HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Steinhandler, L. Articles by Cruickshank, C. Search for Related Content PubMed PubMed Citation Articles by Steinhandler, L. Articles by Cruickshank, C.

Combination of Bacterial Vaginosis and Leukorrhea as a Predictor of Cervical Chlamydial or Gonococcal Infection

From the Department of Obstetrics and Gynecology and the George Anderson Outcomes Measurement Unit, Women and Infants Hospital, Providence, Rhode Island.

Address reprint requests to: Jeffrey F. Peipert, MD, MPH, Women and Infants Hospital, Department of Obstetrics and Gynecology, 101 Dudley Street, Providence, RI 02905; E-mail: jpeipert{at}wihri.org.

	ABSTRACT

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
OBJECTIVE: To evaluate whether the combination of bacterial vaginosis and leukorrhea on microscopic evaluation of a saline wet preparation is associated with cervical infection with Chlamydia trachomatis (C. trachomatis) or Neisseria gonorrhea (N. gonorrhea).

METHODS: A cross-sectional study of 598 patients was performed. Nonpregnant patients undergoing a saline wet preparation and microbiologic testing for C. trachomatis and N. gonorrhea were eligible. Providers prospectively collected data from saline microscopic analysis. Bacterial vaginosis was documented based on Amsel’s criteria, and ligase chain reaction testing of the endocervix was performed for C. trachomatis and N. gonorrhea.

RESULTS: On univariate analysis, young age (less than 25 years), unmarried marital status, black race, and the presence of leukorrhea were all associated with increased rates of testing positive for C. trachomatis and N. gonorrhea (P < .05). We stratified patients into three groups: 1) no evidence of bacterial vaginosis or leukorrhea; 2) evidence of either bacterial vaginosis or leukorrhea, but not both; and 3) evidence of both bacterial vaginosis and leukorrhea. Using logistic regression analysis to control for age, marital status, and race, we found an odds ratio of 3.8 (95% confidence interval 1.3, 11.6) for a positive microbiologic test for either C. trachomatis or N. gonorrhea in women with the combination of bacterial vaginosis and leukorrhea.

CONCLUSION: In this group of high-risk women seen in an urgent care facility, the presence of both bacterial vaginosis and leukorrhea was associated with an increased risk of cervical infection. Future prospective studies should evaluate whether empiric treatment of high-risk women with these findings is justified and cost-effective.

Infection with Chlamydia trachomatis (C. trachomatis) accounts for more than 3 million new infections each year in the United States, and is the most common sexually transmitted pathogen.¹ Neisseria gonorrhea (N. gonorrhea) infection is second only to chlamydial infections in the number of cases reported to the Centers for Disease Control and Prevention.² Infections with either C. trachomatis or N. gonorrhea can result in urethritis, cervicitis, pelvic inflammatory disease, and sequelae such as chronic pelvic pain, tubal factor infertility, and ectopic pregnancy.^1,3 Therefore, the medical and economic implications of untreated C. trachomatis or N. gonorrhea infection are significant. Early identification is an important step in reducing the above sequelae.

Clinical recognition of infection with C. trachomatis or N. gonorrhea depends on a high index of suspicion and a careful clinical examination. Therefore, attempts have been made to identify women with certain demographic risk factors, as well as signs and symptoms which suggest infection, and the need to screen and/or treat.^3,4

There is evidence in the literature that lower genital tract inflammation and bacterial vaginosis are associated with an increased risk of cervical and upper genital tract infection.^5,6 It was our clinical experience that the combination of both excess white blood cells (WBCs) on saline microscopic examination and bacterial vaginosis was a marker for a patient at high risk for C. trachomatis or N. gonorrhea infection. Based on these previous studies, we developed the hypothesis that the combination of both leukorrhea (more WBCs than epithelial cells on saline microscopic examination of vaginal fluid) and bacterial vaginosis is associated with an increased risk for infection with C. trachomatis or N. gonorrhea even after controlling for other demographic characteristics.

	MATERIALS AND METHODS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
After approval from our Institutional Review Board was received, we conducted a cross-sectional study of women presenting to the Women and Infants Hospital Urgent Care/Triage Unit. All nonpregnant patients seen during a 12-month interval who were evaluated with a saline wet preparation and microbiologic testing for C. trachomatis and N. gonorrhea were eligible. The urgent care unit triages approximately 25,000 obstetric and gynecologic visits per year. The prevalence of microbiologically proven cervicitis in this population is approximately 5–7%. Gynecologic patients with complaints of abnormal discharge, pain, dysuria, or concerns about a partner with a sexually transmitted disease are tested with a vaginal saline microscopic examination and microbiogic testing for C. trachomatis and N. gonorrhea.

Obstetrics and gynecology residents who are usually in their second or third postgraduate year typically evaluate patients seen in the urgent care/triage unit. A relatively small number of patients were evaluated solely by attending physicians. To collect the clinical data from over 20 different providers, each clinician completed a specially designed data extraction form at the time of the microscopic examination of the saline wet preparation. Data on the form included: vaginal pH, presence and color of the vaginal discharge, amine odor test, percent of epithelial cells studded with anaerobic bacteria ("clue" cells), quantification of vaginal WBCs (number of WBCs per high-power field), presence of leukorrhea (defined below), cervical friability, presence of mucopus, bimanual examination findings, and the presumed diagnosis. The demographic and reproductive data extracted from the medical record included age, race, marital status, insurance status, gravidity, parity, chief complaint, secondary complaints, and duration of symptoms.

The primary exposures of interest included leukorrhea and bacterial vaginosis. The presence of leukorrhea was defined as a vaginal saline preparation with more WBCs than epithelial cells on high-power microscopic inspection of a saline wet preparation. We used Amsel’s criteria for the definition of bacterial vaginosis.⁷ These criteria include three or four of the following clinical findings: 1) presence of a homogeneous vaginal discharge; 2) vaginal pH greater than 4.5; 3) positive amine odor test with the addition of potassium hydroxide (KOH); and 4) the presence of greater than 20% clue cells. Our primary outcome of interest was testing positive for either C. trachomatis or N. gonorrhea or both. Testing for C. trachomatis and N. gonorrhea was performed using ligase chain reaction as this was the standard of care at our institution during the period of study.

Statistical analysis was performed using Statistical Analysis Software (SAS Institute, Cary, NC) and EpiInfo 6.0 (USD, Inc., Stone Mountain, GA). Descriptive statistics were performed on the entire population. Cross-tabulations were performed on the entire cohort to assess the univariate association of predictors variables and a positive test for C. trachomatis or N. gonorrhea. We then created three groups based on the presence of leukorrhea and bacterial vaginosis. Because many patients were missing at least one of the four clinical criteria for our definition of bacterial vaginosis or information regarding the presence or absence of leukorrhea, we analyzed the subgroup of patients with complete information (n = 380). Group 1 consisted of patients who were negative for bacterial vaginosis and leukorrhea. Patients positive for bacterial vaginosis or leukorrhea, but not both were placed in group 2, and group 3 consisted of patients who were positive for both leukorrhea and bacterial vaginosis. After univariate analysis was performed, we calculated ² for linear trend,⁸ odds ratios (OR), and 95% confidence intervals (CI), and performed logistic regression to assess the association of wet preparation findings with test results while controlling for age, race, and marital status in the model. We also performed our analysis on the entire cohort of women, but used a modification of the clinical criteria for bacterial vaginosis. In this analysis, we used greater than 20% clue cells as the definition of bacterial vaginosis as almost all patients in the dataset had complete information on the percentage of clue cells. Because our results did not differ appreciably from the analysis using Amsel’s criteria, we only report the results of the patients in the database with complete data.

	RESULTS

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
The demographic and reproductive characteristics of the study population were as follows: 47% (281 of 598) white, 27% (161 of 598) black, 22% (134 of 598) Hispanic, 49% (291 of 598) less than 25 years old, 78% (469 of 598) unmarried, 61% (362 of 598) private or health maintenance organization insurance, and 27% (161 of 598) nulliparous. The primary presenting symptom in 54% (322 of 598) of patients was abdominal pain, and 32% (190 of 598) presented with complaints of abnormal vaginal discharge. Of the 598 patients evaluated as part of this study, 75 patients (12.5%) tested positive for C. trachomatis or N. gonorrhea (9.4% positive for chlamydia and 4.3% positive for gonorrhea).

We tested the association of demographic and clinical characteristics and testing positive for C. trachomatis or N. gonorrhea in the entire cohort. Age less than 25 (OR 5.0, 95% CI 2.7, 9.5) and black race (OR 3.1, 95% CI 1.8, 5.2) were associated with testing positive, whereas being married was protective (OR 0.2, 95% CI 0.1, 0.6). The presence of leukorrhea was also associated with testing positive for C. trachomatis or N. gonorrhea (OR 3.2, 95% CI 1.8, 5.5). Among the 386 patients with complete clinical criteria for bacterial vaginosis, 30.3% (117 of 386) were found to have three or four Amsel’s criteria. The association of bacterial vaginosis (in the subgroup with complete clinical criteria) and positive microbiologic test had an OR of 1.8, but this did not reach statistical significance (95% CI 0.9, 3.9).

The characteristics of patients stratified by group are shown in Table 1. Patients in group 3 tended to be younger and unmarried (P < .001). Groups 2 and 3 had a higher percentage of blacks compared with group 1 (41% and 31%, respectively, compared with 21%, P < .001). Because previous studies have shown that age, race, and marital status are associated with testing positive for C. trachomatis or N. gonorrhea, we controlled for these potential confounding variables in the multivariate analysis.

We then entered race (black/nonblack), young age (less than 25 years), marital status, and positive wet preparation for both leukorrhea and bacterial vaginosis into a logistic regression model to assess the independent contribution of each variable. Results of the logistic regression are shown in Table 2. The combination of bacterial vaginosis and leukorrhea was significantly associated with a positive test (OR 3.8, 95% CI 1.3, 11.6) even after adjustment for age, race, and marital status. Black race and young age were also statistically significant in the logistic model, with effect sizes of 5.2 and 3.5, respectively.

	DISCUSSION

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

In the urgent care setting and in symptomatic women, it would be helpful to have a method of determining women at high risk for infection. This is especially true where patient follow-up is not assured, as well as in developing countries where microbiologic testing is not possible because of resource limitations. Numerous demographic and clinical risk factors have been proposed to identify patients in need of screening or who are at high risk for infection. Commonly cited risk factors include young age (ie, less than 25 years), nonwhite race, unmarried marital status, new or multiple sexual partners, lack of condom use, history of other sexually transmitted diseases, douching, mucopurulent cervical discharge, purulent vaginal discharge, cervical edema, ectopy, and friability.^3,4

Leukorrhea seen on vaginal wet preparation has been shown to be a positive predictor for infectious cervicitis and for upper genital tract infection.⁵ In addition, bacterial vaginosis has been associated with coinfection with sexually transmitted diseases and may be an independent risk factor for pelvic inflammatory disease.⁶ Bacterial vaginosis, by itself, is not typically associated with increased vaginal inflammatory cells,¹⁰ and vaginal fluid from patients with bacterial vaginosis has been shown to be toxic to polymorphonuclear leukocytes.¹¹

The data in this study support our hypothesis that the combination of bacterial vaginosis and leukorrhea noted on microscopic examination of vaginal discharge is associated with an increased risk of C. trachomatis or N. gonorrhea infection. Sellors et al¹² tested 596 women in family planning clinics for C. trachomatis and N. gonorrhea. He found that the combination of mucopurulent cervicitis, greater than ten polymorphonuclear leukocytes per high power field, and bacterial vaginosis was 5.4 (P = .04). The likelihood ratio of infection with mucopus and leukorrhea, but no bacterial vaginosis, was 4.3.¹² Similarly, our data support the hypothesis that the combination of bacterial vaginosis and leukorrhea is associated with a higher risk of infection than either criterion alone. Thus, this combination of findings should raise the clinician’s concern that a sexually transmitted infection may be present.

As mentioned earlier, one of the most common clinical findings in women with C. trachomatis or N. gonorrhea cervical infection is mucopurulent cervical discharge.¹³ The number of polymorphonuclear leukocytes in the cervical mucus is correlated with chlamydial cervicitis.¹⁴ However, attention must be paid to careful collection of the specimen and to selective counting of polymorphonuclear leukocytes in the cervical mucus, rather than in areas containing vaginal epithelial cells.³ Given the variability of the providers in our clinical setting and the inability of our laboratory to perform and read Gram stains of cervical mucus on an emergency basis, we did not collect endocervical Gram stains. Thus, we were unable to evaluate polymorphonuclear leukocytes in cervical mucus as a predictor variable or to compare this finding with vaginal leukorrhea or the combination of bacterial vaginosis and leukorrhea. Our primary goal was to assess the combination of two readily available clinical criteria as a potential marker of patients at high risk for infection. It is interesting that our adjusted OR of 3.8 is very similar to the OR provided by Ryan et al for the association of cervical mucopus and N. gonorrhea (OR 3.7, 95% CI 2.4, 5.8) or C. trachomatis (OR 5.6, 95% CI 3.6, 8.7) infection.¹⁵

One of the major limitations of this study is the lack of data concerning patients’ sexual history. Sexual history taking by resident practitioners in the urgent care setting and the documentation of the history is inconsistent. Thus, these data were not included in our analysis. In addition, we were missing data in approximately one-third of patients and could only apply Amsel’s four clinical criteria for diagnosis of bacterial vaginosis in a subset of our 598 patients. However, when an alternative definition of bacterial vaginosis of greater than 20% clue cells was applied to include a larger percentage of the original cohort (n = 560), our results were unchanged. A final limitation of this study concerns external validity. We believe our findings may apply to a population of women at high risk for cervical infection with C. trachomatis and N. gonorrhea. The population studied had a relatively high prevalence of both bacterial vaginosis and infection with C. trachomatis or N. gonorrhea. The findings may not be valid in a private practice setting with women at lower risk for infection. Additional studies are necessary to validate our findings in other high-risk and intermediate or low-risk populations.

In many clinical sites providing care to high-risk women, there is considerable difficulty in notifying patients of positive microbiologic cultures because of change of address, phone number, or simply lack of contact information. In our urgent care setting where follow-up is often a challenge, our practice is to consider empiric therapy in high-risk women with the combination of bacterial vaginosis and leukorrhea. Empiric treatment of a high-risk group based on a saline wet preparation and presenting symptoms may help decrease transmission and sequelae of the infections including pelvic inflammatory disease, infertility, ectopic pregnancy, and chronic pelvic pain. We accept that we may be overtreating a fairly large number of women (75% of women with both criteria). However, we believe this is justified to avoid the individual and public health consequences of untreated C. trachomatis or N. gonorrhea infection in the community and the potential consequences of delay of antibiotic therapy. Future prospective studies should evaluate whether empiric treatment of high-risk women with these findings is justified and cost-effective. Even with screening for leukorrhea or bacterial vaginosis on the vaginal wet preparation, a significant percentage of patients with positive testing would be missed by this approach. In addition, most women with the combination of leukorrhea and bacterial vaginosis will not test positive for C. trachomatis and N. gonorrhea. Liberal use of highly sensitive and specific testing for chlamydia and gonorrhea is essential for the control of these potentially serious sexually transmitted infections.

	Footnotes

 
Presented in part at the ACOG District I Meeting and the ACOG Annual Clinical Meeting. Dr. Steinhandler was the recipient of the 1999–2000 Donald F. Richardson Memorial Prize Paper Award.

Supported in part by NIH Grant K24 HD01298-02, Midcareer Investigator Award in Women’s Health Research, from the National Institute of Child Health and Human Development.

PII S0029-7844(01)01788-4

Received August 14, 2001. Received in revised form November 26, 2001. Accepted December 11, 2001.

	REFERENCES

TOP ABSTRACT MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

 
1. US Preventive Services Task Force. Screening for chlamydial infection: Recommendations and rationale. Am J Prev Med 2001;20:90–4.[Medline]

2. Centers for Disease Control and Prevention, Division of AIDS, STD, and TB Laboratory Research. Neisseria gonorrhoeae. Available at http://www.cdc.gov/ncidod/dastlr/gcdir/gono.html. Accessed 2001 Aug 6.

3. Stamm WE. Chlamydia trachomatis infections of the adult. In: Holmes KK, ed. Sexually transmitted diseases. 3rd ed. New York: McGraw-Hill; 1999:407–22.

4. Hook EW, Handsfield HH. Gonococcal infections in the adult. In: Holmes KK, ed. Sexually transmitted diseases. 3rd ed. New York: McGraw-Hill; 1999:457.

5. Peipert JF, Ness RB, Soper D, Bass D. Association of vaginal white blood cells with objective evidence of pelvic inflammatory disease. Infect Dis Obstet Gynecol 2000;55: 51–7.

6. Peipert JF, Montagno A, Cooper AS, Sung J. Bacterial vaginosis as a risk factor for upper genital tract infection. Am J Obstet Gynecol 1997;177:1184–7.[Medline]

7. Amsel R, Totten PA, Spiegel CA, Chen KC, Eschenbach D, Holmes KK. Nonspecific vaginitis: Diagnostic criteria and microbial and epidemiologic associations. Am J Med 1983;74:14–22.[Medline]

8. Fisher LD, van Belle G. Biostatistics: A methodology for the health sciences. New York: John Wiley & Sons; 1993: 253–6.

9. Scholes D, Stergachis A, Heidrich FE, Andrilla H, Holmes KK, Stamm WE. Prevention of pelvic inflammatory disease by screening for cervical chlamydial infection. N Engl J Med 1996;334:1362–6.[Abstract/Free Full Text]

10. Eschenbach DA, Hillier S, Critchlow C, Stevens C, De-Rouen T, Holmes KK. Diagnosis and clinical manifestations of bacterial vaginosis. Am J Obstet Gynecol 1988; 158:819–28.[Medline]

11. Rein MF, Shih LM, Miller JR, Guerrant RL. Use of a lactoferrin assay in the differential diagnosis of female genital tract infections and implications for the pathophysiology of bacterial vaginosis. Sex Trans Dis 1996;23: 517–21.[Medline]

12. Sellors J, Howard M, Pickard L, Jang D, Mahony J, Chernesky M. Chlamydial cervicitis: Testing the practice guidelines for presumptive diagnosis. CMAJ 1998;158: 41–6.[Abstract]

13. Paavonen J, Stevens CE, Wolner-Hanssen P, Critchlow CW, DeRouen T, Kiviat N, et al. Colposcopic manifestations of cervical and vaginal infections. Obstet Gynecol Surv 1988;43:373–81.[Medline]

14. Brunham RC, Paavonen J, Stevens CE, Kiviat N, Kuo CC, Critchlow CW, et al. Mucopurulent cervicitis — the ignored counterpart in women of urethritis in men. N Engl J Med 1984;311:1–6.[Abstract]

15. Ryan CA, Courtois BN, Hawes SE, Stevens CE, Eschenbach DA, Holmes KK. Risk assessment, symptoms, and signs as predictors of vulvovaginal and cervical infections in an urban U.S. STD clinic: Implications for use of STD algorithms. Sex Trans Infect 1998;74(Suppl. 1):S59–S76.

This article has been cited by other articles:

				W M Geisler, S Yu, M Venglarik, and J R Schwebke Vaginal leucocyte counts in women with bacterial vaginosis: relation to vaginal and cervical infections Sex. Transm. Inf., October 1, 2004; 80(5): 401 - 405. [Abstract] [Full Text] [PDF]

				K Manavi, R Conlan, and G Barrie The performance of microscopic cervicitis for the detection of chlamydial infection Sex. Transm. Inf., October 1, 2004; 80(5): 415 - 415. [Full Text] [PDF]

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Steinhandler, L. Articles by Cruickshank, C. Search for Related Content PubMed PubMed Citation Articles by Steinhandler, L. Articles by Cruickshank, C.