Methods

Setting: Italy, outpatient clinic.

Trial design: multicentre randomized clinical trial, parallel, 2 arms.

Funding sources: not mentioned.

Ethical issues: ethical board and signed consent.

Participants

Age of participants: from 18 to 45 years.

Inclusion criteria:
sexually‐active women with 1 current sexual partner. Clinical diagnosis of bacterial vaginosis (BV and whose partner agreed to be treated. The study authors defined BV as the presence of clue cells on a wet mount slide plus at least 2 of the following.

• Vaginal discharge with pH greater than 4.5.

• Increased thin homogeneous vaginal discharge that adheres to vaginal walls.

• Release of amine odour from a simple of the discharge after addition of 10% KOH.

Exclusion criteria: patients treated with systemic or topical
antibacterial agents in the 2 weeks before diagnosis of BV; use of intrauterine device or condom; clinical evidence of mucopurulent
cervicitis, candidiasis, trichomoniasis,
herpes genitalis, papilloma virus, Chlamydia trachomatis, or Neisseria gonorrhoeae infection.

Population

Number of participants: 139 women, 139 men.

Baseline characteristics: sexually‐active women with 1 current sexual partner. Women were treated with clindamycin 2% cream, administered intravaginally at bed time for 7 days. Participants were invited to abstain from intercourse during the treatment period and for 1 week after the end of treatment. Most participant couples were married (intervention 66% versus placebo 75%), did not use contraception method (intervention 66% versus placebo 63%) and lacked a previous history of pelvic infection (intervention 79% versus placebo 68%) or male urethritis (intervention 98% versus placebo 95%).

Interventions

Total number of intervention groups: Two groups.

Intervention: clindamycin hydrochloride capsules, 150 mg by mouth 4 times daily for 7 consecutive days.
Comparison: placebo.

Outcomes

The trial included evaluation at 1, 4 and 12 weeks after the start of treatment. The participants and their partners had a clinical examination, including the collection of samples of vaginal discharge to check for clue
cells, the determination of vaginal pH, and a
KOH test. At both visits, the participant and her
partner were asked about medication side effects. The study authors defined cure as the absence of clue cells plus at least 2 of the following: vaginal pH less than 4.5; negative 10% KOH sniff test or grossly normal vaginal discharge (defined as translucent white, flocculent, low volume). The study authors defined recurrence as participants previously healthy after treatment that developed a new episode of BV (Amsel criteria). Participants reported adverse events as symptoms.

Notes

None

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Unclear risk

“…This was a double blind, randomised, controlled trial…”
Comment: there was insufficient information to enable us to make a judgement.

Allocation concealment (selection bias)

Unclear risk

Comment: there was insufficient information to enable us to make a judgement.

Blinding of participants and personnel (performance bias)
All outcomes

Low risk

The trial authors did not adequately report the method implemented to blind study participants and personnel from knowledge of which intervention a participant received. "The clinicians were blind to the study treatment…”
Comment: this was probably done.

Blinding of outcome assessment (detection bias)
All outcomes

Unclear risk

The trial authors did not adequately report the method implemented to blind outcome assessment from knowledge of which intervention a participant received. Comment: there was insufficient information to enable us to make a judgement.

Incomplete outcome data (attrition bias)
All outcomes

Unclear risk

For the outcomes clinical cure at 1, 4 and 12 weeks and gastrointestinal symptoms the risk of bias was low (e.g. no missing outcome data; missing outcome data was balanced across groups). However, for the outcome recurrence, the risk of bias was high according to the level of missing data (greater than 20%).

Selective reporting (reporting bias)

Unclear risk

The report did not have sufficient information to permit a judgment of “yes” or “no”.

Other bias

Low risk

This study appeared to be free of other sources of bias.

Methods

Setting: Finland, outpatient of primary care.

Trial design: single randomized clinical trial, parallel, 2 arms.

Funding sources: Orion pharmaceutical.

Ethical issues: not mentioned.

Participants

Age of participants: from 19 to 53 years.

Inclusion criteria:
women with malodorous vaginal discharge and anaerobic vaginosis, which was based on the identification of clue cells in the PAP smear, pH of the vaginal discharge (greater than 4.5) and a positive amine test.

Screening negative for Neisseria gonorrhoea, Chlamydia trachomatis, Candida albicans and Trichomonas vaginalis.

Exclusion criteria:
presence of N. gonorrhoea, C. trachomatis, C. albicans or T. vaginalis.

Population

Number of participants: 90 women, 90 men.

Baseline characteristics: the women were treated with oral tinidazole 500 mg twice a day for 4 days or oral tinidazole 150 mg twice a day for 7 days or intravaginal tinidazole 500 mg at bedtime for 14 days. The trial authors did not provide any more information.

Interventions

Total number of intervention groups:Two groups.
Intervention: oral tinidazol 1 g for 4 days.
Control: no treatment.

Outcomes

The trial included evaluation at 1 and 3 months after the start of treatment. Women had a clinical examination, including the collection of PAP smear to check for clue
cells, the determination of vaginal pH, and a
KOH test.

Cure was defined as the absence of symptoms of BV or absence of clinically verified BV (Amsell criteria)

Relapse was defined as patients who were asymptomatic after treatment and again developed symptoms of BV. Adverse events were not reported.

Notes

None

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Unclear risk

"Prospective randomised trial"

Comment: there was insufficient information to enable us to make a judgement.

Allocation concealment (selection bias)

Unclear risk

Comment: there was insufficient information to enable us to make a judgement.

Blinding of participants and personnel (performance bias)
All outcomes

High risk

There was no blinding or incomplete blinding, and the outcome or outcome measurement was likely to be influenced by a lack of blinding.

Blinding of outcome assessment (detection bias)
All outcomes

High risk

There was no blinding or incomplete blinding, and the outcome or outcome measurement was likely to be influenced by a lack of blinding. The outcomes symptomatic of improvement and relapse were subjectively assessed and the lack of blinding could have affected the results. The outcome of clinical cure was assessed objectively and the lack of blinding could not have affected the results.

Incomplete outcome data (attrition bias)
All outcomes

High risk

We judged the risk of bias as high for all outcomes according to the level of missing data (greater than 20%). For Group A, 7 participants were lost to follow‐up. In Group B 6 participants dropped out and in Group C 4 participants discontinued the trial.

Selective reporting (reporting bias)

Unclear risk

The report had insufficient information to permit judgment of “yes” or “no”.

Other bias

Low risk

Industry sponsored this trial. However, because the result were not positive, the study would be free of other sources of bias.

Methods

Setting: USA, outpatient of primary care.

Trial design: multicentric randomized clinical trial, parallel, 2 arms.

Funding sources: grants from the research committee, American Academy of Familiy Physicians; Family Health Foundation of America; Washington Academy of Family Physicians, and the Washington Familiy Health Foundation.

Ethical issues: ethical board and signed consent.

Participants

Age of participants: from 18 to 40 years.

Inclusion criteria:
women with vaginal discharge fulfilled 3 of the 4 Amsel criteria. Participants were required to have a telephone and only 1 current sexual partner.

Exclusion criteria:
pregnant or menopausal; had used antibiotics or vaginal medication in the previous month; clinical evidence of a mucopurulent cervical discharge or genital herpes, or had T. vaginalis or C. albicans on wet mounts or KOH preparation of their vaginal discharge; contraindications to metronidazole; history of seizure disorder, peripheral neuropathy, cancer or liver disease; prescription with warfarin, phenytoin or phenobarbital.

Population

Number of participants: 161 women, 98 men

Baseline characteristics: sexually‐active women with 1 current sexual partner. Women were treated with metronidazole 500 mg twice a day for 7 days or 2 g single doses. Participants were not invited to abstain from intercourse during the treatment period. The participants included single and married couples, nulliparous and multiparous women. Most participant couples used a contraception method, and lacked a previous history of pelvic infection and male urethritis.

Interventions

Total number of intervention groups: Two groups.
Intervention: 2 g metronidazole oral single dose.
Comparison: placebo.

Outcomes

The trial included evaluations at 2 and 8 weeks after the onset of the treatment. Women had a clinical examination, including the collection of samples of vaginal discharge to check for clue
cells, gram staining and KOH test; sexual partners were contacted by phone for instructions. At the follow‐up visits participants completed a questionnaire on symptoms and medication side effects, and underwent a pelvic examination.

The trial authors defined cure as the absence of at least 3 of the 4 Amsell criteria. They did not define recurrence. Participants reported adverse events as symptoms of BV.

Notes

None

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

"Randomization was accomplished by blocks of varying sizes (4, 8 or 12) so that an equal number of women in each block entered each of the four treatment groups".

Comment: this was probably done.

Allocation concealment (selection bias)

Low risk

"Randomization was accomplished by blocks of varying sizes (4, 8 or 12) so that an equal number of women in each block entered each of the four treatment groups"

Comment: this was probably done. We considered that the authors implemented a valid randomized method which implies that the allocation concealment was probably through the use of consecutively numbered sealed opaque envelopes.

Blinding of participants and personnel (performance bias)
All outcomes

Low risk

“placebo was used to ensure that both subjects and physicians did not know the subject's treatment” "metronidazole was identically coloured and shaped to placebo".
Comment: this was probably done.

Blinding of outcome assessment (detection bias)
All outcomes

Low risk

“placebo was used to ensure that both subjects and physicians did not know the subject's treatment” "metronidazole was identically coloured and shaped to placebo". The outcomes of cure rate and side effect were objectively assessed. Comment: this was probably done.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

20% or fewer participants were excluded and also intention to treat analyses was reported.

Selective reporting (reporting bias)

Unclear risk

The report had insufficient information to permit a judgment of “yes” or “no”.

Other bias

High risk

This trial was supported by grants from the research committee, American Academy of Family Physicians; Family Health Foundation of America; Washington Academy of Family Physicians, and the Washington Family Health Foundation.

Methods

Setting: Sweden, Denmark, Finland and Norway. Outpatient clinic.

Trial design: multicentric randomized clinical trial, parallel, 2 arms.

Funding sources: grants from Rhone‐Paulenc Pharma Norden AS and Örebro County Council Research Committee.

Ethical issues: ethical board and signed consent.

Participants

Age of participants: from 17 to 56 years.

Inclusion criteria:
non‐pregnant sexually‐active women with 3 of the 4 Amsel criteria. Women had 1 male consort with whom they had sexual intercourse without using a condom, at least once a week during the study period.

Exclusion criteria:
pregnancy or lactation; haematological or neurological disease; history of allergy to metronidazole; antibiotic treatment in the preceding week.

Population

Number of participants: 241 women: 241 men
Baseline characteristics: women were treated with 2 g metronidazole on days 1 and 3. A few of the men were circumcized. Participants were not invited to abstain from intercourse during the treatment period. Women had 1 male consort with whom they had sexual intercourse without using a condom. The most frequently used contraception method was oral contraceptives or IUCD.

Interventions

Total number of intervention groups: Two groups.
Intervention: oral metronidazole 2 g, that was repeated 2 days later.
Comparison: placebo.

Outcomes

The triaI included evaluation at 1, 4, and 12 weeks after starting treatment. The trial asked participants about symptoms of BV and had a clinical examination including collection of samples of vaginal discharge to check for odour of the discharge, vaginal pH, KOH test and wetness of the vaginal discharge. The trial authors defined cure as the disappearance of at least 2 previous signs or symptoms reported. The trial authors defined recurrence as participants who were previously healthy after treatment but developed a new episode of BV (Amsell criteria). Adverse events were not reported.

Notes

None

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Unclear risk

"They were randomly allocated to one of two groups for a double blind trial”.

Comment: there was insufficient information to make a judgement.

Allocation concealment (selection bias)

Unclear risk

Comment: there was insufficient information to make a judgement

Blinding of participants and personnel (performance bias)
All outcomes

Low risk

“They were randomly allocated to one of two groups for a double blind trial”, “Whose male consorts were given inert but identical placebo tablets”.

Comment: this was probably done.

Blinding of outcome assessment (detection bias)
All outcomes

Low risk

“They were randomly allocated to one of two groups for a double blind trial”, “Whose male consorts were given inert but identical placebo tablets”.
Comment: this was probably done.

Incomplete outcome data (attrition bias)
All outcomes

Unclear risk

For the outcomes of symptomatic improvement and clinical cure at first and fourth week the risk of bias was low (e.g. no missing outcome data; missing outcome data balanced across groups). For the same outcomes but assessed at 12 weeks, the risk of bias was high according to the level of missing data (greater than 20%).

Selective reporting (reporting bias)

Unclear risk

The report had insufficient information to permit judgment of “yes” or “no”.

Other bias

Low risk

The trial was sponsored by industry. However, because the results were not positives, the study would be free of other sources of bias.

Methods

Setting: USA, outpatient of family practice clinic.

Trial design: single randomized clinical trial, parallel, 2 arms.

Funding sources: Searle Pharmaceuticals and School of Human Medicine, University of Wyoming.

Ethical issues: ethical board and signed consent.

Participants

Age of participants: from 18 to 45 years.

Inclusion criteria:
non‐pregnant women symptomatic for BV; 3 of the 4 Amsel criteria; absence of uterine infection, mucopurulent cervicitis, trichomoniasis or yeast on microscopic examination.

Exclusion criteria:
antibiotics or vaginal cream in the previous 30 days; history of allergy to metronidazole.

Population

Number of participants: 82 women, 82 men.

Number of participants who received the intervention: 14 (single 2‐g dose) and 13 (7‐day regimen).

Number of participants who received other(s) intervention(s) or placebo: No treatment was administred to 55 participants, 32 in yhe 2‐g single dose group and 23 in the seven day regimen group.

Baseline characteristics: women were treated with metronidazole 500 mg twice a day for 7 days or 2 g single doses. The trial authors did not provide more information.

Interventions

Total number of intervention groups: Two groups.
Intervention: single oral 2 g or 500 mg of metronidazole twice daily for 7 days.
Comparison: no intervention.

Outcomes

The trial included evaluation at 7 to 10 and 21 days after starting treatment. All participants had a clinical examination, including the collection of samples of vaginal discharge to check for clue cells, KOH test, vaginal pH and cultures for gonorrhoea. At both visits, the participants were evaluated by a Likert‐type questionnaire for symptoms of vaginitis (vulvar itching, vulvar burning, odour and quantity of discharge).

Cure was defined if G vaginalis was not isolated on culture and if symptoms had markedly improved or been assent (Likert‐type questionnaire). The participants reported adverse events as symptoms.

Notes

None

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

"Computer generated random number list".

Comments: this was probably done.

Allocation concealment (selection bias)

Low risk

"Regimen dispensed by the clinic pharmacist"

Comment: this was probably done. Telephone or central randomization.

Blinding of participants and personnel (performance bias)
All outcomes

High risk

There was no blinding or incomplete blinding, and the outcome or outcome measurement is likely to have been influenced by a lack of blinding.

Blinding of outcome assessment (detection bias)
All outcomes

Low risk

"The clinical practitioner and the laboratory personnel were "blinded" in that they did not know to which treatment group the patient would be assigned".
Comment: this was probably done.

Incomplete outcome data (attrition bias)
All outcomes

High risk

More than 20% were excluded: "18 did no return at 21 days after treatment, 14 were excluded because they failed to return at 7 – 10 day".

Selective reporting (reporting bias)

Unclear risk

The report had insufficient information to permit judgment of “yes” or “no”.

Other bias

Low risk

Searle Pharmaceuticals provided the medication, Other funds for this project were provided by the School of Human Medicine, University of Wyoming.
Comment: sponsored by industry. However, because the result were not positives, the study would be free of other sources of bias.

Methods

Setting: Denmark, outpatient clinic.

Trial design: single randomized clinical trial, parallel, 2 arms.

Funding sources: Rhone‐Poulenc Pharma Norden A/S.

Ethical issues: ethical board and sign consent.

Participants

Age of participants: from 17 to 50 years.

Inclusion criteria:
non‐pregnant women with 3 of the 4 Amsel criteria.

Exclusion criteria:
women diagnosed with Candida or trichomonal infection; mucopurulent cervical discharge; tenderness of the uterus or adnexa; adnexal masses.

Population

Number of participants: 126 women, 126 men.

Baseline characteristics: women were treated with 2 g metronidazole on days 1 and 3. Most participant couples used a contraception method (birth control pill and IUCD). The trial authors did not provide any more information.

Interventions

Total number of intervention groups: Two groups.
Intervention: metronidazole tablets, 2 g on days 1 and 3.
Comparison: placebo.

Outcomes

The trial included evaluation at 1 and 5 weeks after the start of treatment. The women had a clinical examination, including the collection of samples of vaginal discharge to check for clue
cells, the determination of vaginal pH, and a
KOH test. At both visits, the participant and her
partner were asked about medication side effects.

The trial authors defined cure as the absence or symptoms of BV (increased discharge, malodour, burning sensation or itching) and without at least 3 of the 4 Amsell criteria.

The trial authors defined recurrence as participants that were clinically cured after treatment and then developed a new episode of BV (Amsel criteria).

Adverse events were not reported.

Notes

None.

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

“The partners were allocated to receive metronidazole or placebo by random allocation in blocks of four”

Comment: this was probably done.

Allocation concealment (selection bias)

Unclear risk

Comment: there was insufficient information to make a judgement.

Blinding of participants and personnel (performance bias)
All outcomes

Low risk

The trial authors did not adequately report the method implemented to blind study participants and personnel from knowledge of which intervention a participant received. "The women were treated with metronidazole tablets, 2 g on days 1 and 3. The partner was given the same treatment or a placebo".
Comment: this was probably done.

Blinding of outcome assessment (detection bias)
All outcomes

Low risk

The trial authors did not adequately report the method implemented to blind study participants and personnel from knowledge of which intervention a participant received. "The women were treated with metronidazole tablets, 2 g on days 1 and 3. The partner was given the same treatment or a placebo".
Comment: this was probably done.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

20% or fewer participants were excluded and also intention to treat analyses was reported.

Selective reporting (reporting bias)

Unclear risk

The report did not have sufficient information to permit a judgment of “yes” or no”.

Other bias

Low risk

Rhone‐Poulenc Pharma Norden A/S supported the trial.
Comment: this trial was sponsored by industry. However, because the results were not positive, the study would be free of other sources of bias.

Methods

Setting: Thailand, outpatient clinic.

Trial design: single randomized clinical trial, parallel, 2 arms.

Funding sources: Faculty of Medicine, Chiang Mai University.

Ethical issues: ethical board and signed consent.

Participants

Age of participants: from 17 to 40 years.

Inclusion criteria:
symptomatic patients with 3 of the 4 Amsel criteria; 1 current sexual partner with whom had regular sexual intercourse.

Exclusion criteria:
pregnant, lactating, or menopausal women; candidal or trichomonal infection; adnexal masses or mucopurulent cervical discharge, or tenderness of the uterus or adnexa; contraindications to nitro imidazole; inability to attend follow‐up visits; antibiotic or other treatment for vaginitis in the previous month.

Population

Number of participants: 250 women, 250 men.

Baseline characteristics: sexually‐active women with 1 current sexual partner whom had regular sexual intercourse. Women were treated with single dose of tinidazole 2 g. Most participant couples used a contraception method of tubal resection (intervention 32% versus placebo 28%), oral pill (intervention 17% versus placebo 26%) or DMPA (intervention 12% versus placebo 9%) and lacked a previous history of STI diseases (intervention 88% versus placebo 81%).

Interventions

Total number of intervention groups: Two groups.
Intervention: a single oral dose of 2 g tinidazole.
Comparison: placebo.

Outcomes

The trial included evaluation at 1 and 4 weeks after the start of treatment. The women had a clinical examination, including the collection of samples of vaginal discharge to check for clue
cells, the determination of vaginal pH and a
KOH test. At both visits, the participant and her
partner were asked about of any symptoms after taking the medication.

The trial authors defined clinical cure as the proportion of women who remained without at least 2 of the 4 criteria for BV (Amsell criteria). The trial authors defined symptomatic improvement as participants with an absence of symptoms of BV (abnormal vaginal discharge or pruritus vulvae, or both). The participants reported adverse events as symptoms.

Notes

None.

Risk of bias

Bias

Authors' judgement

Support for judgement

Random sequence generation (selection bias)

Low risk

"patients were randomized into two groups using a table of random numbers”.

Comment: this was probably done.

Allocation concealment (selection bias)

Unclear risk

Comment: there was insufficient information to make a judgement.

Blinding of participants and personnel (performance bias)
All outcomes

Low risk

“Either 2 g tinidazole or identical‐looking placebo packed similarly in packets of four tablets”. Comment: this was probably done.

Blinding of outcome assessment (detection bias)
All outcomes

Low risk

“Either 2 g tinidazole or identical‐looking placebo packed similarly in packets of four tablet”. Comment: this was probably done.

Incomplete outcome data (attrition bias)
All outcomes

Low risk

20% or fewer participants were excluded and also intention to treat analyses was reported.

Selective reporting (reporting bias)

Unclear risk

The report had insufficient information to permit a judgment of “yes” or “no”.

Other bias

Low risk

The study was supported by grant from the Faculty of Medicine Endowment Fund for Medical Research, Chiang Mai University. Comment: the industry probably did not influence the results.