Secondary Logo

Journal Logo

Review

Trichomonas vaginalis as a Cause of Perinatal Morbidity: A Systematic Review and Meta-Analysis

Silver, Bronwyn J. MPH*†; Guy, Rebecca J. PhD; Kaldor, John M. PhD; Jamil, Muhammad S. MPH; Rumbold, Alice R. PhD

Author Information
Sexually Transmitted Diseases: June 2014 - Volume 41 - Issue 6 - p 369-376
doi: 10.1097/OLQ.0000000000000134

Trichomoniasis is a sexually transmissible infection (STI) caused by the parasite Trichomonas vaginalis. It is the most common curable STI worldwide.1 Prevalence of T. vaginalis is consistently higher in women than in men and varies substantially across populations, showing a strong association with health and social disadvantage.2,3 It affects women in their peak reproductive years and has been estimated that up to 25 million pregnant women globally are infected with T. vaginalis.4 In women, common symptoms of trichomoniasis include vaginal discharge and itching; however, many will have no symptoms.2

Research conducted in the 1980s found associations between trichomoniasis during pregnancy and adverse pregnancy outcomes, including preterm birth,5 low birth weight (LBW),5,6 and premature rupture of membranes (PROM).7 This research led to clinicians being encouraged to screen and treat infections in pregnancy.8 However, findings were not consistent across all studies. Also, in a randomized trial9 conducted in the late 1990s, asymptomatic women assigned to treatment with metronidazole found an increased risk of preterm birth than in the placebo group. As a result of these studies, practice changed. Currently, screening of asymptomatic women is not routinely recommended during pregnancy, except in endemic or high-risk populations; diagnostic testing is recommended in women who are symptomatic.10 Treatment with metronidazole in pregnancy is currently only advised in symptomatic cases or if asymptomatic, after 37 weeks’ gestation.10

Questions still remain about the role of trichomoniasis in perinatal morbidity. If the infection is indeed associated with preterm birth or other morbidities, strategies need to be found to detect and safely treat it. The studies that have found a relationship may have been compromised by confounding due to one of the many factors that determine pregnancy outcome. On the other hand, the apparently negative studies may have suffered from methodological problems such as misclassification of infection status or outcome, both of which would have diluted any real effect. Given the diversity in research findings, we undertook a systematic review of studies that investigated preterm birth and other perinatal outcomes in pregnant women infected with T. vaginalis and women who were uninfected.

METHODS

This review used the Preferred Reporting Items for Systematic Reviews and Meta Analyses guidelines,11 the Meta-analysis of Observational Studies in Epidemiology criteria,12 and the National Institute for Health and Clinical Excellence guidelines.13

Search Strategy

We searched Medline, EMBASE, and BioMedCentral to 16th May 2013 using the following terms: “Trichomonas” or “Trichomonas vaginalis” AND “adverse pregnancy outcome” or “premature birth” or “premature delivery” or “premature labour” or “premature rupture of membranes” or “low birth weight” or “intrauterine growth retardation” or “small for gestational age” or “gestational age.”

Study Selection

Studies were included if they were published before 16th May 2013 and assessed 1 or more of the specified outcomes in women with T. vaginalis and a control group of women without the infection. The following perinatal outcomes were considered: preterm birth, preterm PROM (PPROM), PROM, LBW, small for gestational age (SGA) infant, stillbirth, neonatal death, and postpartum endometritis. Only full-text English-language articles were included. Studies were excluded if they did not present original data.

The titles and abstracts of all articles were reviewed. If the abstract seemed to meet the inclusion criteria, the full-text article was obtained and reviewed. Reference lists of publications were checked for other potentially relevant studies. Two reviewers (B.S. and A.R.) undertook this process independently and then reached consensus on the final set of studies for inclusion.

Data Collection, Synthesis, and Statistical Analysis

For each study that met the inclusion criteria, information was extracted on the year of study, study design, setting, study population, methods of determining T. vaginalis status, study outcomes, and statistical methods, including adjustment for potential confounders. Where relevant, authors were contacted to provide missing or additional information.

For outcomes reported by 2 or more studies, we pooled data to generate combined estimates of relative risk (RR). The I2 test was used to estimate the proportion of total variability in point estimates that could be attributed to heterogeneity other than that due to chance. If the I2 value was less than 25%, we used a fixed-effects meta-analysis to estimate the combined RR (and 95% confidence interval [CI]), assuming that most between study variability was due to chance.14 If the I2 value was 25% to 75%, a random-effects meta-analysis was used. If the I2 value was greater than 75%, the heterogeneity was considered too great for a summary estimate to be calculated.15 Data were analyzed using STATA 11.2.16

The initial analyses included all studies. For the outcome preterm birth, sensitivity analyses were carried out, which involved excluding studies with very low numbers of T. vaginalis infection. Sensitivity analyses were also carried out to explore the effect of coinfection with other STI and treatment of trichomoniasis. This involved analysis based on studies that had adjusted for STI coinfection, and separately, studies that specified that no treatment was given during pregnancy.

Quality Appraisal and Assessment of Publication Bias

The methodological quality of each study was independently reviewed by 2 authors (B.S. and A.R.) using the National Institute for Health and Clinical Excellence guidelines.13 This involved completing a checklist for each study design that evaluates 4 potential sources of bias: selection bias, performance bias, attrition bias, and detection bias. For each type of bias, the reviewers classified the study as one of the following: low, high, or unclear/unknown risk of bias. An overall risk of bias was determined independently by assessing the major direction of bias and was then discussed and agreed on by 2 authors (B.S. and A.R.). The potential presence of publication bias was assessed using a funnel plot; asymmetry was evaluated visually. We undertook a sensitivity analysis to investigate the impact of removing any studies that seemed to be outliers in the funnel plot analysis.

RESULTS

Selection of Studies

The literature search identified 178 articles, of which 127 were excluded on the basis of information conveyed by the title and abstract (Fig. 1). The remaining 51 articles were reviewed, and of these, 12 reporting on 11 separate studies met the inclusion criteria (Fig. 1).

Figure 1
Figure 1:
Study selection process.

Study Characteristics

Table 1 describes the characteristics of included studies. Briefly, the size of studies varied from 115 to 60,296 (median, 1038), with data collection spanning from 1973 to 2003. Only one study focused on “adolescents” (defined as 13–17 years), the remaining studies recruited women of any “reproductive age” (e.g., 13–49 years).

TABLE 1
TABLE 1:
Characteristics of Included Studies

Most (n = 7) studies were conducted in general antenatal clinics. Four studies classified their population as being at high risk for adverse pregnancy outcomes, based on the presence of one or more of the following risk factors: young age (<18 years), African American or Hispanic ethnicity, unmarried, of low socioeconomic status, uninsured, or had a low income. For the remaining studies, the study population included women who were unselected in regard to risk of adverse pregnancy outcome. Most (n = 8) studies excluded women with multiple pregnancies, and 4 studies included statistical adjustments for a previous preterm birth. Of the 9 studies that reported on preterm birth, 8 excluded multiple pregnancies, 2 excluded women with an existing obstetric or medical condition known to predict preterm birth, and 5 made adjustments to account for differences in smoking and/or other recognized risk factors for preterm birth.

The most frequently reported outcome was preterm birth. Eight studies defined preterm birth as gestation less than 37 weeks and one study as gestation less than 36 weeks. Only one study reported on preterm and very preterm birth. Four studies reported on PROM and LBW. The remaining outcomes were reported by 1 or 2 studies each. The methods used to estimate gestational age varied across the 11 studies (see Table S1, http://links.lww.com/OLQ/A85), with the most frequently used method being based on the date of the last menstrual period and ultrasound (where available). Five studies did not provide any information on method used.

The risk of methodological bias was rated as unclear for most (n = 7) studies (Table S1, http://links.lww.com/OLQ/A85), mostly due to a lack of information regarding performance and detection bias. Three studies17–19 were rated as being at low risk for bias and one rated at high risk.20 A key issue was potential confounding by other STI; coinfection was considered in the analyses of 6 studies, no statistical adjustment was made in 4 studies, and it was unclear in 1 study. Three studies reported that treatment was given to women with trichomoniasis (although none reported the specific regimen), 4 studies stated that no treatment was given, and a further 4 studies stated that it was unclear whether treatment was given.

The timing of screening for trichomoniasis varied, with 2 studies collecting specimens during the first trimester only, 3 in the second only, 1 in the third trimester, 3 allowing for screening in any trimester, and 2 not stating when screening was done. In 2 studies, repeat specimens were collected on all women closer to birth regardless of infection status, and in 1 study, positive women only were repeat tested. The method of diagnosis was microscopy in 4 studies, culture in 1, a combination of both in 5, and unspecified in 1.

Perinatal Outcomes

Nine studies (n = 81,001), including 4 cohort studies, reported on the RR of preterm birth. Because heterogeneity was moderate (I2 = 62.7%, P = 0.006), a random-effects analysis was used, giving a summary RR of 1.42 (95% CI, 1.15–1.75; P = 0.001; Fig. 2).

Figure 2
Figure 2:
Meta-analysis of the association betweenT. vaginalis and preterm birth.

In the sensitivity analysis that excluded 2 studies21,22 with very low numbers of women with trichomoniasis, the summary RR for preterm birth was slightly lower, at 1.28, but remained statistically significant (95% CI, 1.09–1.51; P = 0.003; 7 studies; n = 79,674; I2 = 45.2%). In the analyses that included the 6 studies (n = 20,646) that adjusted for STI coinfection (Table S1, http://links.lww.com/OLQ/A85), the summary RR was 1.34 (95% CI, 1.19–1.51; P ≤ 0.001; I2 = 11.2%). Restricting analyses to the 3 studies (n = 1795) in which no treatment was provided for trichomoniasis (Table S1, http://links.lww.com/OLQ/A85) gave a summary RR of 1.82, which was not statistically significant (95% CI, 0.98–3.41; P = 0.058; I2 = 22.3%). A RR was not calculated for studies of treated women, because in the included studies, it was unclear if all or only some participants were treated.

Two cohort studies (n = 14,843) reported on PPROM. Because there was negligible heterogeneity (I2 = 0.0%, P = 0.859), a fixed-effects analysis was used, giving a summary RR of 1.41 (95% CI, 1.10–1.82; P = 0.007; Fig. 3). Four cohort studies (n = 14,754) reported on PROM including 2 that found a statistically significant increase in risk and 1 that did not. For the fourth study, the 2 women who were positive for T. vaginalis both had PROM,21 so the RR could not be calculated. A summary RR could not be calculated because of a high level of heterogeneity (I2 = 93.9%, P ≤ 0.001).

Figure 3
Figure 3:
Meta-analysis of the association betweenT. vaginalis and preterm PROM.

Four studies reported on LBW (n = 16,908), including 2 that found a statistically significant increase in risk, but a summary RR was not calculated because of significant heterogeneity (I2 = 77.5%, P = 0.004). Two studies reported on SGA infants (n = 72,077) with negligible heterogeneity (I2 = 0.0%, P = 0.631), so a fixed-effect analysis was used, giving a summary RR of 1.51 (95% CI, 1.32–1.73; P ≤ 0.001; Fig. 4).

Figure 4
Figure 4:
Meta-analysis of the association betweenT. vaginalis and SGA.

Two studies (n = 74,569) reported on stillbirth including one that found a significant increase in risk, but no summary RR was calculated because of significant heterogeneity (I2 = 90.0%, P = 0.002). One cohort study reported on the association between T. vaginalis and neonatal death and reported a significant association (1.6% vs. 0.8%; P = 0.005; 13,816 women). One cohort study (n = 13,814) reported on the association between T. vaginalis and postpartum endometritis and reported a significant difference between rates of 6.9% in infected women compared with 4.7% in uninfected women.

Publication Bias

A funnel plot to assess publication bias is presented in Figure S1, http://links.lww.com/OLQ/A85. The graph plots the log risk ratios for preterm birth against trial size as measured by standard error of the log risk ratio. One study21 was removed before the analysis due to small exposure numbers (n = 2). In the bottom right quadrant, there is one small study with a large standard error suggesting that the point estimate in this study may be inflated, possibly because of poor methodological quality. Of note, this study was removed in the sensitivity analyses presented earlier. The plot also shows some asymmetry around the log of the pooled estimate, with fewer studies with point estimates below the pooled estimate and no studies in the bottom left quadrant. This indicates that small studies demonstrating no association between trichomoniasis and preterm birth may be missing.

DISCUSSION

To our knowledge, this is the first comprehensive review and meta-analysis of the relationship between the presence of infection with T. vaginalis and perinatal morbidity. The most commonly reported end point, preterm birth, was 42% more likely (95% CI, 15%–75%) to occur in infected women. Based on much fewer studies, there were also substantial increases in the risk of PPROM (increase in risk, 41%; 95% CI, 10%–82%) and having a SGA baby (increase of 51%; 95% CI, 32%–73%). Because of insufficient data and/or concerns about heterogeneity, we were unable to draw reliable conclusions about the impact of T. vaginalis on other outcomes including PROM, LBW, still birth, neonatal death, or postpartum endometritis.

We followed a systematic process to identify studies and extract data. It is nevertheless possible that we missed 1 or more studies. Data were abstracted by 2 independent reviewers, to minimize any subjectivity in coding. We also undertook a detailed assessment of the quality of each study.

There are several limitations. Moderate heterogeneity was detected for the outcome preterm birth, and significant heterogeneity prevented a summary measure being calculated for several important outcomes. The statistical heterogeneity found may reflect differences in the design of eligible studies as well as the characteristics of participants. The proportion of preterm births in the control groups ranged from 5% to 33%, suggesting a range in baseline maternal risk in eligible studies. There was variation in the degree to which studies accounted for maternal and fetal characteristics that predict preterm birth and other potential confounders, either through the exclusion criteria or through statistical adjustment.

The funnel plot generated for preterm birth (Figure S1, http://links.lww.com/OLQ/A85) indicated that there may be publication bias, such that small studies that failed to show an association between trichomoniasis and preterm birth remain unpublished. However, it is also possible that the asymmetry seen in the plot could be due to differences in the characteristics of women enrolled in the smaller and larger studies rather than publication bias.

A major limitation in all of the eligible studies is the potential misclassification of exposure to T. vaginalis (Table S1, http://links.lww.com/OLQ/A85). All studies were conducted at least 10 years ago and therefore depended on diagnostic techniques such as wet mount microscopy, which may have resulted in an underdetection of T. vaginalis. The more recently developed nucleic acid amplification assays have been shown to almost double the detection of T. vaginalis compared with wet mount microscopy.23 Further misclassification of exposure could arise because of the timing of swab collection, which was mostly at first visit or first trimester, and may have missed infection acquired subsequently.

Several of the included studies did not rule out coinfection with other STIs. In particular, bacterial vaginosis24 and chlamydia25 have both been found to be strong risk factors for preterm birth, and their presence is likely to be correlated with T. vaginalis. However, in the sensitivity analyses that excluded studies that did not account for STI coinfection, the pooled estimate for preterm birth did not change substantially. Therefore, we believe that the observed effect for preterm birth is unlikely to be explained by confounding due to the presence of other infections.

Despite these potential limitations, the finding of a strong relationship between trichomoniasis and preterm birth has very substantial public health implications. Preterm birth is a leading cause of infant mortality26 and has enduring consequences for the health and development of surviving infants.27 The magnitude of the association between preterm birth and trichomoniasis is similar to estimates for other major preventable risk factors such as maternal smoking (odds ratio, 1.27; 95% CI, 1.21–1.33)28 for which interventions have become routine in antenatal care.

The mechanisms underlying the association between trichomoniasis and preterm birth are not well understood but are likely to be linked to a maternal innate immune inflammatory response after infection. Elevated concentrations of cervical interleukin-8 and vaginal defensins have been demonstrated in pregnant women with asymptomatic trichomoniasis.29 Both are markers of neutrophil activation, which has been associated with amniotic fluid infection, preterm birth, and preterm PROM.30–32 In particular, cervicovaginal interleukin-8 is considered to play a critical role in triggering cervical ripening and dilatation, by recruiting neutrophils to cervical tissue resulting in the release of collagenase and elastase.33 Our study also found an increased risk of SGA infants, of which the pathogenesis is unknown and may be quite different to preterm birth. Although speculative, we suggest that trichomoniasis may result in chronic low-grade intrauterine inflammation that interferes with uteroplacental circulation.

Our findings raise important questions about the management of trichomoniasis in pregnancy. First, in settings of high endemnicity, should all women be screened in pregnancy? If so, should symptomatic and asymptomatic infected women be treated? The importance of these questions becomes more pronounced with higher levels of T. vaginalis prevalence. To determine the ideal policy for screening in pregnancy, we must consider prevalence, outcome, and treatment safety.

A single dose of 2 mg metronidazole provides parasitological cure34 and has not been associated with any teratogenic effects in pregnancy,35 even when given in the first trimester. However, concerns about the safety of treatment in pregnancy arose after the publication of 2 studies a decade ago that reported an increased risk of preterm birth and/or LBW associated with metronidazole treatment. Both studies had potential methodological limitations. Klebanoff and colleagues9 administered a dose 4 times higher than the recommended regimen and the trial was stopped early after reaching a third of the target sample when an interim analysis revealed an increased risk of preterm birth in the treatment arm. Kigozi and colleagues36 reported on a small subgroup of women with trichomoniasis participating in a trial of a combination of antimicrobials to prevent perinatal HIV transmission. Conversely, a recent chart review37 of metronidazole use in pregnancy among 2829 women found no association with preterm birth or LBW. There has been very little subsequent research investigating appropriate strategies for managing trichomoniasis in pregnancy with most treatment studies focusing on bacterial vaginosis,38 which is likely to elicit a different immunological response to T. vaginalis. Therefore, whether there are indeed risks or benefits associated with treatment in pregnancy remains unclear,39 and further studies are needed to answer this important question40 to ensure clinical practice and guidelines are supported by a solid evidence base.

Future research should be conducted in a high-prevalence setting, where the impact of policy change will be greatest, and there will be adequate numbers of infected women to provide statistical accuracy. It is recommended that a trial of routine screening and treatment of T. vaginalis is conducted that includes either a strict exclusion criteria for known confounders or, where unfeasible, statistical adjustment. Using a highly sensitive diagnostic technology such as nucleic acid amplification assays would ensure all infections were detected. Swab collection at multiple time points throughout pregnancy and measurement of organism load would also need to be considered. Data should be collected on a comprehensive range of perinatal outcomes, and statistical approaches should be used to take account of treatment of T. vaginalis and STI coinfection, in particular bacterial vaginosis.

CONCLUSIONS

Although our review provides strong evidence that T. vaginalis in pregnancy is associated with preterm birth, there remains a gap in evidence on treatment effects in pregnancy. If treatment is found to be effective, this represents a significant opportunity to make inroads into the prevention of perinatal morbidity in populations where trichomoniasis is endemic. Preventing the acquisition of T. vaginalis, as well as early detection and treatment, should remain a priority.

REFERENCES

1. World Health Organization. Prevalence and Incidence of Selected Sexually Transmitted Infections, Chlamydia trachomatis, Neisseria gonorrhoeae, Syphilis and Trichomonas vaginalis: Methods and Results Used by WHO to Generate 2005 Estimates. Geneva, Switzerland: World Health Organization, 2011.
2. Hobbs M, Sena A, Swygard H, et al. Trichomonas vaginalis and trichomoniasis. In: Holmes K, Sparling P, Stamm W, eds. Sexually Transmitted Diseases. New York: McGraw Hill, 2008: 771–794.
3. Poole DN, Scott McClelland R. Global epidemiology of Trichomonas vaginalis. Sex Transm Infect 2013; 89: 418–422.
4. Walker G. Interventions for Trichomoniasis in Pregnancy: RHL Commentary. The WHO Reproductive Health Library. Geneva, Switzerland: World Health Organization, 2004.
5. Cotch M, Pastorek I, Nugent R, et al. Trichomonas vaginalis associated with low birth weight and preterm delivery. Sex Transm Dis 1997; 24: 353–360.
6. Hardy P, Hardy J, Nell E, et al. Prevalence of six sexually transmitted disease agents among pregnant inner-city adolescents and pregnancy outcome. Lancet 1984; 2: 333–337.
7. Minkoff H, Grunebaum A, Schwarz R, et al. Risk factors for prematurity and premature rupture of membranes: A prospective study of the vaginal flora in pregnancy. Am J Obstet Gynecol 1984; 150: 965–972.
8. World Health Organization. Guidelines for the Management of Sexually Transmitted Infections. Geneva, Switzerland: World Health Organization, 2003.
9. Klebanoff M, Carey J, Hauth J, et al. Failure of metronidazole to prevent preterm delivery among pregnant women with asymptomatic Trichomonas vaginalis infection. N Engl J Med 2001; 345: 487–493.
10. Centers for Disease Control and Prevnetion. Sexually transmitted diseases treatment guidelines. MMWR 2010; 59: 58–60.
11. Moher D, Liberati A, Tetzlaff J, et al. The PRISMA Group (2009) Preferred Reporting Items for Systematic Reviews and Meta-Analyses: The PRISMA statement. PLoS Med 2009; 6: e1000097.
12. Stroup D, Berlin J, Morton S, et al. Meta-analysis of observational studies in epidemiology. A proposal for reporting. JAMA 2000; 283: 2008–2012.
13. National Institute for Health and Clinical Excellence. The Guidelines Manual. London: National Institute for Health and Clinical Excellence, 2012. Available at: http://www.nice.org.uk/guidelinesmanual.
14. Higgins J, Thompson SG, Deeks JJ, et al. Measuring inconsistency in meta-analyses. BMJ 2003; 327: 557–560.
15. Riley RD, Jiggins JPT, Deeks JJ, et al. Interpretation of random effects meta-analyses. BMJ 2011; 342: 964–967.
16. Statistics/Data Analysis (computer program). 10.1 edition. Texas, USA: STATA Corporation; 2008.
17. Johnson H, Ghanem K, Zenilmean J, et al. Sexually transmitted infections and adverse pregnancy outcomes among women attending inner city public sexually transmitted disease clinics. Sex Transm Dis 2011; 38: 167–171.
18. Buchmayer S, Sparen P, Cnattingius S. Signs of infection in Pap smears and risk of adverse pregnancy outcome. Paediatr Perinat Epidemiol 2003; 17: 340–346.
19. Azargoon A, Darvishzadeh S. Association of bacterial vaginosis, trichomonas vaginalis, and vaginal acidity with outcome of pregnancy. Arch Iran Med 2006; 9: 213–217.
20. Mathai E, Muthaiah A, Mathai M, et al. Prevalence and effects of trichomoniasis in pregnancy. Natl Med J India 1998; 11: 151.
21. Rasti S, Behrashi M, Mousavi G, et al. Complications of trichomoniasis on the pregnant women. Jundishapur J Microbiol 2011; 4: 61–63.
22. Kiss H, Petricevic L, Husslein P. Prospective randomised controlled trial of an infection screening programme to reduce the rate of preterm delivery. BMJ 2014; 329: 371.
23. Nye M, Schwebke J, Body B. Comparison of APTIMA Trichomonas vaginalis transcription-mediated amplification to wet mount microscopy, culture, and polymerase chain reaction for diagnosis of trichomoniasis in men and women. Am J Obstet Gynecol 2009; 200: 188.e181–e188 e187.
24. Leitich H, Bodner-Adler B, Brunbauer M, et al. Bacterial vaginosis as a risk factor for preterm delivery: A meta-analysis. Am J Obstet Gynecol 2003; 189: 139–147.
25. de Attayde Silva M, Dantas Florêncio GL, Erbolato Gabiatti JR, et al. Perinatal morbidity and mortality associated with chlamydial infection: A meta-analysis study. Braz Infect Dis 2011; 15: 533–539.
26. Liu L, Johnson H, Cousens S, et al. Global, regional, and national causes of child mortality: An updated systematic analysis for 2010 with time trends since 2000. Lancet 2012; 379: 2151–2161.
27. Saigal S, Doyle L. An overview of mortality and sequelae of preterm birth from infancy to adulthood. Lancet 2008; 371: 261–269.
28. Shah N, Bracken M. A systematic review and meta-analysis of prospective studies on the association between maternal cigarette smoking and preterm delivery. Am J Obstet Gynecol 2000; 182: 465–472.
29. Simhan HN, Anderson BL, Krohn MA, et al. Host immune consequences of asymptomatic Trichomonas vaginalis infection in pregnancy. Am J Obstet Gynecol 2007; 196: 59.e1–e5.
30. Hitti JM, Hillier SL, Agnew KJ, et al. Vaginal indicators of amniotic fluid infection in preterm labor. Obstet Gynecol 2001; 97: 211–219.
31. Simhan HN, Caritis SN, Krohn MA, et al. Elevated vaginal pH and neutrophils are associated strongly with early spontaneous preterm birth. Am J Obstet Gynecol 2003; 189: 1150–1154.
32. Simhan HN, Caritis SN, Krohn MA, et al. The vaginal inflammatory milieu and the risk of early premature rupture of membranes. Am J Obstet Gynecol 2005; 192: 213–218.
33. Tanaka Y, Narahara H, Takai N, et al. Interleukin-1β and interleukin-8 in cervicovaginal fluid during pregnancy. Am J Obstet Gynecol 1998; 179: 644–649.
34. Lossick J. Single-dose metronidazole treatment for vaginal trichomoniasis. Obstet Gynecol 1980; 56: 508–510.
35. Caro-Paton T, Carvajal A, Martin dD, et al. Is metronidazole teratogenic? A meta-analysis. Br J Clin Pharmacol 1997; 44: 179–182.
36. Kigozi G, Brahmbhatt H, Wabwire-Mangen F, et al. Treatment of trichomonas in pregnancy and adverse outcomes of pregnancy: A subanalysis of a randomized trial in Rakai, Uganda. Am J Obstet Gynecol 2003; 189: 1398–1400.
37. Koss C, Baras D, Lane S, et al. Investigation of metronidazole use during pregnancy and adverse birth outcomes. Antimicrob Agents Chemother 2012; 56: 4800–4805.
38. Brocklehurst P, Gordon A, Heatley E, et al. Antibiotics for treating bacterial vaginosis in pregnancy. Cochrane Database Syst Rev 2013; 1:DOI:10.1002/14651858.CD000262.pub4.
39. Gülmezoglu AM, Azhar M. Interventions for trichomoniasis in pregnancy. Cochrane Database Syst Rev 2011; 5:DOI:10.1002/14651858.CD000220.pub2.
40. Muzny C, Schewebke J. The clinical spectrum of Trichomonas vaginalis infection and challenges to management. Sex Transm Infect 2013; 89: 423–425.
41. Germain M, Krohn M, Hillier S, et al. Genital flora in pregnancy and its association with intrauterine growth retardation. J Clin Microbiol 1994; 32: 2162–2168.
42. French J, McGregor J, Parker R. Readily treatable reproductive tract infections and preterm birth among black women. Am J Obstet Gynecol 2006; 194: 1717–1726.
43. Stringer E, Read JS, Hoffman I, et al. Treatment of trichomoniasis in pregnancy in sub-Saharan Africa does not appear to be associated with low birth weight or preterm birth. S Af Med J 2010; 100: 58–64.

    Supplemental Digital Content

    © Copyright 2014 American Sexually Transmitted Diseases Association