Bacterial Vaginosis: Guideline of the DGGG, OEGGG and SGGG (S2k-Level, AWMF Registry No. 015/028, June 2023)

Abstract

Aim This official guideline was coordinated and published by the DGGG, OEGGG and SGGG with the involvement of additional professional societies. The aim of the guideline is to evaluate the relevant literature and use it to provide a consensus-based overview of the diagnosis and management of bacterial vaginosis.

Methods This S2k-guideline was developed by representative members from different medical professional societies on behalf of the guidelines commission of the above-listed societies using a structured consensus process.

Recommendations This guideline provides recommendations on the diagnosis, management, counselling, prophylaxis, and other aspects related to bacterial vaginosis.

I  Guideline Information

Guidelines program of the DGGG, OEGGG and SGGG

For information on the guidelines program, please refer to the end of the guideline.

Citation format

Bacterial Vaginosis: Guideline of the DGGG, OEGGG and SGGG (S2k-Level, AWMF Registry No. 015/028, June 2023). Geburtsh Frauenheilk 2023; 83: 1331–1349

Guideline documents

The complete German-language long version and a slide version of these guidelines as well as a list of the conflicts of interest of all the authors are available on the homepage of the AWMF: http://www.awmf.org/leitlinien/detail/ll/015-028.html

Guideline authors

See Tables 1 and 2 .

Table 1  Lead and/or coordinating guideline author.

Author	AWMF professional society
Ap.Prof. Priv.-Doz. Dr. Dr. Alex Farr, MPH	Austrian Society for Gynecology and Obstetrics ( Österreichische Gesellschaft für Gynäkologie und Geburtshilfe , OEGGG) European Society for Infectious Diseases in Obstetrics and Gynecology (ESIDOG) Austrian Society for Pre- and Perinatal Medicine (Ö sterreichische Gesellschaft für Prä- und Perinatale Medizin , ÖGfPPM)

Table 2  Contributing guideline authors.

Author Mandate holder	DGGG working group (AG)/ AWMF/non-AWMF professional society/ organization/association
* This person attended as an expert (not eligible to vote in the consensus process).
Alex Farr	OEGGG, ÖGfPPM, ESIDOG
Brigitte Frey Tirri	SGGG
Udo Hoyme*	AGII
Werner Mendling	DGGG, AGII
Inge Reckel-Botzem	BVF
Daniel Surbek	SGGG
Sonja Swidsinski	DGHM
Gisela Walter	DSTIG, AEGGF
Birgit Willinger	OEGHMP

The following professional societies/working groups/organizations/associations stated that they wished to contribute to the guideline text and participate in the consensus conference and nominated representatives to attend the conference ( Table 2 ).

Abbreviations

BV

bacterial vaginosis

BVAB

BV-associated bacteria

CDC

Center for Disease Control and Prevention

CFU

colony-forming units

CST

community state types

FISH

fluorescence in situ hybridization

HIV

human immunodeficiency virus

HPV

human papillomaviruses

HSV

herpes simplex virus

IVF

in vitro fertilization

KOH

potassium hydroxide

NAAT

nucleic acid amplification test

NGS

next generation sequencing

PID

pelvic inflammatory disease

spp.

abbreviation of species (plural)

STI

sexually transmitted infection

VVC

vulvovaginal candidiasis

WSW

women having sex with women

II  Guideline Application

Purpose and objectives

The aim of this guideline is to provide optimal care to patients with bacterial vaginosis, whether they are outpatients, day patients or inpatients. The intention must be to provide targeted therapy depending on the symptoms. The aim is also to prevent unnecessary or inadequate therapies. The prevention and early detection of bacterial vaginosis is also one of the aims of this guideline.

Targeted areas of care

Outpatient care, day-patient care, inpatient care and specialized cared.

Target user groups/target audience

This guideline is addressed to the following groups of people: hospital-based gynecologists and gynecologists in private practice, hospital-based microbiologists, and microbiologists in private practice. Other target groups (providing them with information) include: general practitioners, hospital-based midwives and midwives in private practice, nursing staff, biomedical analysts, professional medical societies and associations, public health institutions und decision-makers at national and federal state levels, and funding agencies.

Adoption and period of validity

The validity of this guideline was confirmed by the executive boards/representatives of the participating medical professional societies, working groups, organizations, and associations and the boards of the DGGG, SGGG, and OEGGG and the DGGG/OEGGG/SGGG Guidelines Commission in June 2023 and was thereby approved in its entirety. This guideline is valid from 1 June 2023 through to 30 May 2027. The guideline can be updated earlier if urgently necessary. Similarly, the guidelineʼs period of validity can be extended if the guideline still reflects the current state of knowledge.

III  Methodology

Basic principles

The method used to prepare this guideline was determined by the class to which this guideline was assigned. The AWMF Guidance Manual (version 1.0) has set out the respective rules and requirements for different classes of guidelines. Guidelines are differentiated into lowest (S1), intermediate (S2), and highest (S3) class. The lowest class is defined as consisting of a set of recommendations for action compiled by a non-representative group of experts. In 2004, the S2 class was divided into two subclasses: a systematic evidence-based subclass (S2e) and a structural consensus-based subclass (S2k). The highest S3 class combines both approaches.

This guideline was classified as: S2k .

Grading of recommendations

The grading of evidence based on the systematic search, selection, evaluation, and synthesis of an evidence base which is then used to grade the recommendations is not envisaged for S2k guidelines. The individual statements and recommendations are only differentiated by syntax, not by symbols ( Table 3 ).

Table 3  Grading of recommendations (based on Lomotan et al., Qual Saf Health Care 2010).

Description of binding character	Expression
Strong recommendation with highly binding character	must/must not
Regular recommendation with moderately binding character	should/should not
Open recommendation with limited binding character	may/may not

Statements

Expositions or explanations of specific facts, circumstances, or problems without any direct recommendations for action included in this guideline are referred to as “statements.” It is not possible to provide any information about the level of evidence for these statements.

Achieving consensus and level of consensus

At structured NIH-type consensus-based conferences (S2k/S3 level), authorized participants attending the session vote on draft statements and recommendations. The process is basically as follows. A recommendation is presented, its contents are discussed, proposed changes are put forward, and all proposed changes are voted on. If a consensus (> 75% of votes) is not achieved, there is another round of discussions, followed by a repeat vote. Finally, the extent of consensus is determined, based on the number of participants ( Table 4 ).

Table 4  Level of consensus based on extent of agreement.

Symbol	Level of consensus	Extent of agreement in percent
+++	Strong consensus	> 95% of participants agree
++	Consensus	> 75 – 95% of participants agree
+	Majority agreement	> 50 – 75% of participants agree
–	No consensus	< 51% of participants agree

Expert consensus

As the term indicates, this refers to consensus decisions taken relating specifically to recommendations/statements issued without a prior systematic search of the literature (S2k) or where evidence is lacking (S2e/S3). The term “expert consensus” (EC) used here is synonymous with terms used in other guidelines such as “good clinical practice” (GCP) or “clinical consensus point” (CCP). The strength of the recommendation is graded as previously described in the chapter Grading of recommendations but without the use of symbols; it is only expressed semantically (“must”/“must not” or “should”/“should not” or “may”/“may not”).

IV  Guideline

1  Summary of recommendations

Bacterial vaginosis (BV) is characterized by a strong increase in the number of bacteria (bacterial overgrowth), especially of Gardnerella (G.) species (spp.), a high microbial diversity as well as the displacement of potentially protective lactobacilli in the vaginal fluid.

Gardnerella spp. is the predominant bacterial species in BV and is also the species with the highest virulence potential. They are integrated in a biofilm matrix with other BV-associated types of bacteria (BVAB), and this is what appears to be responsible for treatment failures and chronic recurrences. Chronically recurrent BV is when the patient has at least 3 episodes per year; this may be an indication that the pathogenesis of BV could be biofilm-related. BV is associated with specific risk factors, and particular attention should be paid to them when making the diagnosis. The aim should be to eliminate predisposing host factors where possible.

In clinical practice, women with vulvovaginal symptoms must be investigated for BV, particularly if they present with a thin homogeneously gray discharge (with or without a fishy odor) and an alkaline vaginal pH. Diagnostic investigations must be guided by the patientʼs prior medical history, clinical findings, and the microscopic evidence of clue cells in the unfixed vaginal smear as well as an evaluation based on the Amsel criteria if necessary. Lab diagnostics should include Gram staining with quantitative comparison of different morphotypes using the Nugent scoring system. Lab diagnostics based on molecular genetic procedures only play a minor role and should currently only be used in special cases.

Treatment for BV must only be initiated after a proper diagnostic investigation has been carried out and medically confirmed. Women with vulvovaginal symptoms and confirmed BV must be treated in accordance with medical guidelines, and therapy must consist of oral or topical clindamycin or metronidazole. Alternatively, local antiseptics may be used. The treatment of bacterial vaginosis in pregnancy must consist primarily of vaginal clindamycin or antiseptics. In cases with chronically recurrent BV, treatment should consist of local antiseptics or suppressive maintenance therapy with topical metronidazole followed by vaginal probiotics to reduce the probability of recurrence.

Lactic acid and probiotics appear to have a positive impact in therapy and as recurrence prophylaxis and may be used as a complementary approach after the completion of therapy to regenerate the lactobacilli flora. Treating the patientʼs partner may be considered in cases with chronic recurrence, although the evidence for this is limited. Women with BV must always be informed about which measures could prevent the recurrence of bacterial vaginosis in their case. Women with BV who wish to become pregnant very soon should be treated even if they are asymptomatic. The same applies to women with aerobic or desquamative inflammatory vaginitis.

Future research into BV should focus on reducing the high rate of recurrence and chronic recurrence of disease.

2  Definition

Consensus-based statement 2.S1
Expert consensus	Level of consensus +++
Bacterial vaginosis is characterized by excessive numbers of bacteria, especially Gardnerella species, a high microbial diversity of anaerobic and facultatively anaerobic bacterial species, as well as the displacement of potentially protective lactobacilli in vaginal fluid.

Consensus-based recommendation 2.E1
Expert consensus	Level of consensus +++
Chronic recurrent bacterial vaginosis is defined as the occurrence of at least 3 episodes per year and may indicate that the pathogenesis is biofilm-related.

BV is the most common vaginal disease worldwide with a prevalence of between 23% and 29% in sexually active women 1 . It is considered a form of vaginal dysbiosis where the vaginal microbiota is significantly different to that of healthy women without BV. The difference includes a strong increase in the numbers of bacteria, especially of Gardnerella spp., a high microbial diversity of facultatively and strictly anaerobic bacteria, and the displacement of protective lactobacilli. There are some indications for the presence of a biofilm on vaginal epithelium consisting predominantly of Gardnerella spp. but also including numerous other BVAB. The existence of such a biofilm would explain some of the changes to the vaginal microbiota and could be considered a pathogenetic factor 2 .

3  Microbiology

Consensus-based statement 3.S2
Expert consensus	Level of consensus +++
Gardnerella species is the predominant bacterial species found in bacterial vaginosis and the species with the highest virulence potential. It forms part of a biofilm matrix which includes other BV-associated bacterial species and appears to be responsible for the ultimate failure of therapy and for chronically recurrent disease.

Gardnerella spp. is the predominant bacterial species in 95 – 100% of women with BV 3 , and includes four different named species (G. vaginalis, G. piotii, G. leopoldii, and G. Swidsinskii) and a further 9 as yet unnamed species 4 . Gardnerella spp. are gram-positive bacteria and is part of the Bifidobacteriaceae family 5 . Because of their unusually thin cell walls, in gram-stained specimens they appear as gram-negative or gram-variable coccobacilli 6 . The specific virulence properties of Gardnerella spp. include its pronounced ability to adhere to vaginal epithelial cells and its biofilm-producing capacity 7 . The BV biofilm consists mainly of tightly packed adjacent Gardnerella spp. Many other different BVAB are integrated in its matrix. Their concentrations are significantly higher than in normal vaginal microbiota but still lower than the concentrations of Gardnerella spp. With the exception of Fannyhessia vaginae (previously Atopobium vaginae), none of the non-Gardnerella species are present in > 60% of BV biofilms. The range of different BVAB is enormous; in addition to the most common species Fannyhessia vaginae, Fusobacterium nucleatum, Mobiluncus mulieris, Mycoplasma hominis, Prevotella bivia, and Ureaplasma urealyticum, it also includes lactobacilli such as L. iners 8 ,  9 . The Gardnerella spp. dominate the polymicrobial biofilm; they are responsible for the increased resistance to hydrogen peroxide, lactic acid, bactericides, and host immune defense 10 and are the main cause of the high failure rate of standard antibiotic therapy and for BV recurrence 11 . STI pathogens also benefit from ecological interactions with the BV biofilm. The risk of contracting sexually transmissible infections is significantly higher for women with BV compared to women with a normal vaginal microbiota 12 ,  13 . Women with BV are also more susceptible to other genital infections 14 .

4  Host, virulence and risk factors

Consensus-based recommendation 4.E2
Expert consensus	Level of consensus +++
BV is associated with certain risk factors, and particular attention should be paid to them when making the diagnosis. The aim should be to eliminate predisposing host factors where possible.

No definite gene locus which could be partly responsible for developing BV has been identified yet 15 . However, an association with womenʼs ethnic affiliations has been reported 16 ,  17 . The incidence of BV is higher in women from South and East Africa than in women from West Africa, Europe, Australia, or New Zealand. In the USA, the reported prevalence of BV was 51% in Afro-American women, 32% in Hispanic women, 23% in white American women and 33% in native American women 11 ,  18 ,  19 . Exogenous host factors for BV include smoking, excessive vaginal hygiene, chronic stress, frequent changes of sexual partner, and imminent menstruation 18 ,  20 ,  21 ,  22 . Brookheart et al. 23 reported that BV occurred more often in women with a high body weight or body mass index. There are some indications that BV biofilms may be sexually transmitted. Women who have been previously treated for BV have a higher risk of recurrence if they have sexual intercourse with the same partner again without using a condom 24 ,  25 ,  26 . Women with same-sex partners (WSW, women having sex with women) have a higher risk per se of BV 27 . Taking combined oral contraceptives is associated with a lower prevalence of BV 28  –  30 , although this appears to be associated with the estradiol in the contraceptive 31 . The use of copper intrauterine devices appears to be associated with an increased risk of BV 32 .

5  Symptoms

Consensus-based recommendation 5.E3
Expert consensus	Level of consensus +++
Women with vulvovaginal symptoms, especially with a thin homogeneous grayish discharge (with or without a fishy odor) and an alkaline vaginal pH, must be investigated for bacterial vaginosis.

The characteristic symptom of BV is increased homogeneous vaginal discharge 33 ,  34 . The discharge is thin with a grayish, slightly milky color 35 ,  36 ,  37 , and is accompanied by a fishy smell and a vaginal pH of > 4.5 34 . Additional irritations in the genital area include burning sensation, redness, itching, dyspareunia, or dysuria. Many women present to their GP because of menstrual disorders or bladder infection symptoms 33 ,  38 . In pregnancy, BV may present with symptoms such as preterm labor, cervical shortening, or premature rupture of membranes 39 ,  40 . In women with sexually transmissible infections including HIV, the other infections may mask the symptoms of BV making it more difficult to arrive at the correct diagnosis 41 ,  42 ,  43 . Vulvovaginal candidiasis (VVC) is the most important differential diagnosis, although the main symptom of VVC is vestibular itching 44 ,  45 . The majority of women (85%) with trichomoniasis are asymptomatic 46 ,  47 , although one third of these women go on to develop symptoms within 6 months 48 . Classic symptoms also include vaginal discharge, which is often foul-smelling and has a yellow-green color, accompanied by dysuria, itchiness, and abdominal pain. In contrast to BV, infection with Chlamydia trachomatis is usually characterized by a limited cervical discharge, cervical contact bleeding, urethritis, menstrual disorders, and in some cases endometritis, salpingitis and lower abdominal pain 44 .

6  Diagnosis

Consensus-based recommendation 6.E4
Expert consensus	Level of consensus +++
Diagnostic investigations for bacterial vaginosis must be guided by the patientʼs prior medical history, clinical findings, and the microscopic evidence of clue cells in the unfixed vaginal smear, as well as an evaluation based on the Amsel criteria if necessary.

Consensus-based recommendation 6.E5
Expert consensus	Level of consensus +++
After the initial diagnostic workup, lab diagnostics for bacterial vaginosis should include Gram staining with quantitative comparison of different morphotypes using the Nugent scoring system.

Consensus-based recommendation 6.E6
Expert consensus	Level of consensus +++
Lab diagnostics for bacterial vaginosis based on molecular genetic procedures only play a minor role in clinical practice and should currently only be used in special cases.

The correct diagnosis of BV is of special clinical importance for women with recurrent disease and women with failure of first-line therapy 49 as well as for women wanting to have children or women being investigated for a history of preterm birth 50 ,  51 . Classic microscopic examinations such as unfixed vaginal smear and gram-staining of specimens are still recommended as the reference methods 49 ,  50 . But these methods can only identify the morphotypes and therefore only permit general statements to be made about the existing changes to the vaginal microbiota. In recent years, molecular genetic methods (FISH, NAAT/PCR, NGS) have been developed to diagnose BV, which provide much more detailed information about changes to the vaginal microbiota, permitting a more precise diagnosis and therefore a more targeted therapy 52 .

In patients with the typical symptoms of BV, the diagnosis can be made based on the patientʼs medical history, local findings, examination of an unfixed vaginal smear, and determination of the vaginal pH 44 . With BV, an unfixed smear obtained from the vaginal discharge (phase-contrast microscopy × 400) will show numerous short coccobacilli on the epithelial cells of the vagina, which Gardner has called clue cells 53 . Lactobacilli and other morphotypes or leukocytes cannot be detected or are almost undetectable 44 . The Amsel criteria are met if 3 of the 4 following characteristics are present: homogeneous grayish-white vaginal discharge; vaginal pH > 4.5; vaginal discharge has a fishy smell, especially after the addition of a drop of 10% potassium hydroxide (KOH); and/or confirmation of at least 20% clue cells in relation to the total number of epithelial cells in the unfixed vaginal smear visible in the field of vision 54 .

The Nugent scoring system is a standardized assessment method for Gram-stained vaginal smears where points (0 to 10) are used for a semiquantitative assessment; large gram-positive bacilli are scored as 0 to 4, small gram-variable and gram-negative bacilli as 0 to 4, and curved Mobiluncus-like bacilli as 0 to 2 55 . With this scoring system, women with clinical symptoms of BV usually have scores between 7 and 10, whereas healthy women have scores between 0 and 3. Scores from 4 to 6 are considered “intermediate” and do not permit clinically relevant statements to be made 56 . The Hay-Ison score may be used as an alternative scoring system. It differentiates between 5 categories (grades 0 to 4), whereby grade 0 indicates that only epithelial cells without lactobacilli or indications of BV are present. Grade 1 constitutes a normal state with a predominance of vaginal lactobacilli, grade 2 represents intermediate mixed vaginal flora with some Gardnerella and Mobiluncus morphotypes, grade 3 stands for typical BV with clue cells dominated by anaerobes (without lactobacilli or with only a few lactobacilli). Grade 4 indicates gram-positive cocci, with indications of BV or lactobacilli morphotypes 57 .

Fluorescence in situ hybridization (FISH) is used to obtain a clear taxonomic identification of the microorganisms and to assess the spatial organization and morphological features of the investigated specimen 58 . Next generation sequencing (NGS) is used to identify the genotype of microorganisms in microbial communities even if only small amounts are present in the clinical sample 59 . Quantitative multiplex polymerase chain reaction (qPCR) has a sensitivity of 80% and a specificity of up to 92% 60 . But this method should be reserved for women who wish to become pregnant, women who are already pregnant, and women undergoing IVF, or to screen patients with an increased risk of STI. Additional laboratory tests include the BD Affirm VPIII assay, a synthetic oligonucleotide probe test 61 ,  62 ,  63 ,  64 , and diverse point-of-care tests such as the OSOM BVBLUE test which is based on the detection of sialidase activity 65 ,  66 , and the FemExam test which detects the metabolite trimethylamine and proline aminopeptidase activity 64 ,  67 .

7  Therapy

Consensus-based recommendation 7.E7
Expert consensus	Level of consensus +++
Women with vulvovaginal symptoms and confirmed bacterial vaginosis must be treated in accordance with medical guidelines.

Consensus-based recommendation 7.E8
Expert consensus	Level of consensus +++
The treatment of bacterial vaginosis must consist of oral or topical clindamycin or metronidazole. Alternatively, local antiseptics may be used.

Consensus-based recommendation 7.E9
Expert consensus	Level of consensus +++
The treatment of chronically recurrent bacterial vaginosis should consist of local antiseptics or suppressive maintenance therapy using topical metronidazole, followed by vaginal probiotics to reduce the probability of recurrence after completing therapy.

Consensus-based recommendation 7.E10
Expert consensus	Level of consensus +++
Pregnant women with symptomatic bacterial vaginosis must be treated to reduce symptoms and reduce complications in pregnancy and puerperium.

Consensus-based recommendation 7.E11
Expert consensus	Level of consensus +++
Therapy for symptomatic bacterial vaginosis in pregnancy must consist primarily of treatment with clindamycin. Alternatively vaginal antiseptics may be used.

Consensus-based recommendation 7.E12
Expert consensus	Level of consensus +++
Treatment for bacterial vaginosis must only be initiated after a proper diagnostic workup and a medically confirmed diagnosis.

Consensus-based recommendation 7.E13
Expert consensus	Level of consensus +++
Lactic acid and probiotics appear to have a positive impact during therapy and as recurrence prophylaxis and may be used as a complementary approach.

Consensus-based statement 7.E14
Expert consensus	Level of consensus +++
If a woman suffers from chronic recurrent BV, treating her partner may be considered although the evidence for this is limited.

Consensus-based recommendation 7.E15
Expert consensus	Level of consensus +++
Women with bacterial vaginosis must be informed about measures which can prevent the recurrence of bacterial vaginosis in their case.

Consensus-based recommendation 7.E16
Expert consensus	Level of consensus +++
Women with bacterial vaginosis who wish to become pregnant soon should be treated even if they are asymptomatic. The same applies to women with aerobic or desquamative inflammatory vaginitis.

Treatment of BV is indicated for all patients with vulvovaginal symptoms 49 . But asymptomatic women with BV also benefit indirectly from therapy as treatment reduces the risk of endometritis, PID and STIs, which are associated with possible subsequent infertility 68 . BV can have a significant negative impact on the quality of life of women with recurrent disease or failure of first-line therapy, and treatment is often difficult and protracted.

If metronidazole is used as first-line therapy for BV, the resistance mechanisms of Gardnerella spp. often lead to treatment failure 69 . If oral therapy consists of 500 mg metronidazole, the recommended treatment is to take it twice a day for 7 days or once or twice within 48 hours for 2 g metronidazole. Therapy with 300 mg clindamycin taken two to three times a day for 7 days is less often limited by treatment resistance. Recommended alternative topical therapies include vaginal metronidazole or 2% clindamycin cream once a day for 7 days, as well as vaginal clindamycin ovules for 3 days. In some randomized controlled studies, antiseptics such as 10 mg dequalinium chloride administered once a day for 6 days, octenidine vaginal spray or iodine-containing preparations were found to have good results when treating BV 70  –  74 . Using antiseptics as a treatment alternative appears to be useful, especially given the fact that metronidazole is often ineffective against the BV biofilm and the increasing resistance to antibiotics, even though antiseptics reduce the already low percentage of lactobacilli in women with BV even further 75 ,  76 ,  77 . More rarely used treatment alternatives include 2 g secnidazole taken orally as a one-off therapy or 2 g oral tinidazole for 2 days or 1 g tinidazole taken orally for 5 days. The administration of probiotics after completing antibiotic or antiseptic therapy aims to restore vaginal microbiota 78 . The risk of repeatedly suffering the same symptoms after completing first-line BV therapy is high. Review articles report a recurrence rate of 80% within 3 – 12 months after completion of therapy 34 ,  79 . The biofilms which are usually responsible for recurrence cannot be removed with standard first-line antibiotics 80 ,  81 , although a disruptive effect has been reported in vitro for dequalinium chloride 82 . Local antiseptics could therefore represent a useful alternative to treat recurrent disease 72 . There have been attempts to use some regimens to suppress recurrence but none of them have been consistently effective, meaning that suppressive maintenance therapy may be proposed, as is used to treat chronically recurrent VVC. Topical metronidazole administered twice a week for a total of 16 weeks may be an appropriate therapeutic approach 83 .

Antibiotic treatment of BV is effective in pregnancy, although it does not significantly reduce the overall risk of preterm birth 84 . But treatment is recommended for all symptomatic pregnant women to reduce their symptoms and because symptomatic BV during pregnancy has been found to be associated with premature rupture of membranes, preterm birth, amniotic infection syndrome and postpartum endometritis 85 ,  86 ,  87 . In addition to symptomatic women, there are indications that the diagnosis and therapy of asymptomatic BV before the 23rd week of gestation can reduce the rate of preterm births 88 . Because of its anti-inflammatory and cytokine-inhibiting effect as well as its broad antibiotic spectrum, clindamycin appears to be more suitable for use during pregnancy than metronidazole 49 . The administration of tinidazole should be avoided during pregnancy. Because of their clinically comparable efficacy, other antiseptics such as dequalinium chloride or octenidine may be considered suitable alternatives 71 ,  73 . Povidone-iodine should not be administered during pregnancy.

According to the research to date, full-blown BV and chronically recurrent BV cannot be cured with lactic acid or lactobacilli preparations alone 89 . In their systematic review, Plummer et al. 90 reported that because of the divergent methods and results it is not possible to make a clear recommendation for or against the use of lactic acid as prophylaxis or treatment against BV. Similarly, the administration of vaginal or oral lactobacilli and probiotics as prophylaxis or therapy to treat BV is still controversial 91 ,  92 ,  93 , even though individual studies have shown significant benefits associated with the use of probiotics during or after BV 78 . Oral probiotics are only detectable about 1 – 2 weeks later in the vagina and remain there as long as they continue to be taken orally 94 ,  95 . Other alternative and complementary therapeutics include a vaginal preparation with ascorbic acid 96 and a polymer made of a combination of Aloe barbadensis and lactic acid 97 .

The rate of BV recurrence depends on a number of factors; re-infection with BVAB and re-infection through endogenous sources but also re-infection via the patientʼs sexual partner all play a role 31 ,  98 . However, the Center for Disease Control and Prevention (CDC) does not recommend routine co-treatment of the partner in cases with recurrent BV. The rate of recurrence may be reduced by avoiding other predisposing host or risk factors, for example avoiding stress, having a healthy life style and a normal body weight 99 . There is an association between BV and infertility although the precise pathomechanism of this is not clear. On the one hand, women with BV have a higher risk of developing ascending infections such as cervicitis, endometritis and pelvic inflammatory disease (PID) 68 . Chronic endometritis may often be clinically inapparent for years, leading to inflammatory processes which can interfere with implantation of the fertilized oocyte and may lead to tubal sterility 100  –  104 . On the other hand, BV biofilms also appear to play a role in fertility 58 . Overall, poor IVF outcomes have been reported for women with low microbial diversity and women with a higher percentage of abnormal vaginal microbiota 105 . Antibiotic therapy with doxycycline after hysteroscopy and endometrial scratching improves the pregnancy rate 106 .

8  Outlook

Consensus-based recommendation 8.E17
Expert consensus	Level of consensus +++
Future research into bacterial vaginosis should focus on preventing the high rates of recurrence and chronic recurrence of disease.

Avoiding resistance and chronic recurrence as well as the prevention of BV are therefore at the top of the agenda for future research. The efficacy of alternative therapies and the current treatment plan should be expanded and amended based on the results of new clinical studies into modern approaches and precision medicine 28 ,  31 ,  107 ,  108 . Treatment with Lactobacillus crispatus CTV-05 (LACTIN-V) could also be an interesting approach as it has been shown that the application of LACTIN-V after initial treatment with vaginal metronidazole resulted in a significantly lower rate of BV recurrence after 12 weeks 109 . The current evidence on the efficacy of astodrimer is relatively limited 110 ,  111 as is the evidence for a vaginal polycarbophil cream consisting of 0.04% lauryl glucoside and glycerides, although the data obtained for polycarbophil regarding a reduction in the rate of BV recurrence has been promising 112 ,  113 . The data from clinical studies into TOL-463, an antiseptic based on boric acid which is especially effective against vaginal biofilms of bacteria and fungi, are still not in 114 ,  115 . Finally, research is also increasingly focusing on co-treatment of the patientʼs partner 116 .