An interpretation algorithm for molecular diagnosis of bacterial vaginosis in a maternity hospital using machine learning: proof-of-concept study

Molecular characterization of vaginal microbiota using a new 22-species qRT-PCR test to achieve a relative-abundance and species-based diagnosis of bacterial vaginosis

Background

Numerous bacteria are involved in the etiology of bacterial vaginosis (BV). Yet, current tests only focus on a select few. We therefore designed a new test targeting 22 BV-relevant species.

Methods

Using 946 stored vaginal samples, a new qPCR test that quantitatively identifies 22 bacterial species was designed. The distribution and relative abundance of each species, α- and β-diversities, correlation, and species co-existence were determined per sample. A diagnostic index was modeled from the data, trained, and tested to classify samples into BV-positive, BV-negative, or transitional BV.

Results

The qPCR test identified all 22 targeted species with 95 - 100% sensitivity and specificity within 8 hours (from sample reception). Across most samples, Lactobacillus iners, Lactobacillus crispatus, Lactobacillus jensenii, Gardnerella vaginalis, Fannyhessea (Atopobium) vaginae, Prevotella bivia, and Megasphaera sp. type 1 were relatively abundant. BVAB-1 was more abundant and distributed than BVAB-2 and BVAB-3. No Mycoplasma genitalium was found. The inter-sample similarity was very low, and correlations existed between key species, which were used to model, train, and test a diagnostic index: MDL-BV index. The MDL-BV index, using both species and relative abundance markers, classified samples into three vaginal microbiome states. Testing this index on our samples, 491 were BV-positive, 318 were BV-negative, and 137 were transitional BV. Although important differences in BV status were observed between different age groups, races, and pregnancy status, they were statistically insignificant.

Conclusion

Using a diverse and large number of vaginal samples from different races and age groups, including pregnant women, the new qRT-PCR test and MDL-BV index efficiently diagnosed BV within 8 hours (from sample reception), using 22 BV-associated species.

Vaginal microbiota molecular profiling and diagnostic performance of artificial intelligence-assisted multiplex PCR testing in women with bacterial vaginosis: a single-center experience

Background

Bacterial vaginosis (BV) is a most common microbiological syndrome. The use of molecular methods, such as multiplex real-time PCR (mPCR) and next-generation sequencing, has revolutionized our understanding of microbial communities. Here, we aimed to use a novel multiplex PCR test to evaluate the microbial composition and dominant lactobacilli in non-pregnant women with BV, and combined with machine learning algorithms to determine its diagnostic significance.

Methods

Residual material of 288 samples of vaginal secretions derived from the vagina from healthy women and BV patients that were sent for routine diagnostics was collected and subjected to the mPCR test. Subsequently, Decision tree (DT), random forest (RF), and support vector machine (SVM) hybrid diagnostic models were constructed and validated in a cohort of 99 women that included 74 BV patients and 25 healthy controls, and a separate cohort of 189 women comprising 75 BV patients, 30 intermediate vaginal microbiota subjects and 84 healthy controls, respectively.

Results

The rate or abundance of Lactobacillus crispatus and Lactobacillus jensenii were significantly reduced in BV-affected patients when compared with healthy women, while Lactobacillus iners, Gardnerella vaginalis, Atopobium vaginae, BVAB2, Megasphaera type 2, Prevotella bivia, and Mycoplasma hominis were significantly increased. Then the hybrid diagnostic models were constructed and validated by an independent cohort. The model constructed with support vector machine algorithm achieved excellent prediction performance (Area under curve: 0.969, sensitivity: 90.4%, specificity: 96.1%). Moreover, for subjects with a Nugent score of 4 to 6, the SVM-BV model might be more robust and sensitive than the Nugent scoring method.

Conclusion

The application of this mPCR test can be effectively used in key vaginal microbiota evaluation in women with BV, intermediate vaginal microbiota, and healthy women. In addition, this test may be used as an alternative to the clinical examination and Nugent scoring method in diagnosing BV.

The evaluation of the Allplex™ BV molecular assay for the diagnosis of bacterial vaginosis in symptomatic South African females

Bacterial vaginosis (BV) is a dysbiosis of vaginal microbiota characterized by a shift from Lactobacillus species predomination to a heterogeneous mixture of anaerobes. We compared the performance characteristics of the Allplex ™ BV molecular assay with the reference test, Nugent score microscopy, for vaginal swab specimens from symptomatic South African women. A total of 213 patients were enrolled, of whom 99 (46.5%) and 132 (62.0%) were diagnosed with BV by Nugent and Allplex™, respectively. The Allplex™ BV assay displayed a sensitivity of 94.9% (95% CI, 88.7%-97.8%) and a specificity of 66.7% (95% CI, 57.6%-74.6%), with an agreement of 79.8% (95% CI, 73.9%-84.7%) (κ = 0.60). Assay design may be enhanced for improved specificity by accounting for differences in healthy and BV-associated vaginal microbiomes among women of different ethnicities.

Bacterial Vaginosis and Associated Factors Among Pregnant Women Attending Antenatal Care in Harar City, Eastern Ethiopia

Background

Bacterial vaginosis is one of the most common causes of abnormal vaginal discharge in women of reproductive age, especially pregnant women. It is characterized by the decrement of lactobacilli and increasing proliferation of diverse anaerobic bacteria. The prevalence of bacterial vaginosis and contributing factors vary between countries, within the same country, and among different populations.

Objective

The purpose of this study was to determine the prevalence of bacterial vaginosis and associated factors among pregnant women attending antenatal care from May 5-July 15, 2021 in health-care facilities found in Harar town, eastern Ethiopia.

Methods and Materials

An institutional-based cross-sectional study was conducted among 248 pregnant women attending antenatal care through systematic random sampling. Data were collected using a structured questionnaire and two vaginal swabs were collected using sterile swabs and transported to the microbiology laboratory. Samples were analyzed using Amsel's criteria or using Nugent scoring criteria and culture. Data were entered into EpiData version 3.1 and transferred to SPSS version 25 for analysis. Binary logistic regression was used to identify variables associated with bacterial vaginosis. P-value <0.05 was considered statistically significant.

Results

Altogether 248 pregnant women were included in the study. Overall, 21.4% (95% CI: 16, 27) of study participants had bacterial vaginosis. History of sexually transmitted infection (AOR = 6.0, 95% CI: 1.94, 19.07; P = 0.002), history of spontaneous abortion (AOR = 5.8, 95% CI: 1.55, 22.02; P = 0.009), multiple sex partners (AOR = 8.6, 95% CI: 2.93, 25.79; P = 0.000) and having vaginal discharge (AOR = 5.5, 95% CI: 2.48, 12.41; P = 0.000) were significantly associated with bacterial vaginosis.

Conclusion

The prevalence of bacterial vaginosis is higher among symptomatic pregnant women and associated with a history of sexually transmitted infection, vaginal discharge, multiple sexual partners and spontaneous abortion.

The Female Vaginal Microbiome in Health and Bacterial Vaginosis

The vaginal microbiome is an intricate and dynamic microecosystem that constantly undergoes fluctuations during the female menstrual cycle and the woman's entire life. A healthy vaginal microbiome is dominated by Lactobacillus which produce various antimicrobial compounds. Bacterial vaginosis (BV) is characterized by the loss or sharp decline in the total number of Lactobacillus and a corresponding marked increase in the concentration of anaerobic microbes. BV is a highly prevalent disorder of the vaginal microbiota among women of reproductive age globally. BV is confirmed to be associated with adverse gynecologic and obstetric outcomes, such as sexually transmitted infections, pelvic inflammatory disease, and preterm birth. Gardnerella vaginalis is the most common microorganism identified from BV. It is the predominant microbe in polymicrobial biofilms that could shelter G. vaginalis and other BV-associated microbes from adverse host environments. Many efforts have been made to increase our understanding of the vaginal microbiome in health and BV. Thus, improved novel and accurate diagnosis and therapeutic strategies for BV have been developed. This review covers the features of vaginal microbiome, BV, BV-associated diseases, and various strategies of diagnosis and treatment of BV, with an emphasis on recent research progresses.

Clinical validation of a new molecular test (Seegene Allplex™ Vaginitis) for the diagnosis of vaginitis: a cross-sectional study

OBJECTIVE

To validate the use of Seegene Allplex™ Vaginitis assay in the diagnosis of candidiasis, bacterial vaginosis (BV) and trichomoniasis.

DESIGN

Cross-sectional, prospective study conducted in a single centre.

SETTING

Outpatient clinic of a gynaecology department.

POPULATION

Consecutive symptomatic and asymptomatic women (18-60 years of age).

METHODS

Comparison of the assay test with the reference standards for the diagnosis of vaginitis (cultures for yeasts, Nugent for BV and nucleic acid amplification test for trichomoniasis).

MAIN OUTCOME MEASURES

Performance of the investigational assay, in comparison with the reference standards for the diagnosis of the presence of Candida spp., Trichomonas vaginalis and BV. Secondary objectives are the evaluation of the performance of the test in postmenopausal women and in symptomatic women.

RESULTS

A diagnosis of vaginitis was established in 14.0%. The global prevalences of BV, Candida spp. and T. vaginalis were 22.3%, 13.2% and 2.4%, respectively. The sensitivity and specificity of the assay test for those three causes of vaginitis were as follows: BV 91.7% and 86.6%; any Candida spp. 91.1% and 95.6%; Candida albicans 88.1% and 98.2%, non-albicans Candida 100% and 97.5%, and T. vaginalis 94.4 and 99.9%. The performance of the test was identical in the subgroup of women that reported vulvovaginal symptoms. The presence of multiple infections did not interfere with the performance of the test.

CONCLUSIONS

The Seegene Allplex™ Vaginitis assay has an excellent performance in the diagnosis of the BV and presence of Candida; the results were good for trichomoniasis, but the study was underpowered for this outcome.

TWEETABLE ABSTRACT

Seegene Allplex™ Vaginitis is an excellent option for screening and diagnosis of vaginitis.

Recent advances in cultivation-independent molecular-based techniques for the characterization of vaginal eubiosis and dysbiosis

“The bacterial vaginosis syndrome” has significant adverse effects for women and babies, including preterm birth and increased risk of acquisition of sexually transmitted infections and HIV. Currently, the gold standard for diagnosis is Gram stain microscopy of vaginal secretions, which is not readily available, is somewhat subjective, and does not differentiate between the likely different subtypes of vaginal dysbioses that may have different etiologies, microbiology, responses to antibiotics, and phenotypic outcomes. With new information from molecular-based, cultivation-independent studies, there is increasing interest in the use of molecular techniques for the diagnosis of bacterial vaginosis. We reviewed the current evidence on and the rationale behind the use of molecular techniques for the diagnosis of bacterial vaginosis. We found a number of commercially available molecular diagnostic tests, a few of which have US Food and Drug Administration (FDA) and/or Conformité Européenne in vitro diagnostic (CE-IVD) approval, and we have compared their performance with respect to sensitivities and specificities. Molecular-based tests have the advantage of objectivity, quantification, detection of fastidious organisms, and validity for self-obtained vaginal swabs. The performance of the molecular tests against standard microscopy is impressive, but further education of users on interpretation is needed. Bacterial vaginosis is the major cause of vaginal dysbiosis and should be recognized for the threat it is to women’s genital tract health. Quantitative assessment of microbial abundance, the diversity of other organisms present, specific primers for gene sequence regions, and clades and biovars of target microbes should be recognized and incorporated into future molecular diagnostic tests to better differentiate between vaginal eubiosis and dysbiosis.

Using Artificial Intelligence in Infection Prevention

Purpose of Review

Artificial intelligence (AI) offers huge potential in infection prevention and control (IPC). We explore its potential IPC benefits in epidemiology, laboratory infection diagnosis, and hand hygiene.

Recent Findings

AI has the potential to detect transmission events during outbreaks or predict high-risk patients, enabling development of tailored IPC interventions. AI offers opportunities to enhance diagnostics with objective pattern recognition, standardize the diagnosis of infections with IPC implications, and facilitate the dissemination of IPC expertise. AI hand hygiene applications can deliver behavior change, though it requires further evaluation in different clinical settings. However, staff can become dependent on automatic reminders, and performance returns to baseline if feedback is removed.

Summary

Advantages for IPC include speed, consistency, and capability of handling infinitely large datasets. However, many challenges remain; improving the availability of high-quality representative datasets and consideration of biases within preexisting databases are important challenges for future developments. AI in itself will not improve IPC; this requires culture and behavior change. Most studies to date assess performance retrospectively so there is a need for prospective evaluation in the real-life, often chaotic, clinical setting. Close collaboration with IPC experts to interpret outputs and ensure clinical relevance is essential.