| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
ORIGINAL RESEARCH |
From the Department of Obstetrics and Gynecology at Denver Health Medical Center, and the University of Colorado Health Sciences Center, Denver, Colorado.
Address reprint requests to: Janice I. French, CNM, MS Department of Obstetrics and Gynecology Denver Health Medical Center, Code 0660 777 Bannock Street Denver, CO 80204 E-mail: jan.french{at}dhha.org
 |
Abstract |
|---|
Top
 Abstract Material and MethodsResultsDiscussionReferences |
|---|
Methods: A secondary analysis was conducted of 1100 pregnant women enrolled in a prospective observational study that examined the effects of standardized diagnosis and treatment of lower genital tract infections to prevent preterm birth.
Results: Sixty percent of women with first-trimester bleeding had one or more study infections detected at the initial examination. First-trimester bleeding was associated independently with the presence of bacterial vaginosis (odds ratio [OR] 1.5, 95% confidence interval [CI] 1.0, 2.3), Trichomonas vaginalis (OR 2.3, 95% CI 1.3, 4.2), and Chlamydia trachomatis (OR 2.7, 95% CI 1.4, 5.1). Preterm birth was increased among women with first-trimester bleeding and bacterial vaginosis (relative risk [RR] 4.4, 95% CI 2.0, 9.5) and bacterial vaginosis and T vaginalis (RR 3.0, 95% CI 1.0, 8.8). Systemic antimicrobial treatment reduced the rate of preterm birth among women with bacterial vaginosis without first-trimester bleeding (RR 0.37, 95% CI 0.16, 0.88). Treatment of women with both first-trimester bleeding and bacterial vaginosis reduced preterm birth (RR 0.52, 95% CI 0.18, 1.55), but not significantly.
Conclusion: First-trimester bleeding was increased among women with bacterial vaginosis, T vaginalis, C trachomatis, and combinations of these infections. Women with bacterial vaginosis who also experienced first-trimester bleeding were at heightened risk for preterm birth. Treatment of studied infections reduced significantly the risks of preterm birth among women without first-trimester bleeding.
Preterm birth continues as a major cause of perinatal morbidity and excess health care costs. Both gestational bleeding in the first or second trimesters of pregnancy and presence of bacterial vaginosis and other common genitourinary tract infections have been linked to increased risks of preterm labor, preterm birth, premature rupture of membranes (PROM), low birth weight (LBW), and other adverse pregnancy outcomes.1–11 Gestational bleeding during the first or second trimester occurs in approximately one in eight pregnant women.6 Bacterial vaginosis also is common, occurring among 12–32% of women in various pregnant populations within the United States; trichomoniasis, chlamydia, gonorrhea, and group B streptococcal bacteriuria are less common.8–11
If and how gestational bleeding and common genitourinary tract infections are linked to each other and the risks of preterm birth have not been examined adequately. Early gestational bleeding has been ascribed traditionally to noninflammatory factors such as presence of an abnormal conceptus, uterine anomaly, various hormonal deficiencies, or disruption of the junction between the trophoblast and decidua.2,3 Alternatively, several investigators suggest that ascending reproductive tract infections with presumed subclinical endometritis can affect adversely implantation, early placentation, and growth of the trophoblast and conceptus.12,13 Subclinical endometritis/deciduitis also is thought to cause preterm labor and preterm PROM later in pregnancy.12,13 Finally, microbe-related conditions associated with prematurity (bacterial vaginosis, trichomoniasis, chlamydial infection, and gonorrhea) are all associated with cervicitis, cervical erosion, cervical friability, and/or endometritis and may cause bleeding directly.14,15 In such cases, bleeding may be a marker for the severity of the infection. Bleeding due to cervicitis or deciduitis may impair maternal cervical and lower uterine segment host defense mechanisms (ie, displace cervical mucus and secretory immunoglobulin A antibodies) and/or provide additional nutrients for microbial growth. By disrupting host defense mechanisms or causing uterine contractions, bleeding may enhance ascent of microorganisms into the lower uterine segment early in pregnancy.
Reproductive tract infections are potentially remediable factors that mediate some cases of preterm birth. Evaluation of links between genital tract bleeding and infections may help identify women at heightened risk for associated adverse pregnancy outcomes and allow for focused treatment approaches. This analysis was undertaken to examine the potential role of lower reproductive tract infections as risk factors for early gestational bleeding, to evaluate possible joint effects between gestational bleeding and prevalent genitourinary tract infections and increased risks of adverse pregnancy outcomes, and to determine relative benefits of systematic identification and treatment of common genitourinary tract infections in pregnant women with and without gestational bleeding.
|
Material and Methods |
|---|
|
Top
AbstractMaterial and MethodsResultsDiscussionReferences |
|---|
.
|
The second 8 months of the program (phase II) involved identical screening procedures; however, women with positive findings for bacterial vaginosis, T vaginalis, group B streptococcal bacteriuria as well as N gonorrhoeae and C trachomatis were contacted by program personnel or the clinic staff and provided with specific antimicrobial treatment. Both symptomatic and asymptomatic women received antimicrobial treatment during the intervention phase. Partners of women with sexually transmitted infections were treated similarly. In addition, women were seen by program nurses for test of cure follow-up examinations.
In both phase I and phase II, standard treatments as recommended by the Centers for Disease Control and Prevention were used for N gonorrhoeae, C trachomatis, T vaginalis (metronidazole 500 mg orally, twice daily for 7 days), and bacterial vaginosis (clindamycin 300 mg orally, twice daily for 7 days).17 Women with group B streptococcal bacteriuria were given either penicillin V-K or erythromycin base (500 mg orally, four times daily for 7 days).17 Medical records of mothers and infants were reviewed after delivery.
Bacterial vaginosis was diagnosed using Amsel’s criteria.18 T vaginalis was sought by microscopic examination of a saline wet mount and by culture.16 Patients were asked about the presence or absence of gestational bleeding at each antenatal visit. This information was recorded in their medical records. Subsequently, information regarding gestational bleeding was collected from the medical records and recorded as present or absent for each trimester. Similarly, the patient’s history regarding substance abuse (ie, smoking, alcohol, and street drug use) was collected as present or absent at the initiation of prenatal care. Preterm birth was defined as birth between 22 weeks’ gestation and the completion of 36 weeks’ gestation, in the absence of a clinically recognized medical or obstetric indication for preterm birth (eg, pregnancy-induced hypertension, placenta previa, or abruption). Rupture of fetal membranes before 37 weeks’ gestation and more than 1 hour before the onset of labor among women without a recognized cause of preterm birth (as above) was defined as preterm PROM. Infants weighing less than 2500 g at birth were considered LBW, and those weighing less than 1000 g were considered extremely LBW. Patients who received intrapartum antibiotics for suspected or presumed chorioamnionitis with two or more of the following signs and symptoms: maternal temperature of 37.5C or more, leukocytosis, maternal tachycardia, fetal tachycardia, uterine tenderness, or foul-smelling amniotic fluid were considered to have chorioamnionitis for the study. Records for subjects lost to follow-up were sought by contacting patients, family, and other Denver-area hospitals.
Maternal demographic, historic, and behavioral characteristics as well as microbiologic findings at the start of prenatal care were examined separately for associations with gestational bleeding in each trimester. Categoric data were compared using 
2 tests for proportions and Fisher exact tests with 
set at .05. Crude relative risks (RRs) and 95% confidence intervals (CIs) were calculated to examine the strength of the individual associations.19 Each condition or infection (bacterial vaginosis, T vaginalis, C trachomatis, group B streptococcal bacteriuria, and M hominis) was considered as a possible effect modifier of the association between each of the other individual microorganisms and first-trimester bleeding using stratified analysis. Heterogeneity of the stratum-specific odds ratios (ORs) was considered evidence of effect modification if the Breslow-Day test for heterogeneity was less than .2.19
Adjusted ORs for factors associated with first-trimester bleeding were derived using logistic regression modeling.20 Factors suggested from the literature as well as those associated with first-trimester bleeding using univariate analysis were considered as potential confounding factors (Table 1
). Variables were included for adjustment in the final model of risk factors for first-trimester bleeding if their presence in the logistic models produced at least a 10% change in OR.20 Crude RR and 95% CI also were estimated to evaluate separate associations between first-, second-, and third-trimester bleeding and studied adverse pregnancy outcomes.
Possible joint and independent effects between gestational bleeding in the first trimester, the studied microbiologic factors, and adverse pregnancy outcomes also were evaluated. This was done by creating separate structured variables for each microbial condition that represented the presence of first-trimester bleeding with and without the infection and infection alone compared with those with neither infection nor first-trimester bleeding. The proportion of women with bleeding and no studied infection who delivered preterm was assumed to estimate the direct effects of bleeding on risk of preterm birth. Similarly, those women with infection alone were assumed to estimate the direct effects of the infection on risk of preterm birth. Those with both infection and bleeding were assumed to estimate the risk of preterm birth from the joint effects of bleeding and infection.21
Attributable proportions were calculated to estimate the proportion of preterm births among women exposed to infection with and without first-trimester bleeding that could be attributed to the exposure and potentially reduced if infection and/or first-trimester bleeding were eliminated.19 Further, population-attributable risks were calculated to estimate the proportion of preterm births within the study population that might be eliminated by prevention of the effects of bacterial vaginosis with and without first-trimester bleeding.19 Finally, the effect of treatment for the studied reproductive tract infections among women with and without gestational bleeding was examined using stratified analysis. Data analysis was performed using the Statistical Analysis System for personal computers, version 6.03 (SAS Institute, Cary, NC).
|
Results |
|---|
|
Top
AbstractMaterial and MethodsResultsDiscussionReferences |
|---|
First-trimester antenatal bleeding was recorded for 121 women (11%); second-trimester bleeding was noted among 69 women (6.4%); and third-trimester bleeding was recorded for 43 women (4.1%). Bleeding in both first and second trimester was noted for 17 (1.6%) women. Bleeding was recorded more commonly among black (12.8%), Hispanic (12.4%), and native American women (22.6%) compared with those who were white or Asian (7.3%) (Table 1). Other factors associated with first-trimester bleeding are noted in Table 1.
First-trimester bleeding was associated significantly with the presence of bacterial vaginosis and other lower–reproductive tract infections (Table 2). Fully 60% of the women with first-trimester bleeding and 41% of women without first-trimester bleedings had findings of either bacterial vaginosis, T vaginalis, C trachomatis, group B streptococcal bacteriuria, or combinations of these infections at their first prenatal visit (P < .001). Figure 1 illustrates that first-trimester bleeding was most common among women with combined infections.
|
|
). Each of these infections except group B streptococcal bacteriuria was associated with first-trimester bleeding. To examine more carefully the temporal relationship between first-trimester bleeding and infection, we examined separately the 246 (24.8%) women enrolled in their first trimester. Among this subset of women, 12.2% reported first-trimester bleeding. Fully 53.3% of these women who reported first-trimester bleeding also had bacterial vaginosis. Women with bacterial vaginosis in the first trimester were twice as likely to have first-trimester bleeding (18.6%) compared with women without bacterial vaginosis (8.8%; RR 2.1, 95% CI 1.1, 4.2).
Second-trimester bleeding also was more common among black women (11.6%) compared with Hispanic (6.6%) or white or Asian women (4.6%; P = .01). Second-trimester bleeding also was related to the recovery of T vaginalis (RR 2.8, 95% CI 1.8, 4.5), C trachomatis (RR 2.1, 95% CI 1.1, 3.8), N gonorrhoeae (RR 5.8, 95% CI 1.1, 29.4), bacterial vaginosis (RR 1.3, 95% CI 1.0, 1.8), and M hominis (RR 1.8, 95% CI 1.3, 2.3) and presence of a urinary tract infection (RR 1.9, 95% CI 1.0, 3.8). Bleeding in the third trimester was increased among women with T vaginalis (RR 4.7, 95% CI 2.6, 8.7).
Low birth weight occurred more commonly among women with first-, second-, and third-trimester bleeding compared with those without gestational bleeding (Table 3). Birth weight less than 1 kg was associated strongly with second-trimester gestational bleeding (RR 15.4, 95% CI 3.1, 76.9). Clinically recognized chorioamnionitis also was increased among women with first-, second-, and third-trimester bleeding (Table 3). Among women with clinical chorioamnionitis at term, 23.5% (eight of 34) had reported first-trimester bleeding compared with only 9.2% (85 of 923) of women without clinical chorioamnionitis (RR 2.8, 95% CI 1.3, 6.1; P = .006).
|
). Preterm birth that followed spontaneous onset of labor occurred significantly more often among women with first-trimester bleeding but not second- or third-trimester bleeding (Table 3). On the other hand, women with second- and/or third-trimester bleeding were at significantly increased risk for delivering preterm after PROM (Table 3).
Figure 2 illustrates the relationships between first-trimester vaginal bleeding, bacterial vaginosis, T vaginalis, and risks for preterm birth among the 537 studied women enrolled in the control (observational) phase of the program. The risk of preterm birth associated with reported first-trimester bleeding (RR 1.8, 95% CI 0.6, 5.6) was increased further among women who also had findings at their first antenatal visit of bacterial vaginosis with and without T vaginalis (RR 4.6, 95% CI 0.9, 24.4; and RR 4.4, 95% CI 2.0, 9.5, respectively; Figure 2).
|
). The risk for preterm birth after labor was increased four-fold among women with both first-trimester bleeding and bacterial vaginosis. Similarly, the risk for preterm PROM was increased 5.4-fold for women with both first-trimester bleeding and bacterial vaginosis.
|
Oral antimicrobial treatment during the intervention phase of the study reduced rates of preterm birth among bacterial vaginosis-positive women with and without first-trimester bleeding (Figure 3). Preterm birth was reduced among women without bleeding who were treated programmatically for bacterial vaginosis and other prevalent infections (Figure 3).
|
|
Discussion |
|---|
|
Top
AbstractMaterial and MethodsResultsDiscussionReferences |
|---|
Interpretation of these observations must take into account several limitations of the study methods. Even though this was a relatively large clinical trial, stratification produced a number of small groups, and statistical power was limited. Although the magnitude of the point estimates suggests that the factors may be clinically important, the precision of the estimates is reduced. Misclassification of the studied infections was likely minimal, as sensitive and specific microbiologic methods were utilized to test for each condition at the onset of prenatal care. Possible treatment before joining the study and acquisition of infection after testing would lead to misclassification and would tend to underestimate associations between bleeding and infection.19 Similarly, gestational bleeding information and other risk factors (ie, smoking, alcohol, and drug use) were obtained from review of medical records and may be subject to errors in recording and recall.
Possible selection biases are important to consider. More than three times as many women lost to follow-up had reported early gestational bleeding as compared with women who were observed in the study; whether these women completed their care elsewhere or miscarried is not known. Of similar concern, the majority of women (77.6%) enrolled in this study presented for antenatal care after the first trimester. Thus, women who miscarried before starting prenatal care were not available for inclusion. Both selection bias, engendered with use of a convenience sample of clinic attendees, and bias from survival of healthy pregnancies would lead theoretically to an overestimation of the association between first-trimester bleeding and infection. If most early miscarriages were not related to infection and these women were included in the study, then the association between first-trimester bleeding and infections would be lower. Results in this study may reflect associations between infection and a subgroup of women with first-trimester bleeding whose pregnancies are more likely to survive into the second trimester. Conversely, if there is a relationship between early pregnancy loss and infection, then our estimate may represent an underestimation of the true association between infection and first-trimester bleeding. Further study is needed to address these study design issues and to examine the role of bacterial vaginosis and reproductive tract infections in early miscarriage, as well as in gestational bleeding. This study was conducted among urban women receiving subsidized medical care. It is possible that different groups of women may demonstrate different results. Finally, lack of random assignment to treatment or control group may have introduced unrecognized biases in study results of treatment effectiveness because unrecognized confounding factors may be maldistributed between treatment and control patients.
The present study confirms the findings of others, who examined separately gestational bleeding or reproductive tract infections as risk factors for preterm birth and PROM.1,2,5–12 Comparisons between this study and others examining the effects of gestational bleeding are hindered by use of a wide variety of definitions for gestational bleeding and by omission of information regarding genitourinary tract infections in earlier studies. In general, past studies examined bleeding over broader intervals during pregnancy (ie, before 20 or before 28 weeks’ gestation), whereas this study focused on first-trimester bleeding.1,3–6
Of past studies regarding bleeding and pregnancy outcomes, only that of Strobino and Pantel-Silverman3 evaluated possible infectious processes, ie, cervical inflammation, erosion, or infection as a risk factor for early gestational bleeding. However, these authors did not attempt to describe the specific microbial cause for the cervical condition. The present study expands past work by identifying significant associations between gestational bleeding and common prevalent infections including bacterial vaginosis, T vaginalis, C trachomatis, and N gonorrhoeae.
Data presented here confirm previous work demonstrating strong associations between common prevalent infections such as bacterial vaginosis and T vaginalis and preterm birth.7–9,11,12 To further understand these associations, we explored interrelationships between infection, bleeding, and adverse pregnancy outcomes. Specifically, we sought to describe the proportions of preterm births in which first-trimester bleeding might represent an intermediate step or disorder in the pathophysiologic pathway between genitourinary tract infection and preterm birth, and in which infections such as bacterial vaginosis and T vaginalis act independently to produce preterm birth, as well as the residual preterm births that are associated independently with first-trimester bleeding.
Intermediate associations, (ie, infection is present first; infection leads to bleeding and, subsequently, preterm birth) were suggested previously by Strobino and Pantel-Silverman,3 who noted increased vaginal bleeding among pregnant women who gave a history of cervical inflammation, infection, or erosion in the 6 months preceding conception (RR 2.1, 95% CI 1.2, 3.5). They also noted that rates of bleeding were lower (RR 1.6, 95% CI 1.0, 2.5) if cervical inflammatory changes had been noted more than 6 months before conception.
Associations between genital infections and bleeding are recognized increasingly among nonpregnant women. N gonorrhoeae and C trachomatis are well known to cause cervical erosion and increased friability. Furthermore, each is associated with infection of the endometrium.14,22,23 Abnormal reproductive tract bleeding occurs commonly among 40–60% of women with clinically recognized PID and is considered indicative of endometrial infection.14,22 Korn et al15 demonstrated endometritis among 45% of women with bacterial vaginosis in the absence of N gonorrhoeae or C trachomatis, and no evidence of upper genital tract tenderness. Recently, Hillier et al24 detailed the microbiology of endometritis among nonpregnant women with suspected PID. The type and frequency of the recovered microorganisms among these nonpregnant women were similar to isolates obtained from between fetal membranes in prior studies of women with preterm delivery by these authors.24 Taken together, these findings foster the possibility that in some instances, lower–reproductive tract bacteria are present within the uterus, before, during, or shortly after implantation, and that deciduitis/endometritis is responsible for clinically recognized gestational bleeding, just as it is related to metrorrhagia among nonpregnant women with PID.14
As suggested previously, bleeding from trophoblast and/or the decidua may enhance the ability of cervico-vaginal microbes to ascend into the lower reproductive tract and to begin de novo intrauterine infection during pregnancy. Uterine bleeding may physically displace maternal host defenses such as mucin or enhance microbial growth by increasing the availability of iron and other microbial nutrients. An alternative mechanism relating bleeding, infection, and preterm birth could be the acquisition of, or shift to, relatively more virulent microbes during pregnancy. Examples of such mechanisms include acquisition of a sexually transmitted infection and development/recurrence of bacterial vaginosis. Such microorganisms may defeat cervical host defenses and mediate directly necrosis and bleeding from within the uterus. We suggest that each of the above scenarios can play independent or shared roles in either increasing risks of intrauterine infection or representing a manifestation of a more severe infection, with inflammation and tissue destruction within the uterus (Figure 4).
|
This study sought to measure and inform possible links between early gestational bleeding, common genitourinary tract infections, and preterm birth. Separate prior associations between first-trimester bleeding and preterm birth and common genitourinary tract infections and preterm birth were confirmed. The potential links between infection and early gestational bleeding examined in this study give insight into the mechanisms by which gestational bleeding is related to adverse pregnancy outcomes through reproductive tract tissue infection and inflammation. Which pregnant women should be screened and treated for prevalent reproductive tract infections? This study shows that women with bacterial vaginosis without gestational bleeding received benefit in terms of reduced rates of preterm birth. This study also shows that pregnant women with bleeding and infection suffer relatively increased risks of prematurity. We suggest that all women with studied infections (with and without bleeding) should be identified promptly and treated systemically using standard Centers for Disease Control and Prevention protocols.
|
Footnotes |
|---|
Received June 29, 1998. Received in revised form October 29, 1998. Accepted November 5, 1998.
|
References |
|---|
|
Top
AbstractMaterial and MethodsResultsDiscussionReferences |
|---|
2. Berkowitz GS, Harlap S, Beck GJ, Freeman DH, Baras M. Early gestational bleeding and pregnancy outcome: A multivariable analysis. Int J Epidemiol 1983;12:165–73.
3. Strobino B, Pantel-Silverman J. Gestational vaginal bleeding and pregnancy outcome. Am J Epidemiol 1989;129:806–15.
4. Williams MA, Hickok DE, Zinghiem RW, Mittendorf R, Kimelman J, Mahony BS. Low birth weight and preterm delivery in relation to early-gestation vaginal bleeding and elevated maternal serum alpha-fetoprotein. Obstet Gynecol 1992;80:745–9.
5. Spinillo A, Nicola S, Piazzi G, Ghazal K, Colonna L, Baltaro F. Epidemiological correlates of preterm premature rupture of membranes. Int J Gynaecol Obstet 1994;47:7–15.[Medline]
6. Ananth CV, Savitz DA. Vaginal bleeding and adverse reproductive outcomes: A meta-analysis. Paediatr Perinat Epidemiol 1994; 8:62–78.[Medline]
7. McGregor JA, French JI, Richter R, Vuchetich M, Bachus V, Seo K, et al. Antenatal microbiologic and maternal risk factors associated with prematurity. Am J Obstet Gynecol 1990;163:1465–73.[Medline]
8. Hillier SL, Nugent RP, Eschenbach DA, Krohn MA, Gibbs RS, Martin DH, et al, for the Vaginal Infections and Prematurity Study Group. Association between bacterial vaginosis and preterm delivery of a low-birth-weight infant. N Engl J Med 1995;333:1737–42.
9. Minkoff H, Grunebaum AN, Scharz RH, Feldman J, Cummings M, Cromblehome W, 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;150:965–72.[Medline]
10. Møller M, Thomsen AC, Borch K, Dinesen K, Zdravkovic M. Rupture of fetal membranes and premature delivery associated with group B streptococci in urine of pregnant women. Lancet 1984;2:69–70.[Medline]
11. McDonald HM, O’Loughlin JA, Jolley P, Vigneswaran R, McDonald PJ. Prenatal microbiological risk factors associated with preterm birth. Br J Obstet Gynaecol 1992;99:190–6.[Medline]
12. Hay PE, Lamont RF, Taylor-Robinson D, Morgan DJ, Ison CA, Pearson J. Abnormal bacterial colonization of the genital tract and subsequent preterm delivery and late miscarriage. BMJ 1994;308: 295–8.
13. Witkin S, McGregor JA. Infection-induced activation of cell-mediated immunity: Possible mechanism for preterm birth. Clin Obstet Gynecol 1991;34:112–21.[Medline]
14. Wolner-Hanssen P, Mårdh P-A, Møller B, Weström L. Endometrial infection in women with chlamydial salpingitis. Sex Transm Dis 1982;9:84–8.[Medline]
15. Korn AP, Bolan G, Padian N, Ohm-Smith M, Schachter J, Landers DV. Plasma cell endometritis in women with symptomatic bacterial vaginosis. Obstet Gynecol 1995;85:387–90.[Abstract]
16. McGregor JA, French JI, Parker R, Draper D, Patterson E, Jones W, 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;173:157–67.[Medline]
17. Centers for Disease Control and Prevention. 1989 Sexually transmitted diseases treatment guidelines. MMWR Morb Mortal Wkly Rep 1989;38:1–102.
18. Amsel R, Totten PA, Cpeigel CA, Chen KCS, Eschenbach D, Holmes KK. Nonspecific vaginitis: Diagnostic criteria and microbial and epidemiologic associations. Am J Med 1983;74:14–22.[Medline]
19. Rothman KJ, Greenland S. Modern epidemiology. 2nd ed. Philadelphia: Lippincott-Raven, 1998.
20. Greenland S. Modeling and variable selection in epidemiologic analysis. Am J Public Health 1987;79:340–9.
21. Robins JM, Greenland S. Identifiability and exchangeability for direct and indirect effects. Epidemiology 1992;3:143–55.[Medline]
22. Mårdh P-A, Møller B, Ingerselv HJ, Nessler E, Weström L, Wolner-Hanssen P. Endometritis caused by Chlamydia trachomatis. Br J Vener Dis 1981;57:191–5.[Medline]
23. Hillier SL, Kiviat NB, Hawes SE, Hasselquist MB, Wolner-Hanssen P, Eschenbach DA, et al. Role of bacterial vaginosis-associated microorganisms in endometritis. Am J Obstet Gynecol 1996;175: 435–41.[Medline]
24. Hillier SL, Martius J, Krohn M, Kiviat N, Holmes KK, Eschenbach DA. A case-control study of chorioamnionic infection and histologic chorioamnionitis in prematurity. N Engl J Med 1988;319: 972–8.[Abstract]
This article has been cited by other articles:
|
|
 |
|
  N. Okun, K. A. Gronau, and M. E. Hannah Antibiotics for Bacterial Vaginosis or Trichomonas vaginalis in Pregnancy: A Systematic Review Obstet. Gynecol., April 1, 2005; 105(4): 857 - 868. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. Yang, K. E. Hartmann, D. A. Savitz, A. H. Herring, N. Dole, A. F. Olshan, and J. M. Thorp Jr. Vaginal Bleeding during Pregnancy and Preterm Birth Am. J. Epidemiol., July 15, 2004; 160(2): 118 - 125. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Stelmach, E. Kallas, H. Pisarev, and T. Talvik Antenatal Risk Factors Associated With Unfavorable Neurologic Status in Newborns and at 2 Years of Age J Child Neurol, February 1, 2004; 19(2): 116 - 122. [Abstract] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
