Original Contribution

The rate of sexually transmitted infections in ED patients with Vaginal bleeding

Nicole Berwald MD^a,?, Shahriar Zehtabchi MD^a, Susan Cheng MPH^b, Michael Augenbraun MD^c, Khaled Abu-Lawi PhD^d,

Tom McLaughlin PA^e, Michael Lucchesi MD^a

^aDepartment of Emergency Medicine, State University of New York, Downstate Medical Center/Kings county hospital,

Brooklyn, NY, USA

^bCollege of Medicine, State University of New York, Downstate Medical Center, Brooklyn, NY, USA

^cDepartment of Internal Medicine, Division of Infectious Disease, State University of New York,

Downstate Medical Center/Kings County Hospital, Brooklyn, NY, USA

^dDepartment of Microbiology, State University of New York, Downstate Medical Center/Kings County Hospital, Brooklyn, NY, USA ^eDepartment of Emergency Medicine, State University of New York, Downstate Medical Center, Brooklyn, NY, USA

Received 16 April 2008; revised 10 May 2008; accepted 13 May 2008

Abstract

Background: Early diagnosis of sexually transmitted infections (STI) such as Chlamydia trachomatis (CT) and Neisseria gonorrhoea (NG) is crucial in reducing complications. Vaginal bleeding (VB) has been suggested as a possible presentation of STI.

Objective: To identify the rate of STI in sexually active women presenting to the emergency department (ED) with VB.

Methods: Prospective observational study of females 18-55 years-old presenting to two affiliated urban EDs with VB (convenience sample). Patients with recent STI or antibiotic use (b 3 months) were excluded. To estimate the background rate of STI, we enrolled women with no genitourinary complaints as controls. Specimens for CT and NG were obtained during speculum exam from cases and by self administered vaginal swab in controls. All specimens were analyzed by polymerase chain reaction. Continuous data was presented as mean +- standard deviation and categorical data as percentages with 95% confidence intervals CI). Fisher’s exact test was used to compare the rate of STI between the groups.

Results: From 09.06 to 08.07 a total 273 subjects were enrolled (174 VB and 99 controls, mean age: 33 +- 10). Groups were similar with regards to baseline characteristics. The majority of STI cases were due to CT: 5.8% in VB group (95% CI, 3.2%-10.4%) vs. 7.1% in controls (95% CI, 3.5%-14.0%). The

STI rate was 6.3% (95% CI, 3.5%-11.1%) in VB patients and 8% (95% CI, 3.9%-15.3%) in controls.

Conclusions: Our ED patients with and without vaginal bleeding had comparable rate of STI.

(C) 2009

* Corresponding author. Department of Emergency Medicine, State University of New York, Downstate Medical Center, Box: 1228, Brooklyn, NY 11203, USA. Tel.: +1 718 245 4790; fax: +1 718 245 4799.

E-mail address: [email protected] (N. Berwald).

Introduction

Vaginal bleeding is a common chief complaint in the emergency department (ED) and may represent a variety of pathologies in pregnant and nonpregnant patients. Chlamydia

0735-6757/$ – see front matter (C) 2009 doi:10.1016/j.ajem.2008.05.012

trachomatis and Neisseria gonorrhoeae are 2 common patho- gens responsible for sexually transmitted infection (STI). These infections are often asymptomatic and are a known source of morbidity for women because of their association with Pelvic inflammatory disease (PID), infertility, Ectopic pregnancy, and chronic pain syndromes [1,2]. In addition, Vertical transmission of these infections can cause neonatal pneumonia, conjunctivitis, and infections of the oropharynx and urogenital tract [2].

Previously, it has been suggested that vaginal bleeding is associated with Genital tract infections [1,3-5]. If such association exists, screening for STI in this population could facilitate diagnosis and treatment, and prevent serious complications. Emergency department screening for STI is currently performed mainly on high-risk patients identified by their chief complaint (eg, vaginal discharge) or medical history (eg, multiple sexual partners). This study extends standards already in place by examining whether vaginal bleeding is an associated finding in STI. We sought to determine the prevalence of STI in inner-city ED patients with and without vaginal bleeding.

Methods

Study design

From July 2006 through July 2007, we conducted a prospective, observational cross-sectional study of ED patients with vaginal bleeding. The rate of STI in patients with vaginal bleeding was compared with that of patients with similar demographics who presented to the ED for non- genitourinary complaints (control group). The latter group was chosen to represent the background rate of STI in the community. A convenience sample of patients was enrolled in the study after obtaining informed consent. The joint institutional review boards approved the study.

Study population and setting

This study was conducted at 2 affiliated urban institutions: a tertiary care, university-associated hospital with 60 000 annual ED visits and a municipal academic hospital with approximately 150 000 ED annual visits. These 2 institutions primarily serve a low-income community composed of approximately 80% African and Caribbean Americans, 11% Hispanics, 5% whites, 1% Asians, and 3% other races. Patients in the vaginal bleeding group were selected if they met all of the following criteria: (1) sexually active women 18 to 55 years of age, (2) chief complaint of vaginal bleeding,

(3) no STI or antibiotic treatment within the last 3 months, and (4) consent to having a pelvic examination. Patients were included in the control group if they were sexually active, had no genitourinary symptoms, and did not report STI or any antibiotic treatment in the preceding 3 months. All

patients submitted a written consent to be enrolled in the study as required by the institutional review boards.

Study protocol

On assignED shifts, when academic associates were pre- sent, a consecutive sample of patients was enrolled in the study. Academic associates are medical students or under- graduate students who are trained to assist our department in data collection for specific projects.

All sexually active women aged 18 to 55 years presenting to the adult ED with the chief complaint of vaginal bleeding or spotting were asked to consent to the following: routine history (including sexual, gynecologic, and obstetric his- tory), physical examination (including a pelvic exam), urine pregnancy test, and cervical swab specimens for Chlamydia trachomatis and Neisseria gonorrhoeae. All participants completed a brief questionnaire focusing on pertinent information from their medical and sexual history including age, obstetric history, gynecologic history, sexual history, history of recent antibiotic use, current or recent treatment of STIs, and reason for ED visit. Physical examinations including pelvic examinations and cervical swab specimen collection were performed by an ED attending or resident physician involved in the patients’ care. To validate the reliability of self-administered vaginal swabs, we asked patients with vaginal bleeding to provide a self-administered vaginal swab as well.

In addition, sexually active women presenting to the ED without any genitourinary symptoms (eg, pelvic pain, vaginal discharge/bleeding, and urinary complaints) were offered screening by self-administered vaginal swabs for C trachomatis and N gonorrhoeae. These patients comprised the control group. The schematic study protocol is presented in Fig. 1.

Fig. 1 Schematic study protocol.

All ED staff involved in patient enrollment were provided with a detailed instruction package pertaining to pelvic examination and appropriate collection of cervical swabs. In addition, patients were instructed about proper techniques for obtaining self-administered vaginal swabs by the academic associates. A 1-page detailed instruction was also given to these patients before sample collection (Fig. 2).

Patients testing positive for C trachomatis and/or N gonorrhoeae were contacted and referred to the hospital STI or gynecology clinic for treatment and counseling. Patients with positive tests who were treated in the ED were likewise contacted for referral to the STI or gynecology clinics.

Measurements

Vaginal and cervical specimen swabs were collected and transported in 3-mL M4 culture transport media. After collection, specimens were stored at 2?C to 4?C until ready for the assay. Detection of C trachomatis and N gonorrhoeae

was performed using the COBAS AMPLICOR system (Roche Molecular Systems, Branchburg, NJ). Briefly, 2 mL of thoroughly vortexed specimen was mixed with lysis buffer for 10 minutes to extract DNA. The mix was then diluted and added to an amplification solution containing polymerase and complementary biotinylated primers. Finally, the sample was loaded into the COBAS AMPLICOR Analyzer for Polymerase chain reaction target amplification for both organisms simultaneously, followed by independent specific detection of each organism. Specific detection is accomplished using microscopic beads coated with probes followed by colorimetric detection using avidin/enzyme conjugate and its corresponding substrate. Positive and negative controls were included in each assay.

Data analysis

The primary outcome measure in this study was a positive test result for STI. We defined STI as a positive test for

Fig. 2 Instruction for self-collected specimen.

C trachomatis, N gonorrhoeae, or both. Predictor variables included vaginal bleeding, age, sexual history, number of sexual partners in the past year, gynecologic examination within the past year, chief complaint (including vaginal bleeding), urinary symptoms, physical findings (including vaginal discharge, vaginal bleeding, cervical motion tender- ness, and lower abdominal tenderness), and pregnancy status. Patients were stratified into 2 categories of study and control groups based on presence or absence of vaginal bleeding. Prevalence of these infections was compared in patients with vaginal bleeding with those with nongenitour-

inary symptoms.

Continuous data were reported as mean +- standard deviation. Categorical data were presented as percentages with 95% confidence intervals (CIs). Fisher exact test and Student t test were used to compare the categorical and continuous data, respectively. We used ? statistics to measure the agreement between the test results obtained from the self- administered vaginal swabs and the physician-assisted cervical swabs. All tests were 2-tailed (a = .05).

A sample size analysis revealed that we required 156 patients to reach a 95% CI with +-5% width of the calculated proportion. For this analysis, we used a rate of 11% (the lower limit of 95% CI previously reported for prevalence of STI in ED patients with vaginal bleeding) [7].

Results

A total of 508 patients were approached for enrollment in the study over a 1-year period (Fig. 1). Only 305 patients consented to participate. From this group, 17 patients were excluded because of violation of study protocol (genitour-

Table 1 Baseline characteristics of the study cohort

CT	n = 10	n = 7
	5.8% (3.2%-10.4%)	7.1% (3.5%-14.0%)
NG	n = 1	n = 2
	0.6% (0.1%-3.2%)	2% (0.6%-7.0%)
CT or NG	n = 11	n = 8
	6.3% (3.5%-11.1%)	8% (3.9%-15.3%)
CT and NG	n = 0	n = 1
	0%	1% (0.1%-6%)
a Positive test results by either cervical or vaginal swabs. b Specimen collected by vaginal swab.

inary complaints for control group). Fifteen patients (7 vaginal bleeding and 8 controls) had missing swab specimens or inconclusive results and were also excluded. Data were analyzed for the remaining 273 patients (174 vaginal bleeding and 99 controls). In the vaginal bleeding group, 48% (n = 83) tested positive for pregnancy. Baseline characteristics of study and control groups are compared in Table 1. Apart from the presence of genitourinary symptoms, the groups were similar.

Table 2 Prevalence of STI in vaginal bleeding group compared with the control group

Vaginal bleeding ^a Control ^b

(95% CI) (95% CI)

n = 174 n = 99

In the vaginal bleeding group, C trachomatis was observed in 5.8% of the patients (95% CI, 3.2%-10.4%) and N gonorrhoeae in 0.6% of the patients (95% CI, 0.1%- 3.2%). None of the patients tested positive for both organisms. The overall rate of STI (C trachomatis, N gonorrhoeae, or both) in the vaginal bleeding group was 6.3% (95% CI, 3.5%-11.1%) (Table 2).

Vaginal bleeding (n = 174)		Control (n = 99)	P
Age (mean +- SD, y)	31 +- 9.32	33 +- 10.34	.72
Ethnicity	83% (African American/Caribbean)	81% (African American/Caribbean)	.74
Private physician	60%	60%	1.00
Private gynecologist	51%	55%	.61
Pregnant	48% (n = 83)	–	–
History of STI	22%	13.4%	.8
History of CT	26%	16%	.09
History of GC	3.6%	6.1%	.36
History of PID	4.7%	3.1%	.75
Age at 1st intercourse (mean +- SD, y)	17 +- 3.5	17 +- 3.0	.50
Sex within the past year	94%	84%	.01
No. of sexual partners within the past 6 mo	1.06 +- 0.5	1.14 +- 0.8	.30
Dysuria	14%	8%	.18
Hematuria	28%	1%	.00
Urinary frequency	40%	25%	.01
Vaginal discharge	33%	21%	.04
Dyspareunia	22%	18%	.53
STI indicates sexually transmitted infections; CT, Chlamydia trachomatis; NG, Neisseria gonorrhoea; PID, pelvic inflammatory disease.

Discussion

Table 3 Rate of STI in different age groups (study vs control group)

Age group (y)	Vaginal bleeding	Control	P
18-29	13/87	7/42	.7
	14.9% (8.8%-24%)	16.7% (8%-31%)
30-44	1/64	1/36	1.0
	1.5% (0.01%-9%)	2.8% (0.01%-15%)
45-55	0/23 (0%)	0/21	–

Analysis of the vaginal swab specimens in the control group, which represents the background rate of STI in our community, revealed a rate of 7.1% (95% CI, 3.5%-14.0%) for positive C trachomatis and 2% (95% CI, 0.6%-7.0%) for N gonorrhoeae. One patient tested positive for both C trachomatis and N gonorrhoeae in the control group. The overall rate of STI (C trachomatis, N gonorrhoeae, or both) in the control group was 8% (95% CI, 3.9%-15.3%). The rate of STI in vaginal bleeding group was not significantly different from that of the control group (odds ratio, 0.77; 95% CI, 0.29-1.98). Although we met our required sample size for determining the rate of STI in patients with vaginal bleeding, our sample size might not have adequate power to identify any meaningful statistical significance between the vaginal bleeding and the control groups. Unfortunately, continuation of subject enrollment was not possible because one of the study locations changed the PCR assay used for STI testing.

On subgroup analysis, we compared the rates of STI in various age groups. The data revealed an age-related distribution for positive test results. The 18- to 29-year-old group had the highest rate of infection (14.9% in study group, 16.7% in control group), and the 45- to 55-year-old group had the lowest rates (0% in both control and study group). The rates of STI were not significantly different between the study group and control group for any age range (Table 3).

Analysis of STI-positive patients revealed no significant difference between the 2 groups of vaginal bleeding and controls when compared for ethnicity; pelvic examination in the year before the enrollment; number of sexual partners within the past year; new sexual partners; and history of C trachomatis, N gonorrhoeae, or PID.

We performed a subgroup analysis to compare the rate of STI in pregnant and nonpregnant patients with vaginal bleeding. The rate of STI in pregnant patient with vaginal bleeding (4.8%; 95% CI, 1.5%-12%) was lower than that of nonpregnant patients with vaginal bleeding (9.6%; 95% CI, 5%-17%). However, this difference was not statisti- cally significant.

Finally, the concordance rate between the vaginal swabs and cervical swabs was 98% for C trachomatis and 100% for N gonorrhoeae. The overall Interobserver agreement between self-administered vaginal swabs and physician- assisted cervical swabs in regard to positive and negative STI results was high (?, 0.7).

The purpose of this study was to examine the association between vaginal bleeding and STI in sexually active women 18 to 55 years of age with vaginal bleeding. Our study revealed that the rate of STI in patients presenting with vaginal bleeding was not higher than that of the general ED population.

The rate of STI in the United States is on the rise. The 2006 Sexually Transmitted Disease Surveillance Report, published by the Centers for Disease Control and Prevention, revealed a 5.6% increase in the rate of C trachomatis infections and a 5.5% increase in N gonorrhoeae infections from the previous year [6]. According to this report, the median incidence of C trachomatis and N gonorrhoeae in women 15 to 24 years of age screened at family planning clinics across the United States was 6.7% (range, 2.8%- 16.9%) and 1.1% (range, 0.0%-4.8%), respectively [6].

The prevalence of C trachomatis and N gonorrhoeae is highly variable among female ED patients [7-11]. Different studies have shown these rates to be in the range of 4.5% to 11% for N gonorrhoeae, 9.3% to 11% for C trachomatis, and 13.6% to 17% for either or both [7-11]. In our study, the prevalence of these infections was lower than other ED populations; however, our study included older patients than the aforementioned investigations. The rates of N gonor- rhoeae, C trachomatis, and STI in our study were 0.6%, 5.8%, and 6.3% for patients with vaginal bleeding and 2%, 7.1%, and 8% in the control group, respectively. These rates, however, are consistent with the reported prevalence rates of N gonorrhoeae (2.1%) and C trachomatis (4.7%) in the Northeast United States for young women [12]. Specifically looking at the STI rates in the younger age group (18- to 29-year-old women), our study demonstrates much higher STI rates (14.9% in study group and 16.7% in control group). Interestingly, the reported mean age for patients enrolled in previous studies falls into this age group [7,9-11]. That could explain the higher STI rates reported in some of ED studies. Identifying ED patients at high risk for STI may facilitate diagnosis and screening of such infections. Studies have found that younger age (b25 years), more that one sexual partner in the past year, and no or inconsistent use of contraception are associated with an increased risk of N gonorrhoeae and C trachomatis [5,10,13-15]. Postcoital bleeding and abnormal uterine bleeding have also been associated with increased rates of cervical infections [3,5]. The American College of Emergency Physicians clinical policy for patients presenting to the ED with vaginal bleeding suggests evaluation for cervical infections under most clinical circumstances. However, testing for STI is not a routine part of the workup for vaginal bleeding in most ED settings; and there has not been enough emphasis on the importance of such practice. This could be due to the scarcity of studies directly evaluating the rates of such infections in

female ED patients with vaginal bleeding.

In a small study conducted in 1975, Curran et al found positive N gonorrhoeae cultures in 43.8% of nonpregnant women with abnormal uterine bleeding in an urban ED [3]. Hoffman et al investigated the prevalence of STI in pregnant and nonpregnant ED patients with vaginal bleeding and found that both groups had similar rates of infection [7]. These investigators reported a prevalence of 11% for C trachomatis and 6% for N gonorrhoeae in pregnant ED patients with vaginal bleeding. The nonpregnant cohort had a prevalence of 8% for C trachomatis and 11% for N gonorrhoeae. Overall, Hoffman et al concluded that the 2 groups were at equal risk for STI and suggested that cervical swabs should be obtained for all women presenting to the ED with vaginal bleeding [7]. As opposed to our study, the study of Hoffman et al did not measure the baseline rate of STI in their ED population. Therefore, it is not clear whether vaginal bleeding was associated with higher risk of STI. In addition, the mean age for this study was 24 years for the pregnant patient population and 27 years for the nonpregnant patients. This may be a factor in the high rates of infection detected in this study, as younger age is a risk factor for infection. At the same time, the unusually high rate of N gonorrhoeae in the study of Hoffman et al could possibly be due to increased reported resistance of N gonorrhoeae to antibiotics in the geographic region where Hoffman et al conducted their study (Missouri) [16].

Similar to Hoffman et al [7], the rate of STI in pregnant patients with vaginal bleeding in our study was not significantly different than that in nonpregnant patients with vaginal bleeding. Although this difference was not statistically significant, the rate in nonpregnant women with vaginal bleeding was 2 times higher than that in the pregnant cohort. This lack of significance may be due to sample size of this subgroup analysis and should be evaluated in future studies.

Although we did not find a direct association between vaginal bleeding and STI, the high rate of such infections in ED population [6-11] may still warrant STI screening in these patients. It behooves us, from a public health perspective, to continue to better identify high-risk patients and support ED screening even in asymptomatic patients. Further investiga- tions in this area can potentially lead to clinical decision rules for testing and initiating Empiric therapy.

In addition, it has been suggested that ED patients are often lost to follow-up for a multitude of reasons including low level of education (high school or less), low socio- economic status, minority race, and insurance status [17,18]. This aspect of ED care is essential, especially when patients report that they are unlikely to seek outpatient care or obtain discharge medications. If we identify risk factors, we can decrease delay to therapy and therefore reduce the morbidity associated with untreated infections and disease transmission.

Although our sample size was adequate to achieve the primary goal of determining the rate of STI in the ED patients with vaginal bleeding, our study might not have enough

power to compare this STI rate with that of the general population. This was due to the decision to terminate the subject enrollment promptly because the main study site switched to a different PCR assay for STI testing.

Lastly, our study implemented vaginal swab samples for N gonorrhoeae and C trachomatis. Traditional cervical swab have several limitations for screening. These limita- tions stem from demand on resources and personnel, and patient participation. Previous investigations have looked into less invasive means of screening for STI. Using the Roche Amplicor system, a PCR assay, several studies have demonstrated that vaginal swab specimens (both physician- and self-administered specimens) are at least as sensitive and specific as the current standard testing for N gonorrhoeae and C trachomatis [19-22]. Our data are consistent with prior evidence demonstrating that vaginal swab samples perform as well as physician-assisted cervical samples using this PCR assay.

There are several limitations to our study. This was a convenience sample of patients, and the results may not represent the actual STI rate in our ED. Almost 40% of the identified subjects in our study refused to consent for enrollment. The refusals mostly were due to not feeling well or having been previously tested by their gynecologist. This factor could have skewed the results. In addition, the rates of infection in the control group may be falsely elevated. Patients with risk factors for STI may be more willing to agree to a screening test, whereas patients who are at lesser risk may not see a need to be tested. This factor might have inflated the STI rate in our study. In addition, 80% of our community is African/Caribbean American, which raises the concern that our findings may not be generalizable to some ED populations. This factor should be taken into consideration when applying our results to other communities.

Finally, there has been some concern about the effect of blood in the specimen on the PCR assay. However, this concern has been addressed by several studies excluding such effect [23,24]. In addition, the high concordance between the physician-assisted and self-administered swabs speaks against this. In our study, ED physicians were trained to debride the cervix of discharge, debris, and blood before obtaining the sample, lessening the potential for inhibitors.

Conclusion

According to our data, ED patients with vaginal bleeding are not at increased risk for STI. However, the higher rate of N gonorrhoeae and C trachomatis in ED patients (especially in the younger age group) may warrant STI screening not only in patients with vaginal bleeding but also in asympto- matic patients. This recommendation should be further studied from the standpoint of cost-effectiveness, feasibility, and acceptability.

References

1. American College of Emergency Physicians. Clinical policy for the initial approach to patients presenting with a chief complaint of vaginal bleeding. Ann Emerg Med 1997;29:435-58.
2. Centers for Disease Control and Prevention. Sexually transmitted disease treatment guidelines 2002. MMWR 2002;51(No RR-6):1-80.
3. Curran JW, Rendtorff RC, Chandler RW, et al. Female gonorrhea: its relation to abnormal uterine bleeding, urinary tract symptoms, and cervicitis. Obstet Gynecol 1975;45:195-8.
4. Krettek JE, Arkin SI, Chaisilwattana P, et al. Chlamydia trachomatis in patients who used oral contraceptives and had intermenstrual spotting. Obstet Gynecol 1993;81:728-31.
5. Verhoeven V, Avonts D, Meheus A, et al. Chlamydial infection: an accurate model of opportunistic screening in General practice. Sex Transm Infect 2003;79:313-7.
6. Centers for Disease Control and Prevention. Trends in reportable Sexually transmitted diseases in the United States, 2006 national surveillance data for chlamydia, gonorrhea, and syphilis. Avail- able at: http://www.cdc.gov/std/stats/trends2006.htm (last accessed January 10, 2008) 2007.
7. Hoffman L, Ma OJ, Gaddis G, et al. Cervical infections in emergency department patients with vaginal bleeding. Acad Emerg Med 2002;9 (8):781-5.
8. Doezema D, Hepworth E, Young SA, et al. Low prevalence of Chla- mydia trachomatis by urinary ligase chain reaction in woman patients in the emergency department. Acad Emerg Med 2002;9:646-9.
9. Mehta SD, Rothman RE, Kelen GD, et al. Unsuspected gonorrhea and chlamydia in patients of an urban adult emergency department. Sex Transm Dis 2001;28(1):33-9.
10. Bachmann LH, Pigott D, Desmond R, et al. Prevalence and factors associated with gonorrhea and chlamydial infection in at-risk females presenting to an urban emergency department. Sex Transm Dis 2003;30:335-9.
11. Silva A, Glick NR, Lyss SB, et al. Implementing an HIV and sexually transmitted disease screening program in an emergency department. Ann Emerg Med 2006;49:564-72.
12. Miller WC, Ford C, Morris M, et al. Prevalence of chlamydial and gonococcal infections among young adults in the United States. JAMA 2004;291:2229-36.
13. Burstein G, Zenilman JM, Gaydos CA, et al. Predictors of repeat Chlamydia trachomatis infections diagnoses by DNA amplifica- tion testing among Inner city female. Sex Transm Inf 2001;77: 26-32.
14. Marrazzo J, Handsfield HH, Whittington WL. Predicting chlamydial and gonococcal cervical infection: implications for management of cervicitis. Obstet Gynecol 2002;100:579-84.
15. United States Preventative services Task Force. Screening for gonorrhea: recommendation statement. Ann Fam Med 2005;3: 263-7.
16. Centers for Disease Control and Prevention. Fluoroquinolone- resistance in Neisseria gonorrhoeae, Hawaii, 1999, and decreased susceptibility to azithromycin in N. gonorrhoeae, Missouri. MMWR 2000;49:833-7.
17. Sun BC, Burstin HR, Brennan TA. Predictors and outcomes of frequent emergency department users. Acad Emerg Med 2003;10:320-8.
18. Handel DA, McConnell JK, Allen H, et al. Outpatient follow-up in today’s health care environment. Ann Emerg Med 2007;49: 288-92.
19. Knox J, Tabrizi S, Miller P, et al. Evaluation of self-collected samples in contrast to practitioner-collected samples for detection of Chlamy- dia trachomatis, Neisseria gonorrhoeae, and Trichomonas vaginalis by polymerase chain reaction among women living in remote areas. Sex Transm Dis 2002;29:647-54.
20. Schachter J, McCormack WM, Chernesky MA, et al. Vaginal swabs are appropriate specimens for diagnosis of genital tract infection with

C. trachomatis. J Clin Microbiol 2003;41:3784-9.

1. Gaydos CA, Crotchfelt KA, Shah N, Tennant, et al. Evaluation of dry and wet transported intravaginal swabs in detection of Chlamydia trachomatis and Neisseria gonorrhoeae infections in female soldiers by PCR. J Clin Microbiol 2002;40:758-61.
2. Wiesenfeld HC, Heine RP, Rideout A, et al. The vaginal introitus: a novel site for Chlamydia trachomatis testing in women. Am Journ Obstet Gynecol 1996;174:1542-6.
3. Marrazzo JM, Johnson RE, Green TA, et al. Impact of patient characteristics on performance of nucleic acid amplification tests and DNA probe for detection of Chlamydia trachomatis in women with genital infections. J Clin Microbiol 2005;43:577-84.
4. Verkooyen RP, Luijendijk A, Huisman WM, et al. Detection of PCR inhibitors in cervical specimens by using the AMPLICOR Chlamydia trachomatis assay. J Clin Microbiol 1996;34:3072-4.