Background: Chlamydia trachomatis is a common, often asymptomatic sexually transmitted infection.
Goal: The goal was to estimate the prevalence and predictors of C. trachomatis among young women using self-collected vaginal swabs, and the preferences of women and physicians for self-testing.
Study Design: A total of 514 attendees of university/college health clinics, adolescent birth control clinics, centers providing health services to homeless youth and adults (street health centers), a sexually transmitted diseases clinic, and family practices were tested by ligase chain reaction. Preference for self- versus provider-testing was examined.
Results: Prevalence was 6.0% and was highest (18.2%) in the street health centers. In multivariate analysis, only recent contact with someone with C. trachomatis infection was significantly associated with infection (odds ratio, 7.1; 95% confidence interval, 2.5–20.0). Most women (54.2%; 256 of 472) preferred self-sampling compared with physician sampling (15.9%; 75 of 472). The majority of physicians (75.0%; 9 of 12) reported at the start and end of the study that they would use vaginal swab self-sampling if available.
Conclusions: Prevalence of infection in young women attending homeless youth organizations was high. Self-sampling was acceptable and could facilitate screening in high-risk women who do not regularly access health services.
Prevalence of Chlamydia trachomatis (6.0%) in self-collected vaginal swabs ranged from 2.3% in student health centers to 18.2% in centers servicing homeless young women. Self-sampling was preferred over physician sampling.
*Public Health and Community Services Department, City of Hamilton, the Departments of †Family Medicine, ‡Clinical Epidemiology and Biostatistics, and §Pathology and Molecular Medicine, McMaster University, Hamilton, and ∥Father Sean O'Sullivan Research Center, St. Joseph's Healthcare, Hamilton, Ontario, Canada
This research was supported by grants provided by the Public Health Research, Education & Development Program (PHRED) and Pfizer Canada.
Dr. Sellors is now affiliated with the Program for Appropriate Technology in Health, Seattle, Washington.
Correspondence: Elizabeth Richardson, MD, MHSc, FRCP, City of Hamilton, Public Health & Community Services Department, 1 Hughson St. N., 4th Floor, Hamilton, Ontario, Canada L8R 3L5. E-mail: email@example.com
Received for publication March 4, 2003,
revised July 10, 2003, and accepted July 16, 2003.
CHLAMYDIA TRACHOMATIS IS THE most common bacterial sexually transmitted disease in North America. 1 In Canada, approximately 75% of all reported cases occur in women, and the highest rates of infection occur among young women 15 to 24 years of age. 2C. trachomatis prevalence ranges from 2.5% to 20% in different settings using different testing methods. 3–9 Rates in the city of Hamilton, Canada, have increased by 38% from 1999 to 2000, possibly in part as the result of ongoing research projects and enhanced contact screening. 10
Chlamydial infections among women are often asymptomatic. If untreated, lower genital tract infections can ascend and result in pelvic inflammatory disease, endometritis, salpingitis, ectopic pregnancy, or tubal factor infertility. 11 Neonatal pneumonia and conjunctivitis can be complications in infants born to infected mothers. 12 The Canadian task force on the Periodic Health Examination revised their recommendations for chlamydia screening in 1996. Recommendations included screening women with symptoms of C. trachomatis infection (easily induced mucosal bleeding, intermenstrual bleeding, or mucopurulent cervical discharge), pregnant women in their first trimester, sexually active women under 25 years of age, men or women who have had new or multiple sexual partners in the preceding year, and women using nonbarrier contraceptive methods. 1 Similar criteria for selective screening of young women were derived and validated locally in Hamilton, Canada, in 1992, and included easily induced mucosal bleeding, suspicious discharge, increased urinary frequency, intermenstrual bleeding, or a new sexual partner in the past year. 13
It is difficult to estimate the true prevalence and incidence of C. trachomatis infection among young women, especially among those who are transient and do not access health care. Among those young women who access health care, some might not perceive themselves to be at risk or could be reluctant to undergo pelvic examination. The availability of less-invasive sampling methods could increase the acceptability of C. trachomatis testing for both patients and providers, and expand testing to populations who are difficult to reach. Nucleic acid amplification (NAA) techniques such as ligase chain reaction (LCR) provide increased sensitivity and specificity for detecting C. trachomatis. 6,7,14 In traditional swab specimens and first void urine (FVU) from women, LCR testing for chlamydia with self-obtained vaginal specimens was as sensitive as cervical LCR and more sensitive than cell culture. 15 LCR testing will no longer be available; however, the use of other amplified tests (polymerase chain reaction, strand displacement amplification, and transcription mediated amplification) are available and have been used with vaginal swab specimens.
Recent increases in C. trachomatis prevalence among youth in Hamilton have demonstrated the need to carefully monitor and control the infection. To determine the prevalence and risk factors for C. trachomatis infection in young women in various settings, and to examine the acceptability of self-sampling, we conducted a cross-sectional survey using self-obtained vaginal specimens in Hamilton, Ontario. We also determined the preferences of physicians for vaginal self-sampling.
From May 2000 to February 2001, women 16 to 30 years of age who presented at the study sites in Hamilton, Ontario, were invited to participate in the study. Hamilton has 1 college and 1 university, both with student primary healthcare services. There is also a confidential sexually transmitted disease (STD) walk-in clinic run by the city's Public Health and Community Services Department near the downtown core; several community services available to youth who are transient or homeless, as well as drop-in centers providing meals, shelter, and/or health care in the downtown area; and several planned parenthood or birth control clinics in various areas of the city. Women could have been attending 1 of the sites for any health reason, including birth control, counseling, or STD testing. Sites included the Hamilton STD clinic, 4 sexual health centers servicing youth, 2 community-based organizations providing services to homeless youth, a community-based Street Health Center providing services to homeless adults, the McMaster University and the Mohawk College Student Health Services Clinics, and 12 family physicians’ offices in the community. In all settings except the street centers, recruiting was done in private. In the street centers, display posters invited attendees to approach clinic staff individually if interested in participating in the study. Participants who had taken antibiotics in the preceding 14 days were excluded. After self-collection of the vaginal swab, the women completed a self-administered questionnaire on demographics, sexual health, behavior, and preference for testing (the self-collected vaginal swab versus a hypothetical physician-collected cervical swab during a pelvic examination).
Physicians were asked to complete a questionnaire both at the beginning and end of their study involvement (18 at baseline, 22 at exit, with both completed by 12), about their preference of sampling method, current use of urine testing, and likelihood to use testing based on self-collected vaginal samples if it was available. The preference question for both participants and providers was a 5-point scale ranging from “strongly prefer health professional sampling” to “strongly prefer patient self-sampling,” with an option for “neutral.”
Written informed consent was obtained and the study was approved by the Research Ethics Board at Hamilton Health Sciences.
After enrollment, women were given verbal and written instructions and a diagram showing how to obtain a vaginal specimen. The women were asked to grasp the swab at its midpoint, gently inserting the rounded end into the vagina up to the point where the fingers held the swab device, and rotate the swab while removing it from the vagina. Vaginal swabs were collected using Abbott LCx STD Swab Specimen Collection System (Abbott Laboratories, Abbott Park, IL). The swab was then placed into the tube containing the transport medium. Self-sampling was performed by each participant in a private washroom. Specimens were refrigerated at 4°C and transported to the Chlamydiology Laboratory at St. Joseph's Healthcare within 24 hours.
Chlamydia LCR (LCx Chlamydia Abbott Diagnostics, Chicago, IL) testing of vaginal swabs was performed according to the manufacturer's instructions and has been described previously, 15–17 but was modified to include pooling. 18,19 All specimens were initially tested using pools of 4 vaginal swabs grouped according to study site. Pools that tested positive were reflexed into individual testing to determine which specimens in the pool were LCR-positive. LCR testing of specimens from women at the STD clinic and street health centers revealed that pooling was not appropriate as a result of high rates of infection. Testing was then performed on an individual basis for specimens collected at those sites only.
Data were entered into SPSS version 10.0.5 (Chicago, IL). Univariate analyses were used to determine associations between the demographic, past reproductive health and sexual behavior, and the presence of C. trachomatis infection. The sexual health centers servicing youth, the community-based street health center providing services to homeless adults, and the organizations providing services to homeless youth were grouped together for analyses (termed “street health centers”). Odds ratios (ORs) and 95% confidence intervals (CIs) for the associations between the independent variables and C. trachomatis infection were calculated using EpiInfo version 6.04b (Centers for Disease Control and Prevention, Atlanta, GA; and World Health Organization, Geneva, Switzerland). The Pearson's chi-squared or the Fisher exact test were used, as appropriate, to determine statistical significance. Statistical significance was considered to be a type I (alpha) error level of <0.05 (2-tailed).
Multivariate analysis with logistic regression was used to investigate the relationship of potential predictive factors with C. trachomatis infection while controlling for the effects of covariates. The variables found to be statistically significantly associated with C. trachomatis infection in the univariate analyses (age <20 years, attending a street health center, age of first intercourse 15 years or younger, 5 or more lifetime sexual partners, 2 or more new partners in the previous year, casual partner in the previous year, previous C. trachomatis infection, forced sex in the past, recent contact with someone with C. trachomatis, and presence of symptoms) were forced into a logistic regression model to determine which variables were independent risk factors for infection.
A total of 514 women participated. As a result of the nature of some of the sites and populations, and logistic barriers, we were unable to accurately capture the refusal rate. Of the 285 refusers recorded, age was stated for 256, and approximately two thirds (64.4%, 165 of 256) were 15 to 20 years of age. The mean age of refusers and participants was similar (P = 0.15). The most common reasons for refusal included lack of time (21.4%; 61 of 285) or being uncomfortable with the self-collection (18.2%; 52 of 285). No reason or lack of interest was reported by 29.8% (85 of 285) of refusers.
The demographics and sexual history of the participants at the different sites are shown in Table 1. Nearly half (48.0%; 247 of 514) of the women were under 20 years of age. The mean age of first intercourse was 16.7 years (standard deviation, 5.3 years) and the median number of lifetime sexual partners was 3.
Prevalence of C. trachomatis Infection and Putative Risk Factors for Infection
The overall prevalence of C. trachomatis infection detected in self-collected vaginal swabs was 6.0% (31 of 514). The highest was 18.2% (10 of 55) among women attending the street health centers (Table 2). The lowest prevalences were found in the student health centers (2.8%; 7 of 248) and the family physician offices (3.4%; 3 of 78). The majority of the street health center and birth control clinic participants were aged 20 years or younger. The putative risk factors of young age at first intercourse (15 years or younger) and having 5 or more lifetime sexual partners were reported by approximately half to three fourths of participants from the street health centers and the STD clinic. Participants in the STD clinic reported recent contact with someone with C. trachomatis infection and previous C. trachomatis infection most often.
Variables Associated With C. trachomatis Infection
The behavioral and demographic variables that were statistically significantly associated with C. trachomatis infection in women in the univariable analysis are shown in Table 2. Attending a street health center (OR, 4.6; 95% CI, 2.1–10.5), recent contact with someone with C. trachomatis infection (OR, 9.9; 95% CI, 4.7–20.7), and previous C. trachomatis infection (OR, 5.1; 95% CI, 2.1–12.5) were the strongest predictors of infection.
Multivariate analyses with logistic regression included the variables that were statistically significantly associated with C. trachomatis infection in the univariate analyses. Only recent contact with someone with C. trachomatis infection (adjusted OR, 7.1; 95% CI, 2.5–20.0) remained significantly associated with infection.
Patient Preference of Sampling Method
More participants (54.2%; 256 of 472) mildly or strongly preferred self-sampling and only 15.9% (75 of 472) mildly or strongly preferred physician sampling (Fig. 1). The remaining participants had no preference. The sites in which the largest proportion of participants preferred self-sampling were the street health centers (74.0%; 37 of 50), followed by the student health centers (58.1%; 132 of 227). Approximately half of the participants in the family physicians’ offices (46.3%; 37 of 80), birth control clinics (51.6%; 32 of 62), and the STD clinic (34.0%; 18 of 53) mildly or strongly preferred self-sampling. Significantly more participants mildly or strongly preferred self-sampling in the street health centers (74.0%; 37 of 50) compared with all other sites (51.9%; 219 of 422) (P = 0.005).
We did not ask specifically about reasons for choice of preference in this study; however, written comments at the end of the questionnaire suggested that some patients did not prefer self-sampling because they were having a Pap anyway, they wanted to be tested for all STDs, or they were experiencing a problem such as a urinary tract infection.
Provider Preference for Sampling Methods
Of the 34 physicians who were staffing the clinics (with the exception of the street health centers) during the study period, 29 physicians (85.3%) responded to a questionnaire either at baseline or exit, and 12 (35.3%) responded at both times, regarding their current use and preference for physician versus self-obtained swab collection. The majority of physicians (75.0%; 9 of 12) reported at the start and end of the study that they would use vaginal swab self-sampling in women, if available. Of the 9 providers at the start of the study who had no preference for testing method or who preferred physician testing, 55.6% (5 of 9) reported that they preferred self-sampling at the end of the study. Reasons for not preferring self-sampling given by the 25% (3 of 12) who would not use it were the desire to do a clinical examination, concern that the patient would not take the sample correctly, and the time required to give instructions for self-sampling.
The prevalence of infection among young women varied distinctly across the different settings in our communities, ranging from 2.8% among women attending student health services in the university or college to 18.2% among women attending the street health centers. The high prevalence of chlamydia among young women in the street health centers highlights the need for STD testing and treatment in this group. There was a high prevalence of symptoms (58.9%) and suspected recent sexual contact with infected partners (42.1%) in women attending the STD clinic, and these factors were likely the reason that these women presented to the clinic. However, these risk indicators were also found frequently (30.9% and 32.7%, respectively) among participants at the street health centers where these women were not presenting for testing.
We found that 72.1% of the participants from the street health centers preferred self-sampling to physician sampling, suggesting that increased C. trachomatis screening could be facilitated in this high-risk group if self-sampling was routinely available in settings where these women receive other services. We did not have information on how many women simultaneously or previously received pelvic examination; therefore, not all women could have been able to compare that experience with self-sampling. Previous studies have reported acceptability of vaginal swab self-sampling in young women. 16,17,20–23
The majority of physicians in the current study also reported that they would have patients perform vaginal self-sampling if it was available. Provider endorsement of self-sampling could be beneficial to improve testing coverage. Although physicians might be concerned about lost opportunity for annual cervical cytologic screening among sexually active young women, self-sampling for STDs could be used to avoid additional pelvic examinations in settings such as family practice or student health centers. In addition, self-sampling could be the only feasible method for accessing transient or street youth.
Cervical signs, frequent urination, younger age, and recent new sexual partner were associated with C. trachomatis infection in women in our previous study, 13 and similar factors have been reported elsewhere. 24 The prevalence of signs and symptoms in the current study could also be a reflection of selection bias, because participants were volunteers and could have been more likely to be tested for C. trachomatis if they believed they might be infected.
The use of NAA techniques such as LCR for testing for chlamydia could have contributed to the increased number of cases of C. trachomatis reported between 1997 and 1998 in Canada 2 and in Hamilton. It has been noted that these sensitive tests, which can be used on noninvasive specimens, have facilitated population-based studies that can be used to design better C. trachomatis control programs 25,26 through screening and could improve understanding of the risk factors for infection in different settings. Because NAA assays are more expensive than the previously used antigen detection tests, using them to screen for C. trachomatis establishes an initial substantial increase in costs for testing. 27 We found that the potential benefit of pooling suggested in previous studies 18,19 was in fact useful and decreased the number of tests that would have been performed in the low-prevalence settings.
There were several limitations in this study. The sample size was small and there were a considerable number of refusers, who could have been averse to self-sampling. Although some of the women stated discomfort with self-sampling as the reason for refusal, many women stated lack of interest or time. Participation by women who were willing to engage in self-sampling could have biased our results in favor of preference for self-sampling. It is possible that if vaginal self-testing was offered routinely, especially to transient or homeless youth, that many young women would become more familiar with collecting their own specimen and agree to be tested. It was difficult to track precisely the numbers of women who volunteered, were not approached, and were approached but refused, as a result of differences between sites in staffing. In some situations, for example, if a woman presented at the STD clinic for testing as a result of sexual assault, it was not deemed appropriate to approach her for a research study. It was usually not possible for clinic staff to assess eligibility and approach all potential participants as a result of manpower issues. This is a limitation of the study which could limit the generalizability of the results and could overestimate preference for self-sampling. As pointed out in a recent study of C. trachomatis screening of homeless youth with a 20% acceptance rate, 28 staffing is usually not adequate in settings attended by these youths to approach all potential participants and keep accurate records of refusals.
In the current study, there was a high prevalence of C. trachomatis infection in some settings, yet we were unable to determine independent risk factors for infection, possibly as a result of the relatively high prevalence of many of the common symptoms and risk factors reported by participants. Despite differing risk profiles in the different sites, we combined all participants in our analysis as a result of the limited sample size and the lack of a priori hypotheses regarding which groups would be appropriate to combine. Universal screening of asymptomatic women has been found to be cost-effective in some settings, 29,30 but selective screening of sexually active young women or men is usually more effective. 11,13,24,25,27 The results of this study reveal the pattern of C. trachomatis in young women in different settings, and provide useful information for updating and examining the cost-effectiveness of local screening strategies. The high prevalence of C. trachomatis infection and acceptability of self-sampling techniques in the current study supports opportunistic screening among sexually active young women attending street health centers using a highly sensitive test such as LCR to detect asymptomatic infection.
1. Davies HD, Wang EE. Period health examination, 1996 update: 2. Screening for chlamydial infections. Canadian Task Force on the Periodic Health Examination. Can Med Assoc J 1996; 154: 1631–1644.
2. Health Canada. Canadian sexually transmitted disease surveillance report. Can Commun Dis Rep 2000; 26–16: 133–140.
3. Chernesky MA, Jang D, Sellors J, et al. Urinary inhibitors of polymerase chain reaction and ligase chain reaction and testing of multiple specimens may contribute to lower assay sensitivity for diagnosing Chlamydia trachomatis
infected women. Mol Cell Probes 1997; 11: 243–249.
4. Poulin C, Alary M, Bernier F, et al. Prevalence of Chlamydia trachomatis
and Neisseria gonorrhoeae
among at-risk women, youth sex workers, and street youth attending community organizations in Quebec City, Canada. Sex Transm Dis 2001; 28: 437–443.
5. Sellors J, Howard M, Pickard L, et al. Chlamydial cervicitis: testing the practice guidelines for presumptive diagnosis. Can Med Assoc J 1998; 158: 41–46.
6. Chernesky M, Chong S, Jang D, et al. Ability of commercial ligase chain reaction and PCR assays to diagnose Chlamydia trachomatis
infections in men by testing first-void urine. J Clin Microbiol 1997; 35: 982–984.
7. Chernesky M, Jang D, Lee H, et al. Diagnosis of Chlamydia trachomatis
infections in men and women by testing first-void urine by ligase chain reaction. J Clin Microbiol 1994; 32: 2682–2685.
8. Lee HH, Chernesky MA, Schachter J, et al. Diagnosis of Chlamydia trachomatis
genitourinary infection in women by ligase chain reaction assay of urine. Lancet 1995; 345: 213–216.
9. Jang D, Sellors J, Mahony J, et al. Effects of broadening the gold standard on the performance of a chemiluminometric immunoassay to detect Chlamydia trachomatis
antigens in centrifuged first void urine and urethral swab samples from men. Sex Transm Dis 1992; 19: 315–319.
10. City of Hamilton, Social, Public Health Services Department. Social and Health Issues Report; 2001.
11. Hillis S, Wasserheit J. Screening for chlamydia—a key to the prevention of pelvic inflammatory disease. N Engl J Med 1996; 334: 1399–1400.
12. McGregor JA, French FI. Chlamydia trachomatis
infection during pregnancy. Am J Obstet Gynecol 1991; 164: 1782–1789.
13. Sellors JW, Pickard L, Gafni A, et al. Effectiveness and efficiency of selective vs universal screening for chlamydial infection in sexually active young women. Arch Intern Med 1992; 152: 1837–1844.
14. Chernesky MA, Lee H, Schachter J, et al. Diagnosis of Chlamydia trachomatis
urethral infection in symptomatic and asymptomatic men by testing first-void urine in a ligase chain reaction assay. J Infect Dis 1994; 170: 1308–1311.
15. Hook EW, Smith K, Mullen C, et al. Diagnosis of genitourinary Chlamydia trachomatis
infections by using the ligase chain reaction on patient-obtained vaginal swabs. J Clin Microbiol 1997; 35: 2133–2135.
16. Wiesenfeld HC, Lowry DLB, Heine RP, et al. Self-collection of vaginal swabs for detection of chlamydia, gonorrhea, and trichomoniasis: opportunity to encourage sexually transmitted disease testing among adolescents. Sex Transm Dis 2001; 28: 321–325.
17. Smith K, Harrington K, Wingood G, et al. Self-obtained vaginal swabs for diagnosis of treatable sexually transmitted diseases in adolescent girls. Arch Pediatr Adolesc Med 2001; 155: 676–679.
18. Krepel J, Patel J, Sproston A, et al. The impact on accuracy and cost of ligase chain reaction testing by pooling urine specimens for the diagnosis of Chlamydia trachomatis
infections. Sex Transm Dis 1999; 26: 504–507.
19. Kapala J, Copes D, Sproston A, et al. Pooling cervical swabs and testing by ligase chain reaction are accurate and cost-saving strategies for diagnosis of Chlamydia trachomatis
. J Clin Microbiol 2000; 38: 2480–2483.
20. Sellors JW, Lorincz AT, Mahony JB, et al. Comparison of self-obtained vaginal, vulvar, and urine samples with physician-collected cervical samples for human papillomavirus testing to detect high-grade squamous intraepithelial lesions. Can Med Assoc J 2000; 163: 513–518.
21. Serlin M, Shafer M, Tebb K, et al. What sexually transmitted disease screening method does the adolescent prefer? Arch Pediatr Adolesc Med 2002; 156: 588–591.
22. Holland-Hall CM, Wiesenfeld HC, Murray PJ. Self-collected vaginal swabs for the detection of multiple sexually transmitted infections in adolescent girls. J Pediatr Adolesc Gynecol 2002; 15: 307–313.
23. Newman SB, Nelson MB, Gaydos CA, et al. Female prisoners’ preferences of collection methods for testing for Chlamydia trachomatis
and Neisseria gonorrhoeae
infection. Sex Transm Dis 2003; 30: 306–309.
24. Miller WC. Screening for chlamydial infection: a model based on prevalence. Sex Transm Dis 1998; 25: 201–209.
25. Stamm WE. Expanding efforts to prevent chlamydial infection. N Engl J Med 1998; 339: 768–769.
26. Chernesky MA. Chlamydia trachomatis
diagnostics. Sex Transm Infect 2002; 78: 232–234.
27. Goeree R, Jang D, Blackhouse G, et al. Cost-effectiveness of screening swab or urine specimens for Chlamydia trachomatis
from young Canadian women in Ontario. Sex Transm Dis 2001; 28: 701–709.
28. Van Leeuwen JM, Rietmeijer CA, LeRoux T, et al. Reaching adolescent youths for Chlamydia trachomatis
and Neisseria gonorrhoeae
screening in Denver, Colorado. Sex Transm Infect 2002; 78: 357–359.
29. Howell MR, Gaydos JC, McKee KT Jr, et al. Control of Chlamydia trachomatis
infections in female army recruits: cost-effective screening and treatment in training cohorts to prevent pelvic inflammatory disease. Sex Transm Dis 1999; 26: 519–526.
30. van Valkengoed IG, Postma MJ, Morre SA, et al. Cost effectiveness analysis of a population based screening programme for asymptomatic Chlamydia trachomatis
infections in women by means of home obtained urine specimens. Sex Transm Infect 2001; 77: 276–282.