Imaging characteristics and comparative evaluation of dual-fluorescence assay in various types of vaginitis: a cross-sectional study Authors: WenHui Qi, Liyao Wang, Junyi Bai, Jinfang Chen, Huanrong Li, Mengting Dong, Chen Wang, Fengxia Xue Published: 2026 ABSTRACT AbstractBackgroundRapid and accurate differentiation of common vaginitis remains challenging in routine clinical practice. Dual-fluorescence assay, which simultaneously visualizes microorganisms and host cells under blue light excitation, may serve as an adjunctive tool to conventional microscopy. This study explored the imaging characteristics of dual-fluorescence staining in cultured vaginal epithelial cells and standard reference microbial strains, and further evaluated its diagnostic performance in a clinical population of 758 women undergoing vaginal secretion testing.Results(1) Under 450-500 nm blue light excitation: Red fluorophores specifically labeled microorganisms (including pathogenic fungi, bacteria, and Trichomonas vaginalis), while green fluorophores counterstained host cells (e.g., epithelial cells and leukocytes). The distinct color contrast between microorganisms and host cells, combined with microbial morphological features, enabled rapid localization and identification of pathogens. (2) Dual-fluorescence assay demonstrated strong concordance with conventional microscopy: For bacterial vaginosis (BV), the sensitivity was 0.886 (95% CI: 0.796-0.955), specificity was 0.958 (95% CI: 0.931-0.979), and Kappa was 0.821 (95% CI: 0.756-0.901); for vulvovaginal candidiasis (VVC), sensitivity was 0.953 (95% CI: 0.840-0.988), specificity was 0.978 (95% CI: 0.953-0.990), and Kappa was 0.908 (95% CI: 0.847-0.963); for aerobic vaginitis (AV), sensitivity was 0.800 (95% CI: 0.535-0.848), specificity was 0.959 (95% CI: 0.924-0.995), and Kappa was 0.697 (95% CI: 0.616-0.863); and for trichomoniasis (TV), sensitivity was 0.833 (95% CI: 0.365-0.991), specificity was 0.995 (95% CI: 0.979-0.999), and Kappa was 0.766 (95% CI: 0.331-1.000).ConclusionThe dual-fluorescence assay technique enables simultaneous detection of four common types of vaginitis (including mixed vaginitis) while integrating the diagnostic advantages of both wet mount and Gram staining methods, thus demonstrating high clinical utility.Supplementary InformationThe online version contains supplementary material available at https://doi.org/10.1186/s12866-026-05345-5.Keywords: vaginitis, bacterial vaginosis, vulvovaginal candidiasis, aerobic vaginitis, trichomoniasis, dual-fluorescence assay, vaginal discharge microscopy, diagnostic test Introduction Reproductive tract infections (RTIs) are common diseases in obstetrics and gynecology, accounting for 55% of outpatient visits in gynecology clinics [1]. Without timely diagnosis and proper treatment, RTIs can lead to severe complications such as infertility, ectopic pregnancy, miscarriage, preterm birth, stillbirth, and neonatal infections, significantly impacting women’s reproductive health, family well-being, and public health burden [2–5]. The normal vaginal microbiota is dominated by Lactobacillus species, with only minor proportions of other commensal bacteria. Disruption of this balance, due to invasion by exogenous pathogens or overgrowth of endogenous opportunistic microbes, can lead to various forms of vaginitis. Trichomoniasis (TV) is caused by the protozoan Trichomonas vaginalis (T. vaginalis). Vulvovaginal candidiasis (VVC) results from excessive growth of Candida species, most commonly C. albicans. Bacterial vaginosis (BV) occurs when Lactobacillus species levels decrease and facultative anaerobes (e.g., Gardnerella vaginalis) and strict anaerobes (e.g., Prevotella, Mobiluncus, Bacteroides, Atopobium vaginae) proliferate. Aerobic vaginitis (AV) is characterized by overgrowth of aerobic or facultative anaerobic bacteria, such as Staphylococcus aureus, Streptococcus agalactiae, Enterococcus faecalis, and Escherichia coli, and is often accompanied by inflammation [6]. Clinical guidelines recommend various diagnostic methods for vaginal infections, including microscopy, culture-based assays, point-of-care (POC) tests, and nucleic acid amplification tests (NAATs) [7, 8]. However, due to time constraints, cost considerations, and the need for simultaneous detection of all four major types of vaginitis, microscopic examination of vaginal secretions (including saline wet mount and Gram staining) remains a widely used diagnostic tool globally, especially in resource-limited settings [9]. However, these methods have several limitations: (1) Saline wet mount lacks staining, resulting in low contrast between cells and microorganisms, significant background interference, and difficulty in identifying small pathogens [10], and inherently low diagnostic sensitivity; (2) Gram staining is a simple and widely adopted technique; however, its diagnostic performance for T. vaginalis is poor [11]; (3) Both methods have poor timeliness as they cannot effectively analyze samples with prolonged ex vivo time, often leading to missed diagnoses due to the death of T. vaginalis [12]. To address these challenges, our study evaluated a dual-fluorescence assay technique and assessed its performance in comparison with conventional microscopy. The staining system primarily consists of fluorescein isothiocyanate (FITC) and phycoerythrin, supplemented with additional dyes including eosin and methyl green. Under 450–500 nm blue light excitation, red fluorophores label microorganisms (including fungi, bacteria, and T. vaginalis), while green fluorophores lable host cells (epithelial cells and leukocytes). This distinct color contrast facilitates the rapid localization of potential pathogens, providing an enhanced visualization approach that may serve as an adjunct to conventional microscopic examination. Materials and methods Cell lines and bacterial strains The experimental materials included the vaginal epithelial cell line VK2/E6E7 (ATCC CRL-2616), Lactobacillus spp. (ATCC 7469), C. albicans (CMCC 98001), Gardnerella vaginalis (BNCC 337545), Staphylococcus aureus (ATCC 29213), Streptococcus agalactiae (ATCC 13813), Enterococcus faecalis (ATCC 33186), Escherichia coli (ATCC 25922), and T. vaginalis G3 strain (ATCC 50148™). Laboratory reagents The culture media and reagents used included: keratinocyte serum-free medium (Gibco, Grand Island, NY, USA), de Man, Rogosa and Sharpe (MRS) broth (Becton Dickinson, Franklin Lakes, NJ, USA), Sabouraud dextrose agar (Roche, Basel, Switzerland), Mueller–Hinton agar (MHA) (bioMérieux, Marcy-l’Étoile, France), trypticase–yeast extract–maltose (TYM) medium (prepared in-house according to Diamond’s protocol), Mueller–Hinton broth (MHB) powder (Solarbio, Beijing, China), Gram stain kit (Zhuhai Baso, Guangdong, China), and dual-fluorescence assay solution (Medomics, Jiangsu, China). Laboratory instruments The laboratory equipment comprised: a bacterial incubator (Yamato, Tokyo, Japan), a biological safety cabinet (Thermo Fisher Scientific, Waltham, MA, USA), a phase contrast microscope (OLYMPUS BX53, Tokyo, Japan), an automated Gram staining machine (Zhuhai Baso, Guangdong, China), an automated dual-fluorescence assay machine, and a fluorescent biological microscope MF23-M (Guangzhou Mshot, Guangzhou, China). Clinical specimens A standardized questionnaire was used to collect clinical information from study participants, including demographic characteristics, age, obstetric and gynecologic history, menstrual history, vaginitis-related symptoms, medication history, comorbidities, and findings from gynecological examinations. Participants were consecutively recruited from patients undergoing vaginal swab testing at the Gynecology Outpatient Department of Tianjin Medical University General Hospital between December 2024 and March 2025. During this period, all patients who met the inclusion criteria and had a clinical indication for vaginal swab collection were invited to participate, thereby minimizing selection bias. The enrolled population included both symptomatic individuals (e.g., those presenting with vaginitis symptoms) and asymptomatic individuals (e.g., those undergoing routine preoperative or prenatal screening). All participants were subjected to both conventional microscopy (saline wet mount and Gram staining) and the dual-fluorescence assay. This study was approved by the Ethics Committee of Tianjin Medical University General Hospital (Approval No. IRB2022-WZ-115). Inclusion criteria: (1) Patients deemed by clinicians to require vaginal swab examinations; (2) Ability to provide informed consent (for adults) or assent with guardian consent (for minors). Exclusion criteria: (1) Recent sexual activity within 3 days, vaginal douching, or vaginal medication use within 1 week prior to enrollment; (2) Menstruation or lactation within 8 weeks postpartum; (3) Acute cervicitis, pelvic inflammatory disease, or unexplained cervical/vaginal bleeding; (4) Vulvar dermatoses including vulvar leukoplakia and lichen sclerosus; (5) Severe systemic medical conditions or history of malignancy. The flowchart of participant screening, enrollment, and analysis is shown in Fig. 1. Flowchart of participant screening, enrollment, and analysis Sample size estimation was based on the “target sensitivity” approach for diagnostic accuracy studies. The anticipated prevalence rates at our center were 20.67% for VVC, 25.36% for BV, 6.04% for AV, and 1.05% for TV. Given that BV had the highest expected prevalence, it was selected as the driving condition. Assuming a target sensitivity of 90%, a margin of error of 5%, and a two-sided α level of 0.05, the formula n_event = Z² × P × (1 − P) / δ² yielded 139 required positive cases, corresponding to a total sample size of 548. To compensate for invalid specimens caused by poor specimen quality, staining failure and unevaluable smears during experimental detection, we increased the calculated sample size by an extra 20% as the reserved dropout allowance, resulting in a target sample size of 658. A total of 770 participants were initially recruited. Twelve samples were excluded due to insufficient volume or poor fluorescent signal post-staining, resulting in a final analyzable cohort of 758 participants. This exceeds the target sample size by 15% and ensures adequate statistical power to evaluate the four primary conditions: BV, VVC, AV, and TV. Experimental methods The dual-fluorescence assay solution consists of two components: Solution A and Solution B. Solution A serves as the staining reagent and contains fluorescent components (FITC, PI, Fruit Green, and PerCP), phosphate buffer salts (anhydrous sodium dihydrogen phosphate and sodium dihydrogen phosphate dodecahydrate), isothiazolinone preservatives, and purified water; Solution B serves as a stabilizing and pH-balancing buffer and contains the same phosphate salts, sodium hydroxide (pH adjuster), glycerol (stabilizer), isothiazolinone preservatives, and purified water. The differential staining mechanism of the dual-fluorescence system relies on the differences in membrane permeability and cellular composition between microorganisms and human host cells. Among the fluorescent ingredients, propidium iodide (PI) serves as a classic nucleic acid intercalating dye, which can significantly enhance its fluorescence intensity after binding to nucleic acids. After air-drying and heat fixation, bacteria, fungi, and T. vaginalis exhibit increased membrane permeability. Owing to their compact cellular structure and high nucleic acid density, these microorganisms allow efficient penetration and strong binding of PI to intracellular DNA and RNA, thereby presenting bright red fluorescence. In comparison, FITC and Fruit Green are green-emitting fluorophores with high affinity for cytoplasmic proteins and cellular structural components. Human vaginal epithelial cells and leukocytes are large in size and rich in cytoplasmic proteins, providing abundant binding sites for green fluorescent dyes. These green fluorophores preferentially accumulate in host cells, endowing host cells with dominant green fluorescence signals. Notably, the final fluorescent signal is a composite result affected by membrane integrity, nucleic acid conformation, and cellular metabolic activity, which inevitably causes slight differences in staining intensity among different samples. Therefore, red-fluorescent pathogens and green-stained human host cells can be rapidly distinguished via dual fluorescence, which aids quick pathogen localization and minimizes missed diagnoses. Nevertheless, final microbial identification is determined by typical morphological traits, including bacterial rod/coccus morphology as well as fungal hyphae and spores. The dual-fluorescence assay solution was applied for imaging standardized vaginal epithelial cell lines and reference vaginal microbial strains: cryopreserved VK2/E6E7 vaginal epithelial cells were inoculated in keratinocyte serum-free medium and incubated under standard cell culture conditions. Reference microbial strains, including Lactobacillus vaginalis in MRS broth, C. albicans in Sabouraud dextrose agar, Gardnerella vaginalis on surface-dried Columbia blood agar plates, and Staphylococcus aureus, Streptococcus agalactiae, and Escherichia coli on Mueller-Hinton agar were cultured overnight at 37℃ with 5% CO2 in a bacterial incubator. Separately, T. vaginalis G3 strain was cultivated axenically in TYM medium supplemented with 10% heat-inactivated horse serum at 37℃ in a humidified incubator without CO2 for 48–72 h until mid-log phase. Suspensions of all microorganisms were then prepared, with one droplet spread on glass slides, heat-fixed, cooled, and sequentially stained with dual-fluorescence Solutions A and B (1 min each) before microscopic observation and imaging under a fluorescence biological microscope. Vaginal secretion samples were collected from the upper third of the vaginal sidewall using sterile swabs after cervical exposure with a speculum. Samples were used to prepare Gram-stained smears, saline wet mounts, and dual-fluorescence stained slides according to routine laboratory procedures. For dual-fluorescence microscopy, specimens were eluted in 500 µL of sterile saline, and the diluted sample was spread over approximately 4 cm², heat-fixed, cooled, and sequentially stained with Solutions A and B (1 min each) before examination under a fluorescence microscope. Slides were initially screened at 100× magnification to assess sample adequacy and identify regions of interest, followed by detailed examination at 400× magnification for the detection of fluorescently labeled microorganisms and cellular components. We employed routine vaginal discharge microscopy (saline wet mount and Gram staining) as the reference method to validate the diagnostic capability of the dual-fluorescence assay. The corresponding diagnostic criteria for various vaginitis types are described in detail below. For BV diagnosis, the Hay-Ison criteria were applied. This standard is based on a semi-quantitative grading system of Gram-stained smear morphology and does not require precise absolute counts of individual microorganisms. The diagnosis relies on the observed qualitative shift in the vaginal flora (from a Lactobacillus-dominated microbiota to a mixed flora or one dominated by Gardnerella morphotypes) [13]. For VVC diagnosis, Gram-stained microscopy was used, defined by the presence of budding yeast, hyphae, or pseudohyphae [7]. For AV diagnosis, the Donders AV scoring system [14] was employed, assessing lactobacillary grades (LBG), leukocyte count, toxic granulation-containing leukocytes/leukocyte ratio, background flora, and parabasal epithelial cells (PBC)/epithelial cell ratio in saline wet mounts, with a cumulative score ≥ 3 indicating AV. For TV diagnosis, both guidelines [7, 13] affirm the utility of wet mount microscopy as the most cost-effective and widely used method. A key drawback of this technique is that microscopic sensitivity drops to 20% if specimens are not examined within 1 h, this limitation can be alleviated via alternative sample-handling strategies. In this study, we implemented combined warm specimen preservation and point-of-care microscopic testing to avoid this pitfall and mitigate time-dependent sensitivity decay. All microscopic evaluations were performed independently by two trained observers who were blinded to clinical information. In cases of discordance, a third senior expert conducted a blinded assessment, and their judgment was taken as the final determination. PCR-based NAATs and culture were not used as reference standards because the objective of this study was to compare the dual-fluorescence assay directly with routine laboratory microscopy, which—despite its known limitations in sensitivity—remains the predominant diagnostic approach in real-world outpatient practice. Statistical methods All data were analyzed using SPSS 20.0 (IBM Corp., Armonk, NY, USA). Quantitative variables were expressed as mean±standard deviation (x̄±s) and categorical variables as number (percentage). For diagnostic accuracy analyses, 2 × 2 contingency tables were constructed using conventional microscopy (Gram-stained smears and saline wet mounts, according to contemporary international guidelines) as the reference method for each type of vaginitis (BV, VVC, AV, and TV). Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and Youden’s index were calculated. The 95% confidence intervals (95% CIs) for sensitivity, specificity, PPV, and NPV were obtained using the exact (Clopper–Pearson) method. Agreement between dual-fluorescence assay and the reference microscopy-based methods was assessed using Cohen’s Kappa (κ), and 95% CIs for κ were calculated based on the standard error of κ. κ values were interpreted as follows: κ < 0, poor agreement; 0.00–0.20, slight; 0.21–0.40, fair; 0.41–0.60, moderate; 0.61–0.80, substantial; and 0.81–1.00, almost perfect agreement. Intergroup comparisons of categorical variables were performed using the chi-square test (or Fisher’s exact test, where appropriate). All tests were two-sided, and a P-value < 0.05 was considered statistically significant. Result Dual fluorescent stains for imaging human vaginal epithelial cells and labeled bacterial species in the vaginal environment Fluorescence microscopy was performed using blue light excitation (450–500 nm) to visualize dual-fluorescence assay human vaginal epithelial VK2/E6E7 cells with representative microbial species, including Lactobacillus spp., Gardnerella vaginalis, Staphylococcus aureus, Streptococcus agalactiae, Escherichia coli, and T. vaginalis (Fig. 2). The selected microorganisms were chosen for their representative roles in the vaginal microbiome: Lactobacillus species are the dominant members of the healthy vaginal flora, whereas the other organisms are commonly associated with vaginitis and hold significant clinical relevance. The morphological characteristics of these cells and microorganisms under dual-fluorescence assay are as follows: Vaginal epithelial VK2/E6E7 cells: Large, polygonal, flat squamous epithelial cells with well-defined cell borders; Lactobacillus spp.: Slender, rod-shaped (bacilli); Gardnerella vaginalis: Small bacilli or coccobacilli with a slightly curved, short rod-shaped morphology; Escherichia coli: Short, straight rods with rounded ends, occurring singly or in pairs; Enterococcus faecalis: Spherical or ovoid cells, usually arranged in pairs or short chains; Staphylococcus aureus: Spherical cells, appearing singly or in characteristic irregular clusters resembling grape bunches; Streptococcus agalactiae: Spherical or ovoid cells that form long chains; C. albicans: Oval yeast cells with visible budding projections; T. vaginalis: Pear-shaped or ovoid protozoa, significantly larger in size than bacterial cells. Fluorescent staining of cultured vaginal epithelial cells (VK2/E6E7) and standard strains of vaginal microbial species with dual dyes. Images were acquired on a fluorescence microscope (blue excitation 450-500 nm, 40×objective, total magnification 400×). a Vaginal epithelial VK2/E6E7 cells; b Lactobacillus spp.; c Gardnerella vaginalis; d Escherichia coli; e Enterococcus faecalis; f Staphylococcus aureus; g Streptococcus agalactiae; h C. albicans; i T. vaginalis This dual-fluorescence assay produced a consistent two-channel contrast in vaginal smears: host epithelial cells and leukocytes were predominantly green fluorescent, whereas microorganisms (including bacteria, fungi, and T. vaginalis) were predominantly red fluorescent. Mixed or intermediate fluorescence signals were occasionally observed. Therefore, fluorescence was used for rapid localization and as an auxiliary cue, while definitive identification was based on established morphological criteria (e.g., bacterial morphotypes/arrangement, fungal hyphae/spores, and morphology of T. vaginalis). Validation of dual fluorescent staining solution for clinical samples Description of Clinical Data for Patients: the baseline demographic and clinical characteristics of the 758 enrolled participants are summarized in Table 1. Consecutive recruitment of patients clinically indicated for a vaginal swab yielded a cohort with a mean age of 38.59 ± 6.76 years, encompassing both symptomatic individuals and those undergoing routine screening. Specifically, 609 patients (80.3%) presented with at least one symptom or sign suggestive of vaginitis, while 149 (19.7%) were asymptomatic at enrollment. Clinically, the most common reasons for seeking medical attention were itching (422 cases, 55.78%), increased vaginal discharge (368 cases, 48.53%), yellow or green discharge (210 cases, 27.66%), dyspareunia (163 cases, 21.54%), burning pain (103 cases, 13.59%), contact bleeding (82 cases, 10.88%), and foul-smelling discharge (22 cases, 2.94%). Notably, 149 cases (19.73%) presented with no symptoms or signs. Among the 609 symptomatic patients, 180 (29.56%) reported only one symptom, whereas 429 (70.44%) exhibited two or more concurrent symptoms, highlighting the frequent coexistence of mixed clinical manifestations in real-world vaginitis presentations. (see Table 1). Table 1Demographic and clinical characteristics of participantsCharacteristicn (%) or Mean ± SDAge (years)38.59 ± 6.76Range12 ~ 75Age Group Distribution < 18 years5 (0.66%) 18–45 years654 (86.28%) > 45 years99 (13.06%)Reason for the visit Symptomatic609 (80.27%) Asymptomatic (routine screening)149 (19.7%)Presenting Symptoms Itching55.78% (422/758) Increased vaginal discharge48.53% (368/758) Yellow or green discharge27.66% (210/758) Dyspareunia21.54% (163/758) Burning pain13.61% (103/758) Contact bleeding10.88% (82/758) Foul-smelling discharge2.95% (22/758)Symptom count (n = 609) Only 1 symptom180 (29.56%) ≥2 symptoms429 (70.44%)As patients may have multiple symptoms, each symptom was grouped independently, allowing overlap in statistical counts Demographic and clinical characteristics of participants As patients may have multiple symptoms, each symptom was grouped independently, allowing overlap in statistical counts Imaging manifestations of dual fluorescent staining in different types of vaginitis The dual-fluorescence assay performs well in experimental settings by using a red dye to target microorganisms and a green dye to highlight cell morphology, allowing rapid localization of microbes in vaginal cells and cultured bacterial specimens. However, the natural vaginal environment is more complex, with microorganisms coexisting and interacting in ways that differ from laboratory conditions. Clinical trials are therefore indispensable to validate its real-world applicability. The imaging characteristics of the dual-fluorescence assay and conventional microscopy (saline wet mount and Gram staining) in various types of vaginitis are presented in Fig. 3. VVC: the dual-fluorescence assay highlights yeast cells and pseudohyphae as bright red structures against green host cells, improving contrast and pathogen detection. Conventional methods (Gram/wet mount) may miss low-burden or atypical infections due to poor resolution. BV: loss of Lactobacillus and overgrowth of anaerobic coccobacilli, clearly visualized in red with fluorescence assay. Gram stain shows similar detail; wet mounts lack resolution due to small organism size and low contrast. TV: large organisms with irregular red fluorescence. Saline wet mount remains the reference method (based on motility), but requires prompt evaluation. Gram stain is insensitive and not recommended. AV: Diagnosed using a scoring system based on Lactobacillus dominance, leukocyte count, presence of parabasal cells, background flora, and toxic leukocytes. The dual-fluorescence assay enables reliable assessment of the first four parameters, with performance comparable to Gram staining. Although wet-mount microscopy allows evaluation of all five criteria, it provides suboptimal resolution for microbial morphology and bacterial grading. Together, these two dyes form a multidimensional fluorescent signal system through differential labeling strategies, synergistically enabling the accurate discrimination of host cells and microorganisms in complex samples. Dual-fluorescence assay overcomes the limitations of saline wet mounts, which lack staining effects, exhibit unclear backgrounds, and make it difficult to identify small microbes, while also addressing the complexity of the Gram staining procedure and its inability to reliably detect T. vaginalis. Therefore, dual-fluorescence assay combines the advantages of both saline wet mounts and Gram staining. Comparison of dual-fluorescence assay and conventional wet-mount microscopy in diagnosing vaginitis Comparison of dual fluorescent staining and routine vaginal discharge microscopy results In line with our primary aim of comparing a novel laboratory assay with current routine laboratory practice, we deliberately used microscopy-based criteria as the reference method. According to the aforementioned diagnostic criteria for vaginitis, mixed vaginitis cases were individually counted under each respective category due to the possibility of co-infections [15, 16]. The laboratory diagnostic results of vaginal discharge microscopy and dual-fluorescence assay are presented in Table 2, 3, 4 and 5. Based on the data, we calculated the diagnostic sensitivity, specificity, and Cohen’s kappa coefficient for agreement of the dual fluorescence assay against the respective reference methods for each type of vaginitis. Detailed performance metrics are provided in Table 6. For the diagnosis of BV, the dual-fluorescence assay method demonstrated a sensitivity of 0.886 (95% CI:0.796–0.955), a specificity of 0.958 (95% CI:0.931–0.979), and almost perfect agreement with Gram staining (κ = 0.821; 95% CI:0.756–0.901). In the diagnosis of VVC, the assay achieved a sensitivity of 0.953 (95% CI:0.840–0.988) and a specificity of 0.978 (95% CI:0.953–0.990), showing almost perfect agreement with Gram staining (κ = 0.908; 95% CI:0.847–0.963). Notably, the dual fluorescence method identified yeast cells in seven samples that were negative by Gram staining. For TV, the dual-fluorescence assay exhibited a sensitivity of 0.833 (95% CI:0.365–0.991) and a specificity of 0.995 (95% CI:0.979–0.999), with substantial agreement with wet-mount microscopy (κ = 0.766; 95% CI: 0.331–1.000). It detected T. vaginalis in four samples that were negative by wet-mount microscopy. Regarding AV, the method showed a sensitivity of 0.800 (95% CI:0.535–0.848), a specificity of 0.959 (95% CI:0.924–0.995), and substantial agreement with wet-mount microscopy (κ = 0.697; 95% CI: 0.616–0.863). Table 2BV diagnosis: Dual-Fluorescence Assay vs. Gram StainingDual-Fluorescence Assay(BV)Gram PositiveGram NegativeTotalPositive12426150Negative16592608Total140618758BV was diagnosed employing the Hay-Ison criteria, wherein a Grade 2 (intermediate) finding was included in the negative group for the purpose of statistical analysis BV diagnosis: Dual-Fluorescence Assay vs. Gram Staining BV was diagnosed employing the Hay-Ison criteria, wherein a Grade 2 (intermediate) finding was included in the negative group for the purpose of statistical analysis Table 3VVC diagnosis: Dual-Fluorescence Assay vs. Gram StainingDual-Fluorescence Assay(VVC)Gram PositiveGram NegativeTotalPositive12214136Negative6616622Total128630758 VVC diagnosis: Dual-Fluorescence Assay vs. Gram Staining Table 4AV diagnosis: Dual-Fluorescence Assay vs. Wet Mount MicroscopyDual-Fluorescence Assay(AV)Wet Mount PositiveWet Mount NegativeTotalPositive562884Negative14660674Total70688758 AV diagnosis: Dual-Fluorescence Assay vs. Wet Mount Microscopy Table 5TV diagnosis: Dual-Fluorescence Assay vs. Wet Mount MicroscopyDual-Fluorescence Assay(TV)Wet Mount PositiveWet Mount NegativeTotalPositive10414Negative2742744Total12746758 TV diagnosis: Dual-Fluorescence Assay vs. Wet Mount Microscopy Table 6Clinical Performance of the Dual-Fluorescence Assay for VaginitisSensitivity (95%CI)Specificity (95%CI)PPV (95%CI)NPV (95%CI)Kappa Value (95%CI) BV 0.886 (0.796–0.955)0.958 (0.931–0.979)0.827 (0.730–0.910)0.973 (0.951–0.990)0.821 (0.756–0.901) VVC 0.953 (0.840–0.988)0.978 (0.953–0.990)0.897 (0.793–0.954)0.990 (0.970–0.998)0.908 (0.847–0.963) AV 0.800 (0.535–0.848)0.959 (0.924–0.995)0.667 (0.645–0.937)0.979 (0.946–0.985)0.697 (0.616–0.863) TV 0.833 (0.365–0.991)0.995 (0.979–0.999)0.714 (0.303–0.949)0.997 (0.983-1.000)0.766 (0.331-1.000)PPV: Positive predictive value; NPV: Negative predictive value;Kappa values are interpreted as follows: 0.00-0.20, slight agreement; 0.21–0.40, fair agreement; 0.41–0.60, moderate agreement; 0.61–0.80, substantial agreement; 0.81-1.00, almost perfect agreement Clinical Performance of the Dual-Fluorescence Assay for Vaginitis PPV: Positive predictive value; NPV: Negative predictive value; Kappa values are interpreted as follows: 0.00-0.20, slight agreement; 0.21–0.40, fair agreement; 0.41–0.60, moderate agreement; 0.61–0.80, substantial agreement; 0.81-1.00, almost perfect agreement Discussion Vaginitis is the most common reproductive tract infection among women, with over 300 million cases reported globally each year [17]. Current clinical guidelines recommend various diagnostic methods for vaginitis [2, 18, 19], but due to limitations in time and cost, these approaches still fail to fully meet clinical needs. Accurate diagnosis is crucial for the prevention and treatment of vaginitis. Therefore, this study aims to develop a novel detection method, and evaluate its diagnostic performance by comparison with conventional microscopy. Diagnostic accuracy is a pivotal determinant in the clinical translation of novel detection methods. Our clinical validation verified that the dual fluorescence assay performs excellently for BV and VVC, with nearly perfect consistency with Gram staining. Its higher analytical sensitivity is further supported by the detection of seven yeast-positive samples missed by Gram staining, indicating strong capability for identifying low-abundance fungal pathogens. This favorable performance is attributed to the high nucleic acid affinity of fluorescent dyes and the distinct color contrast, which together enhance signal clarity and minimize background interference. For TV diagnosis, the dual-fluorescence assay achieved a higher detection rate than wet-mount microscopy. It identified T. vaginalis in four wet-mount-negative samples. Wet-mount microscopy only detects motile parasites and is prone to false negatives for non-viable organisms. Fluorescent nucleic acid staining can detect residual microbial nucleic acids and sustain diagnostic performance for specimens with compromised pathogen viability, which is consistent with the findings of Radonjic et al. [20]. These additional positives in both TV and VVC likely reflect the method’s higher analytical sensitivity, though their true clinical relevance—actual infection versus false positives—still requires confirmation by reference methods such as NAAT or culture. For AV detection, the dual-fluorescence assay showed high specificity (0.959) but moderate sensitivity (0.800). The reduced sensitivity mainly reflects intrinsic limitations of the staining and scoring process. First, the method highlights nuclei and bacteria rather than cytoplasmic changes, making toxic granules less visible and lowering AV composite scores. Second, borderline cases near the diagnostic cutoff are prone to small scoring shifts, and the superior visualization of Lactobacillus and background flora under fluorescence may lead to discrepancies with wet-mount microscopy. Third, AV relies on a multi-parameter scoring system that is inherently susceptible to observer variation. Overall, despite moderate sensitivity, the assay’s high specificity and substantial agreement with routine microscopy support its utility as a reliable adjunct tool, consistent with findings by Dong et al. [21]. Notably, our performance evaluations rely on conventional microscopy as the comparator, which brings inherent study limitations. Conventional microscopy, although guideline-endorsed and widely used, has suboptimal sensitivity—particularly for low-abundance yeasts and non-viable T. vaginalis. The absence of systematic culture or NAAT confirmation means that misclassification inherent to routine microscopy may influence the apparent performance of the dual-fluorescence assay. Therefore, our results primarily reflect concordance with current clinical practice rather than absolute accuracy against a composite gold standard. Despite these constraints, the dual‑fluorescence assay demonstrates strong agreement with routine microscopy and remains a promising adjunct tool for vaginal infection diagnosis. In addition to diagnostic accuracy, testing turnaround time is a critical factor influencing clinical decision-making. The dual fluorescence assay offers high efficiency, requiring only a two-minute staining step and rapid whole-slide screening at 400×, while its dual-color labeling provides clear chromatic contrast that enables fast and accurate pathogen identification, improving throughput for high-volume outpatient testing. This assay also provides several technical advantages in clinical practice. First, it shows strong anti-interference performance: nucleus-free red blood cells remain unstained, reducing background noise in hemorrhagic samples and improving diagnostic clarity. Second, it tolerates pre-analytical delays by detecting non-viable T. vaginalis, overcoming the major limitation of wet-mount microscopy, which fails once parasite motility is lost. Third, this single-stain, multi-target approach supports simultaneous detection of multiple vaginitis subtypes and mixed infections, offering broader diagnostic coverage. Technologists familiar with routine smear microscopy can quickly master the assay after brief training, and standardized procedures further enhance diagnostic consistency and laboratory workflow efficiency. Despite its promising clinical performance, the dual-fluorescence assay has several technical limitations for widespread routine use. First, it relies on specialized fluorescence microscopes, which are unavailable in most primary care facilities. The associated equipment cost and moderate learning curve restrict its application in resource-limited regions. Second, fluorophore photobleaching requires slide interpretation within 30 min, prohibiting slide archiving, retrospective analysis and repeat consultation. Third, its reagent cost is twice that of conventional Gram staining, though still 30% lower than standard PCR assays locally. Fourth, non-specific red fluorescence labeling cannot distinguish pathogenic from commensal microbes, relying on manual morphological identification for final diagnosis. Notwithstanding the above limitations, the dual-fluorescence assay integrates the complementary strengths of Gram staining and wet-mount microscopy. Compared with conventional microscopy, the assay maintains high diagnostic consistency and exhibits superior sensitivity for low-abundance and non-viable pathogens, making it a reliable first-line adjunct screening tool in outpatient and emergency settings to reduce diagnostic errors. The assay’s high-contrast dual-color imaging provides clear separation between host cells and pathogens, generating distinct morphological features that support auxiliary AI-assisted analysis. This visual advantage reduces subjective interpretation variability and enhances detection of low-abundance fungal pathogens. However, AI development is still limited by insufficient annotated datasets and the lack of standardized cross-device validation criteria. Further prospective studies and algorithm refinement are needed to achieve reliable automated diagnosis. Even so, the dual-fluorescence assay maintains strong potential for broader clinical adoption, offering meaningful value for vaginitis diagnosis and gynecological infection prevention as technology continues to advance. Supplementary Information Supplementary Material 1. Acknowledgements We would like to express our sincere gratitude to all the patients whose participation made this study possible. We are also deeply grateful to the medical staff of Tianjin Medical University General Hospital for their invaluable assistance in sample collection and processing. Additionally, we thank Medomics for providing the dual fluorescence staining kits and technical support. Finally, we acknowledge the financial support provided by the Tianjin Science and Technology Commission. Abbreviations Reproductive tract infections Bacterial vaginosis Aerobic vaginitis Vulvovaginal candidiasis Trichomoniasis point-of-care tests nucleic acid amplification tests fluorescein isothiocyanate Bacterial Vaginosis-Associated Bacterium 2 Authors’ contributions Wenhui Qi oversaw project design, data collection, manuscript drafting, and overall supervision; Liyao Wang optimized reagents and performed clinical specimen staining; Wenhui Qi and Liyao Wang contributed equally as co-first authors; Junyi Bai and Huanrong Li collected clinical cases; Jinfang Chen and Mengting Dong conducted laboratory testing; Fengxia Xue and Chen Wang determined the research direction and designed the study plan to ensure scientific rigor. Funding This work was supported by the Tianjin Science and Technology Planning Project (No. 22JCQNJC01530) and the Special Research Fund for Central Universities, Peking Union Medical College (No. 3332024105) and Tianjin Key Medical Discipline (Specialty) Construction Project(TJYXZDXK-3–007 C). Data availability The datasets supporting the conclusions of this article are included within the article. Declarations Ethics approval and consent to participate This study is an in vitro methodological comparison and belongs to non-clinical interventional laboratory research, which does not involve any prospective patient interventions, modifications to treatment regimens, or assessments of clinical outcomes. All procedures performed in this study involving human samples were conducted in accordance with the ethical standards outlined in the Declaration of Helsinki. This study was approved by the Ethics Committee of Tianjin Medical University General Hospital (Approval No. IRB2022-WZ-115). Written informed consent was obtained from all participants. Consent for publication Not applicable. No identifiable individual patient images or personal clinical data are presented in this manuscript. All authors have read and approved the final manuscript and agree to publication. Competing interests The authors declare no competing interests. Footnotes Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Contributor Information Chen Wang, Email: wangchener32@163.com. Fengxia Xue, Email: xuefengxia@tmu.edu.cn. References Associated Data Supplementary Materials Supplementary Material 1. Data Availability Statement The datasets supporting the conclusions of this article are included within the article.