B R I E F R E P O R T
Establishing and Sustaining a Healthy
Vaginal Environment: Analysis of
Data From a Randomized Trial of
Periodic Presumptive Treatment for
Jennifer E. Balkus,1Barbra A. Richardson,2Kishorchandra Mandaliya,3
James Kiarie,4Walter Jaoko,4Jeckoniah O. Ndinya-Achola,4Jeanne
Marrazzo,5Carey Farquhar,1,5,6and R. Scott McClelland1,4,5,6
1Department of Epidemiology, and2Department of Biostatistics, University of
Washington, Seattle,3Coast Provincial General Hospital, Mombasa, Kenya;
4Universityof Nairobi, Kenya;5Department of Medicine, and6Departmentof Global
Health, University of Washington, Seattle
Data from a randomized trial of oral periodic presumptive
treatment (PPT) to reduce vaginal infections were analyzed
to assess the effect of the intervention on a healthy vaginal
environment (normal flora confirmed by Gram stain with no
candidiasis or trichomoniasis). The incidence of a healthy
vaginal environment was 608 cases per 100 person-years in
the intervention arm and 454 cases per 100 person-years in
the placebo arm (hazard ratio [HR], 1.36; 95% confidence
interval [CI], 1.17–1.58). Sustained vaginal health (healthy
vaginal environment for R3 consecutive visits) was also
more frequent in the intervention arm (HR, 1.69; 95% CI,
1.23–2.33). PPT is effective at establishing and sustaining
a healthy vaginal environment.
Vaginal conditions including bacterial vaginosis (BV), vulvo-
vaginal candidiasis (VVC), Trichomonas vaginalis, andabnormal
vaginal flora are highly prevalent among reproductive-aged
at the same time, are also common. Disturbances in the vaginal
environment due to abnormal vaginal flora and vaginal
infections have been associated with increased risk of sexually
transmitted infections (STIs), including human immunodefi-
ciency virus type 1 (HIV-1) [1–5]. Conversely, the presence of
normal vaginal flora has been associated with the lowest risk of
HIV-1 and STI acquisition [3, 6, 7].
Interventions that improve vaginal health by establishing and
sustaining a healthy vaginal environment, defined as the absence
of any vaginal infections or abnormal vaginal flora, may reduce
susceptibility to HIV-1 and other STIs. Recently, a randomized
controlled trial (RCT) assessed the effect of monthly periodic
presumptive treatment (PPT) with 2 g of oral metronidazole
plus 150 mg of fluconazole versus placebo on the incidence of
among Kenyan women . The intervention reduced the
incidence of BV and increased colonization with Lactobacillus
species. In this secondary analysis, we take an additional step,
assessing the effect of PPT on establishing and sustaining
a healthy vaginal environment.
Detailed methods for the RCT were published previously .
We conducted the trial in Mombasa, Kenya, between May 2003
and December 2006. Female sex workers who were enrolled in
an open cohort study of risk factors for HIV-1 acquisition 
were eligible to participate if they were HIV-1 seronegative, aged
18–45 years,and nonpregnant. To avoid open-label treatment at
enrollment, women with abnormal vaginal discharge or itching
were ineligible to enroll. The study was approved by the
institutional review boards at Kenyatta National Hospital
(Nairobi, Kenya)andthe UniversityofWashington(Seattle).All
participants provided written, informed consent.
At enrollment and at each of 12 monthly follow-up visits, we
conducted a brief face-to-face interview to collect information
on medical and sexual history. A physical examination, in-
cluding speculum-assisted pelvic examination, was performed,
and we collected specimens to diagnose genital tract infections.
We collected blood for HIV-1 testing and performed a urine
pregnancy test. Participants were randomized to receive 2 g of
metronidazole plus 150 mg of fluconazole or identical placebo
monthly. At monthly visits, we administered study product
orally as directly observed treatment. We syndromically treated
women who reported abnormal vaginal discharge or vulvova-
ginal itching with a single 2 g dose of oral metronidazole plus
clotrimazole 200 mg vaginal suppositories nightly for 3 nights.
When this treatment was dispensed, we withheld study product.
Received 1 November 2010; accepted 9 February 2011.
Potential conflicts of interest: none reported.
Presented at: International Conference on Microbicides, Pittsburgh, Pennsylvania,
22–25 May 2010.
Correspondence: Jennifer E. Balkus, MPH, Department of Epidemiology, University of
Washington, Box 359909, 325 9th Avenue, Seattle, WA 98104 (email@example.com).
The Journal of Infectious Diseases
? The Author 2011. Published by Oxford University Press on behalf of the Infectious Diseases
Society of America. All rights reserved. For Permissions, please e-mail: journals.permissions@
0022-1899 (print)/1537-6613 (online)/2011/2042-0023$14.00
d JID 2011:204 (15 July)
We treated other genital infections according to World Health
Organization guidelines .
We performed all laboratory procedures in Mombasa. A
Gram stain of vaginal fluid was evaluated for diagnosis of BV by
Nugent criteria . A vaginal saline wet mount was examined
microscopically for the presence of motile trichomonads and
fungal elements. A drop of 10% potassium hydroxide was added
to the slide and evaluated again for the presence of yeast buds or
hyphae. T. vaginalis culture was performed in Diamond’s
modified medium. HIV-1 testing was performed using an
enzyme-linked immunosorbent assay (ELISA, Detect-HIV
[BioChem ImmunoSystems]). A positive ELISA result was
confirmed using a second ELISA (Recombigen [Cambridge
Biotech] or Vironostika [bioMerieux]).
The objectives of this analysis were to assess the effect of the
intervention on the presence of a healthy vaginal environment
(defined as a Nugent score of 0–3 with no yeast on wet mount
sustained vaginal health (defined as the presence of a healthy
vaginal environment for R3 consecutive visits). Women were
included in the analysis if they were randomized and returned
for at least 1 follow-up visit. We conducted an intent-to-treat
analysis using an Andersen–Gill proportional hazards model
that allows for recurrent events to estimate the effect of the
environment during follow-up. Participants were censored if
they became pregnant or HIV-1 seropositive, or at 420 days
following the date of enrollment (administrative censoring).
Because the outcome for the sustained vaginal health analysis
was a composite variable based on vaginal health status at
3 consecutive visits, this analysis was restricted to women who
were returned for R3 follow-up visits. We compared baseline
characteristics by study arm using chi-squared tests for cate-
gorical outcomes and Wilcoxon rank sum tests for continuous
test, anda Coxproportional hazards model toestimatetheeffect
of the intervention on the incidence of sustained vaginal health
repeated our analysis using the definitions of R2 and R4
consecutive follow-up visits with a healthy vaginal environment.
All statistical tests used a 2-sided a of 0.05. Analyses were con-
ducted using Stata version 11.0 (StataCorp).
This trial screened 378 women, of whom 310 were enrolled.
Demographic and clinicalcharacteristics ofenrolled participants
were presented previously, and were similar between study
arms . The median age, duration of sex work, and number of
sexual episodes in the past week were: 32 years (interquartile
range[IQR], 27–38y), 4 years (IQR, 1–9 y), and 1 episode (IQR,
0–2 episodes), respectively. Condom use was high, with
a median of 100% use (IQR, 0%–100%) among those who re-
ported sex in the past week. Most women (266 women [94%])
reported vaginal washing in the past week, 88 (29%) reported
using hormonal contraception, and 106 (35%) had a healthy
vaginal environment at enrollment. Three hundred and two
women returned for R1 follow-up visit (151 women in each
arm). The number of follow-up visits attended was also similar
between study arms (median [IQR] in the intervention arm
5 12 [7–12] versus 12 [9–12] in the placebo arm; P 5 .7). The
intervention was well tolerated with similar rates of adverse
events reported by arm .
The frequency of a healthy vaginal environment is presented
in Table 1. The proportion of women who had a healthy vaginal
environment at every visit was identical between the study arms
(6 [4%] women in each arm, P 5 1.0). Conversely, the pro-
portion of womenwho never had a healthyvaginal environment
was lower intheintervention arm than in the placebo (9[6%] vs
24[16%];P 5.006). Women intheinterventionarm weremore
likely to have a healthy vaginal environment at any visit than
were women in the placebo arm (hazard ratio [HR], 1.36, 95%
confidence interval [CI] 1.17–1.58) (Table 1).
We also conducted an exploratory analysis, stratifying by
vaginal conditions present at enrollment (Table 1). Among
women who had a healthy vaginal environment at enrollment,
the intervention had no effect on the presence of a healthy
vaginal environment during follow-up. However, among
women who did not have a healthy vaginal environment at
enrollment, those in the intervention arm had an increased in-
cidence of a healthy vaginal environment during follow-up
compared with placebo.
The analysis of sustained vaginal health included 284 women
who attended R3 follow-up visits, comprising 142 women in
each arm. Women included in this analysis were slightly older
(median age, 32 y vs 28 y; P 5 .002) and reported a longer
duration of sex work (median duration, 4 y vs 2 y; P , .001)
compared with women who were excluded due to ,3 follow-up
visits. Baseline characteristics were similar by study arm among
the subgroup who attended R3 follow-up visits (data not
shown). There were 151 episodes of sustained vaginal health
(Table 1). Median time to sustained vaginal health was 30 weeks
in the intervention arm versus 56 weeks in the placebo arm (log-
rank test P 5 .001) (Figure 1). Women in the intervention arm
were more likely to havesustained vaginal healththan womenin
the placebo arm (HR, 1.69; 95% CI, 1.23–2.33). Results were
similar in sensitivity analyses that assessed the effect of the
intervention using shorter (R2 mo) or longer (R4 mo) periods
to define sustained vaginal health (data not shown).
In this secondary analysis, monthly oral treatment with metro-
nidazole and fluconazole increased both the frequency of having
d JID 2011:204 (15 July)
d BRIEF REPORT
a healthy vaginal environment and the incidence of sustained
vaginal health compared with placebo. In addition, we observed
that the impact of the intervention was greatest among women
lacking a healthy vaginal environment at enrollment. This
analysis complements the primary RCT analysis that reported
a decrease in theincidence of BV and an increase in Lactobacillus
colonization . Interventions that promote the establishment
and maintenance of a healthy vaginal environment are impor-
tant for potentially reducing susceptibility to HIV-1 and STIs.
The issue of sustained vaginal health is of critical importance
given theprevalence ofvaginal infectionsand thefrequencywith
which they recur. A prospective study of 121 women treated for
symptomatic BV reported that 23% experienced another epi-
sode of symptomatic BV within 1 month and 58% experienced
a recurrence by month 12 . In addition, vaginal infections
can occur in combination (mixed infections) or sequentially.
T. vaginalis and BV are frequently detected together , and
several studies have reported an association between BV treat-
ment with metronidazole and subsequent symptomatic VVC
[12, 14]. Finally, when considering the question of vaginal
health, the absence of vaginal infections does not guarantee
the presence of healthy vaginal flora; abnormal vaginal flora
(Nugent score 4–6) has also been associated with increased risk
of HIV-1 acquisition .
In both arms of the study, we observed equal proportions of
women who experienced a healthy vaginal environment at every
follow-up visit. It is possible that microbiologic, immunologic,
behavioral, or other host factors that are unique to these women
may contribute to reduced susceptibility to abnormal flora and
vaginal infections. The identification of factors that are associ-
ated with prolonged vaginal health could inform the de-
velopment of future vaginal health interventions.
With repeated use of antibiotic regimens, it is important to
consider the issue of antimicrobial resistance. A study of
resistance patterns associated with BV treatment reported met-
ronidazole resistance in ,1% of anaerobic bacterial isolates
suggest that metronidazole resistance is likely to be infrequent.
The high rate of retention in the trial, combined with
monthly measurement of biological outcomes, allowed us to
gain a substantial level of precision in this longitudinal
characterization of vaginal health. Nonetheless, these findings
should be interpreted in the context of several limitations.
First, this is a secondary analysis in which our sustained
vaginal health analysis used a subgroup of participants from
the RCT, potentially introducing bias. Because the endpoint
Table 1. Frequency of a Healthy Vaginal Environment and Sustained Vaginal Health by Study Arm
Occurrences Rate per 100 person-years (95% CI)Hazard ratio (95% CI)
Healthy vaginal environmenta
Intervention (n 5 151)
Placebo (n 5 151)
815 608 (568–652) 1.36 (1.17–1.58)
635 454 (420–491)1.0 (Reference)
Women with a healthy vaginal environment at enrollment
Intervention (n 5 59)
Placebo (n 5 47)
365 676 (610–749) 0.99 (0.84–1.17)
293 679 (605–761) 1.0 (Reference)
Women without a healthy vaginal environment at enrollment
Intervention (n 5 92)
Placebo (n 5 104)
Sustained vaginal healthb
Intervention (n 5 142)
Placebo (n 5 142)
450563 (513–617) 1.63 (1.32–2.02)
342 354 (318–393) 1.0 (Reference)
86 102 (83–126)1.69 (1.23–2.33)
65 63 (49–80)1.0 (Reference)
NOTE. CI, confidence interval.
aNugent score of 0–3 with no yeast on wet mount and no Trichomonas vaginalis on wet mount or culture.
bHealthy vaginal environment at R3 consecutive follow-up visits.
d JID 2011:204 (15 July)
of sustained vaginal health required 3 months of follow-up, Download full-text
the analysis necessarily restricted the population to women
who had R3 follow-up evaluations. Second, our definition of
sustained vaginal health was somewhat arbitrary. We hy-
pothesized that 3 consecutive visits with a healthy vaginal
environment would reflect a more substantial shift in the
vaginal flora than would 2 visits. Although the clinical im-
plications of having sustained vaginal health have not yet been
explored, we are reassured by the consistency of the findings
observed in the sensitivity analyses. Third, wet mount mi-
croscopy alone was used to diagnose VVC. The combination
of wet mount with culture may provide greater sensitivity and
specificity. Lastly, behavioral characteristics that are unique to
the study population could limit generalizability. Women in
our study reported high rates of vaginal washing and condom
use. It is possible that the effect of the intervention may differ
in other populations.
There is mounting evidence that disruptions in the vaginal
environment due to vaginal infections and abnormal flora in-
crease susceptibility to HIV-1 infection. Among women at risk
for HIV-1 infection, PPT could be used to promote vaginal
health and potentially reduce susceptibility to HIV-1. As anti-
retroviral-based HIV-1–prevention strategies such as Tenofovir
gel  continue to be evaluated, interventions that promote or
preserve a healthy vaginal environment could be used to aug-
ment the effect of these HIV-1–prevention strategies, especially
if antiretroviral-based strategies provide only partial protection.
Additional studies are needed to assess the effects of a healthy
vaginal environment and sustained vaginal health on risk of
HIV-1 and STI acquisition.
Funding for this study was provided by the National Institutes of
Health (grant K23 AI52480); the University of Washington Center for
AIDS and STDs (grant T32 AI007140-32, to J. E. B.); and the University of
Washington Center for AIDS Research, an NIH funded program (grant
P30 AI027757) that is supported by the following NIH Institutes and
Centers: National Institute of Allergy and Infectious Diseases, National
Cancer Institute, National Institute of Mental Health, National Institute
on Drug Abuse, National Institute of Child Health and Human De-
velopment, National Heart, Lung and Blood Institute, and National
Center for Complementary and Alternative Medicine.
1. Van Der Pol B, Kwok C, Pierre-Louis B, et al. Trichomonas vaginalis
infection and human immunodeficiency virus acquisition in African
women. J Infect Dis 2008; 197:548–54.
2. Atashili J, Poole C, Ndumbe PM, Adimora AA, Smith JS. Bacterial
vaginosis and HIV acquisition: A meta-analysis of published studies.
AIDS 2008; 22:1493–501.
3. van de Wijgert JH, Morrison CS, Cornelisse PG, et al. Bacterial
vaginosis and vaginal yeast, but not vaginal cleansing, increase HIV-1
acquisition in African women. J Acquir Immune Defic Syndr 2008;
4. Cherpes TL, Meyn LA, Krohn MA, Lurie JG, Hillier SL. Association
between acquisition of herpes simplex virus type 2 in women and
bacterial vaginosis. Clin Infect Dis 2003; 3:319–25.
5. Schwebke JR, Desmond R. A randomized trial of metronidazole in
asymptomatic bacterial vaginosis to prevent the acquisition of sexually
transmitted diseases. Am J Obstet Gynecol 2007; 196:517.e1–6.
6. van de Wijgert JH, Morrison CS, Brown J, et al. Disentangling
contributions of reproductive tract infections to HIV acquisition in
African women. Sex Transm Dis 2009; 36:357–64.
7. Martin HL, Richardson BA, Nyange PM, et al. Vaginal lactobacilli,
microbial flora, and risk of human immunodeficiency virus type 1 and
sexually transmitted disease acquisition. J Infect Dis 1999; 180:1863–8.
8. McClelland RS, Richardson BA, Hassan WM, et al. Improvement of
vaginal health for Kenyan women at risk for acquisition of human
immunodeficiency virus type 1: Results of a randomized trial. J Infect
Dis 2008; 197:1361–8.
9. Martin HL Jr, Jackson DJ, Mandaliya K, et al. Preparation for AIDS
vaccine evaluation in Mombasa, Kenya: Establishment of seronegative
cohorts of commercial sex workers and trucking company employees.
AIDS Res Hum Retroviruses 1994; 10:S235–7.
10. World Health Organization (WHO). Guideline for the management of
sexually transmitted infections. Geneva: WHO, 2003.
11. 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; 29:297–301.
12. Bradshaw CS, Morton AN, Hocking J, et al. High recurrence rates of
bacterial vaginosis over the course of 12 months after oral metroni-
dazoletherapyandfactors associatedwith recurrence.J Infect Dis2006;
13. McClelland RS, Sangare L, Hassan WM, et al. Infection with Tricho-
monas vaginalis increases the risk of HIV-1 acquisition. J Infect Dis
14. Sobel JD, Ferris D, Schwebke J, et al. Suppressive antibacterial therapy
with 0.75% metronidazole vaginal gel to prevent recurrent bacterial
vaginosis. Am J Obstet Gynecol 2006; 194:1283–9.
15. Beigi RH, Austin MN, Meyn LA, Krohn MA, Hillier SL. Antimicrobial
resistance associated with the treatment of bacterial vaginosis. Am
J Obstet Gynecol 2004; 191:1124–9.
16. AbdoolKarim Q, AbdoolKarim SS,Frohlich JA, et al. Effectiveness and
safety of tenofovir gel, an antiretroviral microbicide, for the prevention
of HIV infection in women. Science 2010; 329:1168–74.
d JID 2011:204 (15 July)
d BRIEF REPORT