions

M.C. conceived the study design, coordinated data collection, performed statistical analysis, and drafted the manuscript. O.K. contributed to patient recruitment, data collection, and manuscript revision. M.G. participated in data collection and clinical assessments. A.C. supervised the study design, provided clinical oversight, and critically reviewed the manuscript. All authors read and approved the final manuscript.

Funding

No specific funding was received for this research.

Data availability

The datasets analyzed during this study are available from the corresponding author upon reasonable request, subject to institutional data sharing policies and patient privacy requirements.

Declarations

Ethics approval and consent to participate

This study was conducted in accordance with the Declaration of Helsinki and received approval from the Human Research Ethics Committee of Haseki Training and Research Hospital (Registry no 83-2024, dated October 31, 2024). As a retrospective study analyzing existing medical records, specific study consent was waived by the ethics committee in accordance with institutional policies for retrospective chart reviews. All patients had provided written informed consent for treatment and data utilization as part of routine clinical care.

Consent for publication

All authors consent to the publication of this manuscript. Patient data are presented in anonymized format to protect confidentiality.

Competing interests

The authors declare no competing interests.