Abstract

BackgroundWe wished to determine recurrences of bacterial vaginosis (BV) after treatment over the course of 12 months and to establish factors associated with recurrence

MethodsWomen with symptomatic BV (a Nugent score [NS] of 7–10 or of 4–6 with ⩾3 Amsel criteria) were enrolled. BV was treated with 400 mg of oral metronidazole twice a day for 7 days. Participants completed a questionnaire and vaginal swabs were collected at 1, 3, 6, and 12 months; the study end point was an NS of 7–10

ResultsA total of 121 (87%) women with an NS of 7–10 and 18 (13%) with an NS of 4–6 and ⩾3 Amsel criteria were enrolled; 130 (94%) returned ⩾1 vaginal samples. Sixty-eight women (58% [95% confidence interval {CI}, 49%–66%]) had a recurrence of BV (NS 7–10), and 84 (69% [95% CI, 61%–77%]) had a recurrence of abnormal vaginal flora (NS 4–10) by 12 months. A past history of BV, a regular sex partner throughout the study, and female sex partners were significantly associated with recurrence of BV and abnormal vaginal flora by multivariate analysis; the use of hormonal contraception had a negative association with recurrence

ConclusionCurrent recommended treatment is not preventing the recurrence of BV or abnormal vaginal flora in the majority of women; factors associated with recurrence support a possible role for sexual transmission in the pathogenesis of recurrent BV

Bacterial vaginosis (BV) is the most common cause of abnormal vaginal discharge in women of reproductive age, and it is associated with serious sequelae such as spontaneous abortion, preterm delivery, and increased susceptibility to HIV and other sexually transmitted infections (STIs) [1–5]. It is characterized by a disturbance in normal vaginal flora, with loss of lactobacilli and increasing numbers of anaerobes and gram-negative rods; however, despite an increasing understanding of the pathogenesis of BV, the causative organism(s) remain unknown

Current recommended treatment regimens for BV are based not on the eradication of a known pathogen but on the restoration of a pattern of normal vaginal flora as established historically using culture-based methods. The recommended first-line therapy is metronidazole or clindamycin [6]; these antibiotics have differing spectra of activity but seemingly equivalent short-term efficacy, with cure rates of 80%–90% at 1 month [7]. Long-term recurrence rates after treatment and factors associated with recurrence have not been established. Few studies have extended follow-up to 3 months in nonpregnant women [8, 9], and data on the long-term efficacy of treatment are limited to 1 observational study, which reported recurrence rates of BV at nonscheduled visits over the course of 6 years of 51% (n=52) after 10 days of oral metronidazole treatment [10]. Whether recurrences are due to failure to eradicate the causative organism(s) or are a consequence of reinfection from sex partners remains unclear. The treatment of male sex partners has failed to alter recurrence rates in previous studies [11–16]; however, many of the behavioral practices seen in women with BV are common to women with STIs [17–23]. Determining the long-term efficacy of current therapies is critical to establishing any clinical or public health benefit of treatment, and identifying the factors associated with recurrence will assist in further elucidating the etiology of BV. To quantify the long-term recurrence rates of BV and to investigate the factors associated with recurrence, we conducted a study, over the course of 12 months, of women who had been treated for 7 days with oral metronidazole

Subjects, Materials, and Methods

Study populationWomen presenting to the Melbourne Sexual Health Centre, Melbourne, Australia, between July 2003 and August 2004, with symptoms of BV (abnormal vaginal discharge or odor) were screened for BV. Those with a diagnosis of symptomatic BV who were treated with metronidazole were offered participation in the cohort study. Women were excluded if they were pregnant, HIV infected, postmenopausal, menstruating, could not participate for 12 months, or had used intravaginal preparations (lubricants or intravaginal creams) within the preceding 72 h

Clinical and laboratory methodsInformed consent was obtained at enrollment, and participants completed a questionnaire regarding symptoms (rating scale, 1–5) and behavioral practices. Clinicians performed a speculum examination and recorded clinical and laboratory findings on a standardized data-collection sheet. A sample of vaginal secretions was obtained for the determination of vaginal pH (Spezialindikator strips [pH 2–9]; Merck), microscopic analysis (wet preparation and Gram’s stain), and the amine test [24]. The evaluation of vaginal Gram-stained smears, wet preparations, and amine tests was performed on site by 2 laboratory staff researchers trained in both the Nugent [24] and Amsel [25] methods. A Nugent score (NS) of 0–3 was classed as normal flora, 4–6 as intermediate flora, and 7–10 as BV. Amsel’s criteria were recorded: homogenous white adherent vaginal discharge, vaginal-fluid pH >4.5, positive amine test, and the presence of Clue cells (⩾20% of epithelial cells). Slides were rescored by an independent microscopist who was blinded to the results of the other microscopists. Forty-eight (11%) of 452 Gram’s stains differed by the Nugent category allocated by the 3 microscopists and were reread by all 3 microscopists, who were blinded to the results; a consensus was reached on all slides

The diagnosis of BV for the purposes of the cohort was made if a woman had an NS of 7–10 or intermediate flora (NS, 4–6) and ⩾3 Amsel criteria. Women with BV were treated with 400 mg of oral metronidazole twice a day for 7 days and underwent screening for the presence of Chlamydia trachomatis by strand-displacement amplification (BD-ProbeTec-ET-CT-Amplified DNA Assay; Becton-Dickinson), Neisseria gonorrhoeae by culture (modified Thayer-Martin medium), Trichomonas vaginalis by culture (modified Diamonds medium), and Candida species by culture (in both media)

Women were scheduled for follow-up at 1, 3, 6, and 12 months. They completed a questionnaire regarding practices and symptoms, oral and vaginal self-treatments, and clinician-prescribed therapies for each interval. Women smeared a self-collected vaginal swab on a glass slide and returned the swab, slide, and questionnaire by mail. Reminders in the form of phone calls, letters, or text messages were sent when the pack was mailed and weekly after the due date for 4 weeks, after which a woman was classified as a nonresponder for that interval. Nonresponders were contacted at the subsequent follow-up point. Women were encouraged to attend a visit or to call the free study telephone number if they had symptoms of BV during the study, rather than attending other clinics. Women reached the study end point if their vaginal smear had an NS of 7–10 or if they received treatment for presumed BV during the study. The Human Research and Ethics Committee of the Department of Human Services, Victoria, approved the study

Statistical analysisData were entered and stored in Microsoft Access and analyzed using SPSS (version 12.0; SPSS) and STATA (version 9.0; StataCorp). Proportions were compared using χ2 and Fisher’s exact tests where appropriate, and 95% confidence intervals (CIs) were calculated. Kaplan-Meier methods were used to generate survival curves for time until the recurrence of BV and abnormal vaginal flora. Behavioral factors associated with the recurrence of BV and abnormal vaginal flora were investigated using a discrete-time version of the proportional-hazards regression model proposed by Carlin et al. [26]. Rate ratios and robust SEs were generated using this methodology. Factors that either reached significance in the univariate analysis or were clinically important were included in the multivariate model; if 2 factors were highly correlated, only 1 was included in the model. Patients were excluded from the analysis when clinical information or specimens were not available

Results

Demographic, behavioral, and clinical characteristics of thestudy population at enrollmentThree hundred forty-two women were identified by clinicians as having symptoms of abnormal vaginal odor or discharge during the study period, and symptomatic BV was diagnosed in 157 (46%); 139 (89%) women with BV were eligible and enrolled in the cohort study. The mean age of the study population was 29.8 years (SD, 8.2 years); demographic and behavioral characteristics of the cohort are described in table 1. Fifty-nine (43%) women reported a past history of BV; the median number of prior BV episodes was 1 (range, 1–10), and the median time of the last episode was 24 weeks (range, 1–64 weeks) before enrollment; 5 women reported their last episode as having been >2 years earlier. Nine women (7%) had received oral antibiotics in the preceding month, 18 (14%) had used a vaginal cream, and 4 (3%) had used both oral antibiotics and vaginal cream (types unknown)

Table 1

Demographic and behavioral characteristics of the study population at enrollment (n=139)

Table 1

Demographic and behavioral characteristics of the study population at enrollment (n=139)

One hundred twenty-one women (87%) had an NS of 7–10, and 18 (13%) had an NS of 4–6; clinical and laboratory characteristics are shown in table 2. All participants were HIV negative, and STIs were uncommon. The median duration of symptoms of BV prior to enrollment was 21 days (range, 1–300 days), with 4 women reporting symptoms for >1 year

Table 2

Clinical and laboratory characteristics of study population at enrollment (n=139)

Table 2

Clinical and laboratory characteristics of study population at enrollment (n=139)

Adherence to metronidazole, adverse effects, and loss tofollow-up during the studyThe self-reported rate of adherence to metronidazole was high, and adverse effects were common (table 3). Three women received treatment for intermediate flora during the study, and follow-up was terminated at the point of treatment. Four women received treatment for presumed BV (unknown) from a private practitioner; all 4 women had a microbiologically confirmed recurrence at the next study review. One hundred thirty (94%) women returned ⩾1 vaginal sample during follow-up. Nine women failed to return any slides and were excluded from the subsequent analysis; they did not differ significantly in demographic or behavioral characteristics from those who remained in the cohort, except that they were more likely to have had <13 years of education and to have had ⩽1 male sex partner during the preceding 3 months (P=.04). Of the 130 women who returned vaginal samples, 96 (74%) returned all samples until the study end point, 13 (10%) missed 1 sample, and 21 (16%) missed 2–3 samples. Eight women failed to return any questionnaires; therefore, 122 women were included in the behavioral analysis

Table 3

Adherence to and adverse effects of metronidazole in the study cohort (n=130)

Table 3

Adherence to and adverse effects of metronidazole in the study cohort (n=130)

Treatment response and recurrence of BV and abnormalvaginal floraThere was a significant improvement in reported vaginal symptoms and their impact on sexual enjoyment and daily life 1 month after therapy (P<.01) (table 4). Cumulative recurrence rates of BV and abnormal vaginal flora were calculated for each interval of follow-up in an on-treatment analysis (table 5), from which women who failed to return slides were excluded. One month after treatment, 25 women (23%; range, 16%–32%) had had a recurrence of BV (NS, 7–10), and 36 (33%; range, 25%–42%) had had a recurrence of abnormal vaginal flora; by month 12, 68 women (58%; range, 49%–66%) had had a recurrence of BV, and 84 (69%; range, 61%–77%) had had a recurrence of abnormal vaginal flora. The Kaplan-Meier survival curves for recurrences of BV and abnormal vaginal flora are shown in figures 1 and 2. Of women who had recurrences, the median time to diagnosis of BV was 176 days (range, 27–391 days), and the median time to diagnosis of abnormal vaginal flora was 99 days (range, 25–291 days). Data regarding symptoms of BV at the study end point were available for 111 (85%) women. Women with a recurrence of BV were more likely to have symptoms (abnormal vaginal discharge or odor) of BV (57%) than were women with an NS of 0–6 (22%) (P<.001)

Table 4

Impact of metronidazole on symptoms and lifestyle in the study cohort (n=130)

Table 4

Impact of metronidazole on symptoms and lifestyle in the study cohort (n=130)

Table 5

Cumulative recurrence of bacterial vaginosis (BV) and abnormal vaginal flora over the course of 12 months

Table 5

Cumulative recurrence of bacterial vaginosis (BV) and abnormal vaginal flora over the course of 12 months

Figure 1

Kaplan-Meier survival curve demonstrating recurrence of bacterial vaginosis over the course of 12 months after treatment for 7 days with oral metronidazole (n=130). Dashed lines indicate 95% confidence intervals

Figure 1

Kaplan-Meier survival curve demonstrating recurrence of bacterial vaginosis over the course of 12 months after treatment for 7 days with oral metronidazole (n=130). Dashed lines indicate 95% confidence intervals

Figure 2

Kaplan-Meier survival curve demonstrating recurrence of abnormal vaginal flora over the course of 12 months after treatment for 7 days with oral metronidazole (n=130). Dashed lines indicate 95% confidence intervals

Figure 2

Kaplan-Meier survival curve demonstrating recurrence of abnormal vaginal flora over the course of 12 months after treatment for 7 days with oral metronidazole (n=130). Dashed lines indicate 95% confidence intervals

Clinical, laboratory, and behavioral associations with the recurrence of BV and abnormal vaginal floraIn the univariate and multivariate analyses (table 6), having a past history of BV, a regular sex partner throughout the study, and female sex partners were associated with recurrence of BV and abnormal vaginal flora, and engaging in sex work had a borderline association with abnormal vaginal flora. The use of hormonal contraception for any period during the study had a negative association with recurrence of BV and abnormal vaginal flora. Having a regular sex partner during the study was highly correlated with a lack of a new sex partner, so the latter was excluded from the multivariate model. The recurrence of BV or abnormal vaginal flora was not associated with smoking, vaginal douching or washing, adherence to metronidazole therapy, number of male sex partners, or condom use for vaginal sex. Confining the analysis to heterosexual women did not change the associations. Oral sex and anal sex did not occur in the absence of vaginal sex, so any independent effect of these practices on recurrence could not be examined. If <50% condom use for vaginal sex was included in the multivariate model, the associations remained the same, except that hormonal contraception had a borderline association with the recurrence of BV (P=.09) and abnormal vaginal flora (P=.10)

Table 6

Behavioral and contraceptive practices associated with recurrences of bacterial vaginosis (BV) and abnormal vaginal flora over the course of 12 months

Table 6

Behavioral and contraceptive practices associated with recurrences of bacterial vaginosis (BV) and abnormal vaginal flora over the course of 12 months

Interestingly, 4 women who had a recurrence of BV reported no sexual activity between treatment and recurrence; 3 had an early recurrence by month 1, and 1 had a recurrence between months 1 and 3. All reported completing metronidazole therapy. There was no significant difference in recurrence rates of BV or abnormal vaginal flora (NS, 4–10) after treatment between women with an NS of 7–10 and those with an NS of 4–6 at enrollment, and no clinical feature or laboratory characteristic at enrollment was associated with a recurrence of BV or abnormal vaginal flora

Discussion

We report high recurrence rates of BV (58%; range, 49%–66%) and abnormal vaginal flora (69%; range, 61%–77%) over the course of 12 months in women treated for 7 days with oral metronidazole. Recurrences of BV and abnormal vaginal flora were associated with a prior history of BV, lack of hormonal contraceptive use, and specific behavioral practices, including having a regular sex partner throughout the study and having a female sex partner. This is the first large prospective study, to our knowledge, to establish recurrence rates of and risk factors for BV over the course of 12 months. Recurrence rates of abnormal vaginal flora and BV after the completion of recommended therapy are unacceptably high and highlight the need to improve current approaches to the management of BV

The recurrence rate of BV at 1 month (23%; range, 16%–31%) in this cohort was within the range reported in a recent meta-analysis (15%–27%) [7]. High levels of adherence and a significant reduction in vaginal symptoms were reported at 1 month, although adverse effects from metronidazole were common. Microbiologically confirmed recurrence of BV and abnormal vaginal flora was associated with a recurrence of vaginal symptoms. The 3-month recurrence rate in our cohort was 43% (range, 34%–52%). Only 1 study has examined the 3-month recurrence rate of BV in nonpregnant women after 7 days of oral metronidazole, and those researchers reported a recurrence rate of 62% [9]; however, there was a 52% loss to follow-up. Comparisons between trials are complicated by the use of different methods for diagnosis of BV. We used the Nugent method, which is reported to have the least inter- and intraobserver variability; however, a significant proportion of women with intermediate flora have BV according to Amsel’s criteria; therefore, the use of the Amsel method may mean that higher recurrence rates are reported. Recurrence rates for BV beyond 3 months have been reported in only 1 study of nonpregnant women (n=52) treated with oral metronidazole for 10 days [10]. Outpatients’ attendance was retrospectively reviewed, and follow-up at 6 years (n=44) revealed a cumulative recurrence rate of BV of 51%; 73% of recurrences had occurred within 12 months. Although no long-term studies of clindamycin therapy for BV have been reported, short-term cure rates appear to be similar to those after 7 days of oral metronidazole [7, 8], so it is doubtful that clindamycin would be more effective over the course of 12 months

Recurrences of BV and abnormal vaginal flora were associated with specific behavioral practices, which supports a possible role for sexual transmission in the pathogenesis of recurrent BV. Having a regular sex partner throughout the study was associated with a higher recurrence rate of both BV and abnormal vaginal flora, whereas having a new partner during the study period was associated with a lower recurrence rate. No difference in other behavioral practices was found between women with and without a regular partner in our cohort, except that the former were more likely to have used condoms <50% of the time; however, condom use was not associated with recurrence in the cohort, and the association remained if this variable was included in the multivariate model. This could suggest that women treated for BV are reinfected by their regular sex partner but by not new partners, who would be less likely to carry the causative agent for BV, or it may be due to unmeasured behavioral factors. Treatment trials for heterosexual partners using imidazoles and clindamycin have not shown a benefit in preventing recurrence [11–16]; however, cross-sectional studies have shown associations with BV and high-risk practices observed in women with STIs [17–23], and BV-associated organisms have been identified in the prepuce and urethra [27, 28]

There are more data to support the sexual transmission of BV between women than within heterosexual partnerships. Gardner and Dukes [29] demonstrated that transmission between women is possible, and BV has been reported to be more prevalent in lesbians than in heterosexual women [18, 30]. There is high concordance of BV within monogamous lesbian partnerships [31, 32], and an association with BV and sharing of sex toys has been observed for lesbians [32]. Women with recurrences of BV and abnormal vaginal flora were more likely to have had female sex partners prior to recurrence in our cohort. Although we attempted to relate specific sexual practices between women and recurrences of BV, there were insufficient numbers to examine practices within this group, and there were no differences in reported practices—such as douching or smoking—between lesbians and heterosexual women. To our knowledge, these are the only published data to associate a recurrence of BV after therapy with female sex partnerships

Cross-sectional studies have shown a higher prevalence of BV in sex workers [33–36], and sex work was associated with recurrence of abnormal vaginal flora but not of BV in our cohort. Sex workers are an established risk group for STIs. Whether the association between recurrence of abnormal vaginal flora and sex work provides further evidence for sexual transmission or is due to other unmeasured factors—such as the disruption of vaginal flora after increased exposure to spermicides and lubricants—or to other behavioral factors can only be postulated. No difference between sex workers and non–sex workers was apparent for practices such as smoking and douching, but sex workers were more likely to have a past history of BV and to have a greater number of male partners (P<.05)

If sexual transmission is involved in the pathogenesis of BV, one would expect a lack of consistent condom use to be associated with BV and recurrent BV. Inconsistent condom use was not significantly associated with recurrences of BV or abnormal vaginal flora in this cohort. Some, but not all, cross-sectional studies have reported an association between BV and inconsistent condom use [23, 37, 38]. One of the difficulties with self-reported condom use is that it can be inaccurate and prone to recall bias. Unprotected vulvo-penile contact and partial penetration commonly occur during foreplay. Couples may use a condom only near ejaculation and report this as 100% protected sex, despite having engaged in contact capable of transmitting an infectious agent

The use of hormonal contraceptives at any point during the present study was associated with a reduced risk of the recurrence of BV and abnormal vaginal flora. A negative association between BV and hormonal contraceptive use has been shown in several cross-sectional studies [20, 38, 39]; however, there have been no published prospective data. Estrogen increases levels of epithelial-cell glycogen, a substrate for lactic-acid production by lactobacilli, and lactic acid is a potent inhibitor of BV-associated organisms [39]. Supraphysiological vaginal estrogen levels may exert a protective effect against BV; however, women who use hormonal contraceptives may also differ behaviorally from those who do not. In our cohort, women who used hormonal contraceptives were less likely to report a past history of BV (P=.05), but there were no other differences in reported behavioral practices, including condom use

The behavioral practices that were associated with recurrences of BV and abnormal vaginal flora in the present trial can only lead to speculation regarding the pathogenesis of BV and recurrent BV. Some of these factors are interrelated, and although we attempted to control for this in a multivariate model, there may be an effect from unmeasured confounding. Many of the factors associated with the recurrence of BV in our study have been reported in cross-sectional studies and are biologically plausible. Although several of them lend support to the theory that sexual transmission is involved in the pathogenesis of BV, some could indicate that there is an influence from other factors (e.g., hormonal factors) or that current therapy is failing to eradicate persistent infection

The strengths of the present study include its size, prospective nature, and high rate of follow-up. Comprehensive behavioral data were collected for the study period, and the use of coded self-answer questionnaires, self-collected sampling, intensive recall, and postage of specimens may have optimized follow-up over the course of the prolonged period. Self-sampling has been shown to be comparable to and more acceptable to subjects than practitioner-collected samples in the diagnosis of BV [40, 41]. Experienced microscopists scored vaginal smears according to the Nugent method, and 3 microscopists reviewed discrepancies in classification independently, so that a consensus was reached. However, limitations include the fact that self-report can be prone to recall bias and the possibility that selection bias could have occurred: women with more-severe symptoms may have enrolled, we may have selected for women with factors favoring recurrence, or, given the intensive follow-up, women with stable lifestyles and few risk practices may have enrolled. We included women with intermediate flora but ⩾3 Amsel criteria in the study. This decision was made because, clinically, this group fulfils the criteria for BV according to the Amsel method, and women with intermediate flora are at a risk of developing adverse sequelae similar to that of women with BV. Importantly, the inclusion of this group demonstrated that there was no difference in treatment response between women with an NS of 4–6 those with an NS of 7–10. The treatment dose of oral metronidazole for BV recommended by the US Centers for Disease Control and Prevention is 500 mg twice a day for 7 days [6]; however, because only the 400-mg oral formulation is available in Australia, standard practice is to use 400 mg twice a day for 7 days. It is unlikely but possible that this slightly reduced daily dose of metronidazole resulted in lower cure rates. Although recurrence rates did not differ between women who did and women who did not use oral and topical medications during the study, their use may have affected vaginal flora in some way. However, the use of narrow- or broad-spectrum antibiotics was not associated with BV recurrences in a previous study [10]

We report unacceptably high recurrence rates of BV and abnormal vaginal flora over the course of 12 months in women treated for 7 days with oral metronidazole, the recommended therapy [6]. Recurrence was associated with specific behavioral practices that support a possible role for sexual transmission in the pathogenesis of recurrent BV; however, prospective studies in other populations are needed to further investigate these associations. It remains unclear whether recurrence is due to the use of antibiotics that do not eradicate persistent infection or to reinfection from sex partners; both factors may contribute to the recurrence of BV in sexually active women. Current therapies are not effectively preventing the recurrence of BV and abnormal vaginal flora in the majority of women. Despite the recent identification of novel BV-associated bacteria [42–44], progress in improving clinical management is impeded by our lack of knowledge of the specific causative agent(s)

Acknowledgments

We thank all practitioners at the Melbourne Sexual Health Centre, for the recruitment of participants; Karen Kolosar and Mary Santoro, for their contribution to the project; and James Beeson, for reviewing the manuscript

References

1
Hay
PE
Lamont
RF
Taylor-Robinson
D
Morgan
DJ
Ison
C
Delivery and late miscarriage
BMJ
 , 
1994
, vol. 
308
 (pg. 
295
-
8
)
2
Hillier
SL
Nugent
RP
Eschenbach
DA
, et al.  . 
Association between bacterial vaginosis and preterm delivery of a low-birth-weight infant. The Vaginal Infections and Prematurity Study Group
N Engl J Med
 , 
1995
, vol. 
333
 (pg. 
1737
-
42
)
3
McGregor
JA
French
JI
Parker
R
, et al.  . 
Prevention of premature birth by screening and treatment for common genital tract infections: results of a prospective controlled evaluation
Am J Obstet Gynecol
 , 
1995
, vol. 
173
 (pg. 
157
-
67
)
4
Sewankambo
N
Gray
RH
Wawer
MJ
, et al.  . 
HIV-1 infection associated with abnormal vaginal flora morphology and bacterial vaginosis
Lancet
 , 
1997
, vol. 
350
 (pg. 
546
-
50
)
5
Taha
TE
Hoover
DR
Dallabetta
GA
, et al.  . 
Bacterial vaginosis and disturbances of vaginal flora: association with increased acquisition of HIV
AIDS
 , 
1998
, vol. 
12
 (pg. 
1699
-
706
)
6
Centers for Disease Control and Prevention
Sexually transmitted diseases treatment guidelines 2002
MMWR Recomm Rep
 , 
2002
, vol. 
51
 (pg. 
1
-
78
)
7
Koumans
EH
Markowitz
LE
Hogan
V
for the CDC BV Working Group
Indications for therapy and treatment recommendations for bacterial vaginosis in nonpregnant and pregnant women: a synthesis of data
Clin Infect Dis
 , 
2002
, vol. 
35
 
Suppl 2
(pg. 
S152
-
72
)
8
Austin
MN
Beigi
RH
Meyn
LA
Hillier
SL
Microbiologic response to treatment of bacterial vaginosis with topical clindamycin or metronidazole
J Clin Microbiol
 , 
2005
, vol. 
43
 (pg. 
4492
-
7
)
9
Sobel
JD
Schmitt
C
Meriwether
C
Long-term follow-up of patients with bacterial vaginosis treated with oral metronidazole and topical clindamycin
J Infect Dis
 , 
1993
, vol. 
167
 (pg. 
783
-
4
)
10
Boris
J
Pahlson
C
Larsson P-G
Six years observation after successful treatment of bacterial vaginosis
Infect Dis Obstet Gynecol
 , 
1997
, vol. 
5
 (pg. 
297
-
302
)
11
Vutyavanich
T
Pongsuthirak
P
Vannareumol
P
Ruangsri
RA
Luangsook
P
A randomized double-blind trial of tinidazole treatment of the sexual partners of females with bacterial vaginosis
Obstet Gynecol
 , 
1993
, vol. 
82
 (pg. 
550
-
4
)
12
Colli
E
Landoni
M
Parazzini
F
Treatment of male partners and recurrence of bacterial vaginosis: a randomised trial
Genitourin Med
 , 
1997
, vol. 
73
 (pg. 
267
-
70
)
13
Vejtorp
M
Bollerup
AC
Vejtorp
L
, et al.  . 
Bacterial vaginosis: a double-blind randomized trial of the effect of treatment of the sexual partner
Br J Obstet Gynaecol
 , 
1988
, vol. 
95
 (pg. 
920
-
6
)
14
Moi
H
Erkkola
R
Jerve
F
, et al.  . 
Should male consorts of women with bacterial vaginosis be treated
Genitourin Med
 , 
1989
, vol. 
65
 (pg. 
263
-
8
)
15
Potter
J
Should sexual partners of women with bacterial vaginosis receive treatment
Br J Gen Pract
 , 
1999
, vol. 
49
 (pg. 
913
-
8
)
16
Hamrick
M
Chambliss
ML
Bacterial vaginosis and treatment of sexual partners
Arch Fam Med
 , 
2000
, vol. 
9
 (pg. 
647
-
8
)
17
Larsson
PG
Platz-Christensen
JJ
Sundstrom
E
Is bacterial vaginosis a sexually transmitted disease
Int J STD AIDS
 , 
1991
, vol. 
2
 (pg. 
362
-
4
)
18
Bailey
JV
Farquhar
C
Owen
C
Bacterial vaginosis in lesbians and bisexual women
Sex Transm Dis
 , 
2004
, vol. 
31
 (pg. 
691
-
4
)
19
Avonts
D
Sercu
M
Heyerick
P
Vandermeeren
I
Meheus
A
Incidence of uncomplicated genital infections in women using oral contraception or an intrauterine device: a prospective study
Sex Transm Dis
 , 
1990
, vol. 
17
 (pg. 
23
-
9
)
20
Bradshaw
CS
Morton
AN
Garland
SM
Morris
MB
Moss
LM
Higher-risk behavioral practices associated with bacterial vaginosis compared with vaginal candidiasis
Obstet Gynecol
 , 
2005
, vol. 
106
 (pg. 
105
-
14
)
21
Nilsson
U
Hellberg
D
Shoubnikova
M
Nilsson
S
Mardh
PA
Sexual behavior risk factors associated with bacterial vaginosis and Chlamydia trachomatis infection
Sex Transm Dis
 , 
1997
, vol. 
24
 (pg. 
241
-
6
)
22
Desmond
RA
Oh
MK
Predictors of bacterial vaginosis in adolescent women who douche
Sex Transm Dis
 , 
2004
, vol. 
31
 (pg. 
433
-
6
)
23
Smart
S
Singal
A
Mindel
A
Social and sexual risk factors for bacterial vaginosis
Sex Transm Infect
 , 
2004
, vol. 
80
 (pg. 
58
-
62
)
24
Amsel
R
Totten
PA
Spiegel
CA
Chen
KC
Eschenbach
D
Nonspecific vaginitis: diagnostic criteria and microbial and epidemiologic associations
Am J Med
 , 
1983
, vol. 
74
 (pg. 
14
-
22
)
25
Nugent
RP
Krohn
MA
Hillier
SL
Reliability of diagnosing bacterial vaginosis is improved by a standardized method of gram stain interpretation
J Clin Microbiol
 , 
1991
, vol. 
29
 (pg. 
297
-
301
)
26
Carlin
JB
Wolfe
R
Coffey
C
Patton
GC
Analysis of binary outcomes in longitudinal studies using weighted estimating equations and discrete-time survival methods: prevalence and incidence of smoking in an adolescent cohort
Stat Med
 , 
1999
, vol. 
18
 (pg. 
2655
-
79
)
27
Serour
F
Samra
Z
Kushel
Z
Gorenstein
A
Dan
M
Comparative periurethral bacteriology of uncircumcised and circumcised males
Genitourin Med
 , 
1997
, vol. 
73
 (pg. 
288
-
90
)
28
Burdge
DR
Bowie
WR
Chow
AW
Gardnerella vaginalis-associated balanoposthitis
Sex Transm Dis
 , 
1986
, vol. 
13
 (pg. 
159
-
62
)
29
Gardner
HL
Dukes
CD
Haemophilus vaginalis vaginitis: a newly defined specific infection previously classified non-specific vaginitis
Am J Obstet Gynecol
 , 
1955
, vol. 
69
 (pg. 
962
-
76
)
30
Bailey
JV
Farquhar
C
Owen
C
Mangtani
P
Sexually transmitted infections in women who have sex with women
Sex Transm Infect
 , 
2004
, vol. 
80
 (pg. 
244
-
6
)
31
Berger
BJ
Kolton
S
Zenilman
JM
Cummings
MC
Feldman
J
Bacterial vaginosis in lesbians: a sexually transmitted disease
Clin Infect Dis
 , 
1995
, vol. 
21
 (pg. 
1402
-
5
)
32
Marrazzo
JM
Koutsky
LA
Eschenbach
DA
Agnew
K
Stine
K
Characterization of vaginal flora and bacterial vaginosis in women who have sex with women
J Infect Dis
 , 
2002
, vol. 
185
 (pg. 
1307
-
13
)
33
Hart
G
Factors associated with trichomoniasis, candidiasis and bacterial vaginosis
Int J STD AIDS
 , 
1993
, vol. 
4
 (pg. 
21
-
5
)
34
Laurent
C
Seck
K
Coumba
N
, et al.  . 
Prevalence of HIV and other sexually transmitted infections, and risk behaviours in unregistered sex workers in Dakar, Senegal
AIDS
 , 
2003
, vol. 
17
 (pg. 
1811
-
6
)
35
Ramjee
G
Karim
SS
Sturm
AW
Sexually transmitted infections among sex workers in KwaZulu-Natal, South Africa
Sex Transm Dis
 , 
1998
, vol. 
25
 (pg. 
346
-
9
)
36
Riedner
G
Rusizoka
M
Hoffmann
O
, et al.  . 
Baseline survey of sexually transmitted infections in a cohort of female bar workers in Mbeya Region, Tanzania
Sex Transm Infect
 , 
2003
, vol. 
79
 (pg. 
382
-
7
)
37
Richey
CM
Weiss
HL
Correlation of behaviors with microbiological changes in vaginal flora
J Infect Dis
 , 
1999
, vol. 
180
 (pg. 
1632
-
6
)
38
Calzolari
E
Masciangelo
R
Milite
V
Verteramo
R
Bacterial vaginosis and contraceptive methods
Int J Gynaecol Obstet
 , 
2000
, vol. 
70
 (pg. 
341
-
6
)
39
Shoubnikova
M
Hellberg
D
Nilsson
S
Mardh
PA
Contraceptive use in women with bacterial vaginosis
Contraception
 , 
1997
, vol. 
55
 (pg. 
355
-
8
)
40
Sturm
PD
Moodley
P
Nzimande
G
Balkistan
R
Connolly
C
Diagnosis of bacterial vaginosis on self-collected vaginal tampon specimens
Int J STD AIDS
 , 
2002
, vol. 
13
 (pg. 
559
-
63
)
41
Boskey
ER
Atherly-Trim
SA
O’Campo
PJ
Strobino
DM
Misra
DP
Acceptability of a self-sampling technique to collect vaginal smears for gram stain diagnosis of bacterial vaginosis
Womens Health Issues
 , 
2004
, vol. 
14
 (pg. 
14
-
8
)
42
Fredricks
DN
Fiedler
TL
Marrazzo
JM
Molecular identification of bacteria associated with bacterial vaginosis
N Engl J Med
 , 
2005
, vol. 
353
 (pg. 
1899
-
911
)
43
Verstraelen
H
Verhelst
R
Claeys
G
Temmerman
M
Vaneechoutte
M
Culture-independent analysis of vaginal microflora: the unrecognized association of Atopobium vaginae with bacterial vaginosis
Am J Obstet Gynecol
 , 
2004
, vol. 
191
 (pg. 
1130
-
2
)
44
Verhelst
R
Verstraelen
H
Claeys
G
, et al.  . 
Cloning of 16S rRNA genes amplified from normal and disturbed vaginal microflora suggests a strong association between Atopobium vaginae, Gardnerella vaginalis and bacterial vaginosis
BMC Microbiol
 , 
2004
, vol. 
4
 pg. 
16
 
(See the editorial commentary by Marrazzo, on pages 1475–7.)
Presented in part: Australasian Sexual Health Conference, Adelaide, Australia, 31 March–3 April 2004 (abstract 1343); 13th International Union against Sexually Transmitted Infections, Asia Pacific Conference, Chiangmai, Thailand, 6–9 July 2004 (abstract 20); Annual Scientific Symposium of the Infectious Diseases Society for Obstetrics and Gynecology, San Diego, 5–7 August 2004 (abstract 15); 16th Biennial Meeting of the International Society for Sexually Transmitted Diseases Research, Amsterdam, Netherlands, 10–13 July 2005 (abstract TO-104); Australasian Sexual Health Conference, Hobart, Australia, 22–24 August 2005 (abstract P42)
Potential conflicts of interest: none reported
Financial support: National Medical and Research Council (research scholarship to C.S.B.)